This text is made available for information purposes only.

A summary of this decision is published in all Community languages in the Official Journal of the European Union.

Case No COMP/M.5658 – UNILEVER/SARA LEE

Only the English text is authentic.

REGULATION (EC) No 139/2004

MERGER PROCEDURE

Article 8 (2)

Date: 17/11/2010

EUROPEAN COMMISSION

Brussels, 17.11.2010

C(2010) 7934final

PUBLIC VERSION

COMMISSION DECISION

of 17.11.2010

declaring a concentration to be compatible with the internal market and the EEA Agreement

(Case No COMP/M.5658 – Unilever/Sara Lee Body Care)

2

COMMISSION DECISION of 17.11.2010

declaring a concentration to be compatible with the internal market

and the EEA Agreement

(Case No COMP/M.5658 – Unilever/Sara Lee Body Care)

(Only the English text is authentic)

(Text with EEA relevance)

THE EUROPEAN COMMISSION,

Having regard to the Treaty on the Functioning of the European Union,

Having regard to the Agreement on the European Economic Area, and in particular Article 57 thereof,

Having regard to Council Regulation (EC) No 139/2004 of 20 January 2004 on the control of 1concentrations between undertakings, and in particular Article 8(2) thereof,

Having regard to the Commission's Decision of 31 May 2010 to initiate proceedings in this case,

Having given the undertakings concerned the opportunity to make known their views on the objections raised by the Commission,

2Having regard to the opinion of the Advisory Committee on Concentrations,

3Having regard to the final report of the Hearing Officer in this case ,

1 OJ L 24, 29.1.2004, p. 1. With effect from 1 December 2009, Articles 81 and 82 of the EC Treaty have become Articles 101 and 102, respectively, of the Treaty on the Functioning of the European Union ("TFEU"). The two sets of provisions are, in substance, identical. For the purposes of this Decision, references to Articles 101 and 102 of the TFEU should be understood as references to Articles 81 and 82 of the EC Treaty where appropriate. The TFEU also introduced certain changes in terminology, such as the replacement of "Community" by "Union" and "common market" by "internal market". The terminology of the TFEU will be used throughout this Decision.

2 OJ C

WHEREAS:

(1) On 21 April 2010, the Commission received a notification of a proposed concentration pursuant to Article 4 of Council Regulation (EC) No 139/2004 (hereinafter referred to as "the Merger Regulation") by which the undertakings Unilever N.V., the Netherlands, and Unilever Plc, United Kingdom, (hereinafter collectively referred to as "Unilever" or as "the notifying party") acquire control of the whole of the undertaking Sara Lee Household and Body Care International, the Netherlands (hereinafter referred to as "Sara Lee"), belonging to Sara Lee Corporation, USA (hereinafter referred to as "Sara Lee Corp"), within the meaning of Article 3(1)(b) of the Merger Regulation, by way of irrevocable binding offer announced on 25 September 2009. Unilever and Sara Lee are collectively referred to as "the Parties" throughout this Decision.

(2) After examination of the notification, the Commission adopted a Decision on 31 May 2010 (hereinafter referred to as “the 6(1)(c) Decision”), concluding that the operation falls within the scope of the Merger Regulation and raises serious doubts as to its compatibility with the internal market and the functioning of the EEA Agreement and initiated proceedings pursuant to Article 6(1)(c) of the Merger Regulation.

(3) On 12 August 2010, a Statement of Objections was sent to Unilever pursuant to Article 18 of the Merger Regulation.

(4) Unilever replied to the Statement of Objections on 27 August 2010.

(5) At the request of Unilever, the time-period for taking a final decision in this case was extended by 13 working days on 26 August 2010, pursuant to the second subparagraph of Article 10(3) of the Merger Regulation.

(6) The time-period for taking a final decision in this case was further extended by two working days on 16 September 2010, pursuant to the second subparagraph of Article 10(3) of the Merger Regulation.

(7) On 21 September 2010, Unilever offered commitments with a view to rendering the proposed concentration compatible with the internal market. Those commitments were modified and the final version of the commitments was submitted to the Commission on 12 November 2010.

(8) The time-period for taking a final decision in this case was again further extended by five working days on 11 October 2010, pursuant to the second subparagraph of Article 10(3) of the Merger Regulation.

I. THE PARTIES

4(9) Unilever, an Anglo-Dutch company, is a worldwide supplier of fast-moving consumer goods and has primary share listings on Euronext Amsterdam (through

3 OJ C

4 Unilever has a dual-listed structure comprising Unilever N.V. and Unilever Plc. The two entities exist as separate companies but operate as a single economic unit.

2

Unilever N.V.) and the London Stock Exchange (through Unilever Plc). Its principal businesses are in the food, home care and personal care categories. In the home care sector, Unilever is a leading supplier of products for fabric and surface cleaning and hygiene. Unilever’s personal care division supplies deodorants, bath and shower products, skin care products, oral care products and hair care products.

(10) Sara Lee Corp is a worldwide supplier of branded consumer goods, operating in the meats, bakery, beverage and household and body care sectors with its headquarters in the United States of America and listed on the New York and Chicago Stock Exchanges. The target company - Sara Lee - is comprised of the global body care business which manufactures and supplies bath and shower products, deodorants, baby care products, men’s toiletries and oral care products worldwide on one hand and the European laundry care business supplying fabric cleaning and conditioning products and laundry aids on the other.

II. THE OPERATION AND THE CONCENTRATION

(11) On 25 September 2009, Unilever made an irrevocable binding offer for Sara Lee Corp’s worldwide body care and European laundry care businesses. The acquisition by Unilever is structured as a number of purchases of shares and assets comprising Sara Lee Household and Body Care International, as set out in the draft Sale and Purchase Agreement (the “SPA"). Sara Lee Corp agreed to accept the binding offer subject to consultation with relevant employee representatives, which have been fulfilled.

(12) Since after completion of the notified transaction Unilever will own all the shares and assets of Sara Lee, the proposed transaction constitutes a concentration within the meaning of Article 3(1)(b) of the Merger Regulation.

III. UNION DIMENSION

5(13) The undertakings concerned have a combined aggregate world-wide turnover of more 5 than EUR 5 billion(Unilever: EUR 39 823 million, Sara Lee: EUR 756 million). Both undertakings have a Union-wide turnover in excess of EUR 250 million ∗(Unilever: [...], Sara Lee: [...]*), but they do not achieve more than two-thirds of their aggregate Union-wide turnover within one and the same Member State. The notified operation therefore has a Union dimension.

5 Turnover calculated in accordance with Article 5(1) of the Merger Regulation.

∗ Parts of this text have been edited to ensure that confidential information is not disclosed; those parts are enclosed in square brackets and marked with an asterisk.

3

IV. COMPETITIVE ASSESSMENT

IV.1. Introduction

6(14) Unilever and Sara Lee are both active in the supply of personal care and home care 6products. According to the Parties, their activities overlap in the following categories: deodorants, skin cleansing (products for personal washing like bath and shower and soap products), skin care (products designed to moisturise and nourish skin on the hands and body), fabric care (detergents, conditioners and laundry aids), aftershave treatments, oral care (toothpaste), hair care (shampoo, conditioners and styling products), and household cleaning (multi-purpose cleaners). An overview of the 7Parties' market shares in different product categories is provided in Table 1.

Table 1: Overview of Parties' Market Shares in Member States 2008 – Source: Form CO

6 See Form CO, paragraph 6.3.

7 Table 1 only provides a broad overview based on the Form CO, in particular the categories mentioned in the table should not be interpreted as relevant product markets.

4

(15) All the segments affected by the transaction have similar features, being differentiated product markets where the customers of the Parties are mainly large retailers. However, the degree of product differentiation is highest in deodorants, followed by bath&shower products, with fabric care products being slightly less differentiated. Correspondingly, private label products have a lower share in deodorant markets compared to bath & shower products or detergents.

(16) In each affected market, two stages exist in the sales process, with manufactures such as Unilever and Sara Lee supplying products to retailer customers and then final consumers purchasing products from retailers.

IV.2. Deodorants

The deodorant category: a differentiated market and dominated by brands

5

(17) Deodorants are products which minimise or eliminate the negative effects of sweating through control of odour and/or wetness.

(18) The deodorant category in the Union has a total value of approximately EUR 2.7 billion (2009) and has been growing constantly over the last ten years. According to the Parties, value growth was between [5-10]*% in 2008. Looking at the national markets covered in more detail in section IV.2.3.2, the United Kingdom is the largest national market with value sales to final customers of [...]*, followed by Germany ([...]*) and France ([...]*), Spain ([...]*) and the Netherlands ([...]*) and the smaller markets are Belgium ([...]*), Portugal ([...]*), Denmark ([...]*) and Ireland ([...]*).

(19) Deodorants are increasingly differentiated according to gender and most brands have variants marketed specifically to male and female consumers (hereinafter referred to as "male variants" and "female variants"). Certain brands are entirely dedicated to either male or female consumers (such as Axe which is targeting male consumers only). Male-specific and female-specific deodorants (hereinafter referred to as "male deodorants" and "female deodorants") can generally be distinguished by their fragrances, marketing strategies and packaging (typically white and pale coloured packaging for female deodorants and black, dark blue or silver packaging for male- deodorants). In some Member States (especially Spain) a "unisex" category exists, with products appealing to both male and female consumers. Female deodorants together with unisex deodorants are collectively referred to as "non-male deodorants" (and similarly, female variants together with unisex variants are referred to as "non-male variants") throughout this Decision.

(20) Deodorants are also sold in different formats. The notifying party distinguishes between contact (primarily roll-ons, creams, stick and wipes) and non-contact formats. Most of the sales of a brand often derive from one specific format, although the brand is available also in other formats.

(21) Deodorant brands are typically associated with certain core functionalities, the three 8main ones being "efficacy", "skin caring" or "fragrance". In practice, brand equity is built under one of these three main propositions although the brand may exist in variants claiming properties extending beyond their core brand proposition. Brands, or variants of 9 the brand, may also claim value-added benefits such as "no white marks"or hair minimisation or be particularly targeted to a certain market segment such as teenage girls.

(22) Finally, in relation to pricing, brands occupy a wide range of price positions falling into three broad categories according to their relative positions: premium-priced, mid- priced and low-priced brand. Pricing varies in national markets; in general non-contact deodorant variants are sold at prices between EUR 7 and EUR 40 per litre whereas contact deodorants are more expensive (on average between EUR 20 and EUR 80 per

8 Efficacy in terms of credentials for minimizing sweat and odour.

9 "no white marks" is a term used in the industry to describe a deodorant with credentials for not leaving any white marks/traces on the consumer's clothing.

6

10litre) . Such a large price spectrum highlights the differentiated nature of the deodorant market.

The brands sold by the Parties

(23) Sara Lee’s leading brand is Sanex, which is present in all the national markets analysed in detail in section IV.2.3.2. Its marketing focuses primarily on the promises of healthy skin, although customers also value Sanex for its efficacy credentials. The focus on health is exemplified by the launch of the Sanex “Nature Protect” range which contains natural mineral alum. Sanex has a female bias, only a relatively small proportion of its sales ([20-30]*%) are achieved with male variants. Sanex is sold as an antiperspirant aerosol and as a roll-on (tending to be stronger in roll-ons) and, in some Member States, in the stick format. In terms of price points, Sanex normally occupies a mid-price position.

(24) Sara Lee also supplies deodorants under other brands including Radox (United Kingdom and Ireland), Williams (Belgium, Denmark, France and Spain), Duschdas (Germany); Monsavon (France) and Neutral (Denmark, the Netherlands and Sweden). These brands are much less important in the Sara Lee deodorant portfolio compared to the main brand Sanex.

(25) Unilever has three core Union-wide brands: Axe, Rexona and Dove. In addition to the core brands, it has two brands present in a certain national markets: Vaseline (Vasenol in Portugal) and Impulse.

(26) Unilever's largest brand, Axe (known as Lynx in the United Kingdom and Ireland) is an exclusively male deodorant. It is heavily targeted at young males (between 15 and 25) with strong efficacy claims and a distinctive black/silver packaging. Its marketing is based upon its "seductive fragrances". Its self-professed mission is to give the Axe consumer the edge in attracting the opposite sex with its so-called "Axe effect". Axe is sold predominantly as an aerosol/spray although it is also available in a roll-on format. Axe is a premium-priced product.

(27) Rexona (known as Sure in the United Kingdom and Ireland) is positioned as an efficacy product. Its marketing proposition is its ability to continue to work even during heavy sweating activities such as sports. Rexona also has more neutral propositions promoting skin care and more fragrance based variants targeting teenage girls. Like Sanex, Rexona is available in male and female variants although its credentials are stronger in the female- area. It is mainly sold in aerosol format but it is also available as an antiperspirant roll-on and stick. Rexona is a mid-priced product (on average [...]* cheaper than Axe).

(28) Dove deodorants (part of the wider Dove personal care portfolio) were initially launched in 1998 and have a particular focus on superior moisturising to fight skin dryness. Dove tends to have a stronger position in aerosols although it is also available in contact formats such as roll-ons and sticks. The brand used to be exclusively

10 Form CO, Annex DEO 6.14 "Finnegan – Deodorant price charts". Prices for deodorants in a contact- format (format like stick or roll-on which is in direct contact with the skin, as opposed to sprays which are a non-contact form) are lower in the United Kingdom (between EUR 10 and EUR 40 per litre) and significantly higher in France (between EUR 50 and EUR 95 per litre).

7

targeted at women but a range of products under the Dove Men+Care brand, including deodorants, was launched across several Member States in the Union in January 2010. Dove is more of a premium product targeted for sale at a [...]* price premium to Rexona.

(29) Vaseline is marketed with a strong focus on skin health and contains an ingredient which mimics the skins own protection layer and makes the skin less prone to irritation. Vaseline is available as an aerosol spray, roll-on and cream. Vaseline products have a female bias but Vaseline for Men was launched in some national markets in early 2008 and the range became more widely available in early 2009. 11[...]* .

(30) Impulse is a fragranced body spray deodorant marketed to teenage girls and younger women. Unilever's intention is to target women with a young, fun and sexy image. This range, marketed only in some Member States (the United Kingdom and Ireland, the Netherlands, Spain and Portugal), is priced as a premium product.

Main competitor brands in the EEA

(31) Beiersdorf markets Nivea, one of the main deodorant brands in the EEA. Nivea deodorants exist in both male and female variants ([40-50]*% of all Nivea deodorant sales are in male variants in the markets affected by the transaction). The brand is available in all main formats but has a particularly strong market presence in roll-ons. Nivea deodorants for the female market are sold with a skin-caring proposition, using as a marketing argument "smooth and beautiful underarms, ease of shaving and softer skin". Nivea deodorants generally occupy a mid-market price position.

(32) Henkel is an established supplier in the deodorant market with its brand Fa. Fa is primarily a female-oriented brand marketed under a fragrance based/feminine proposition. In spring 2008, Fa included more natural lines in its product range such as Fa Rice Dry and Fa Natural. It is positioned as a lower priced product. Through the supply of the Right Guard brand in the United Kingdom and Ireland, Henkel has also developed a position in male products in those Member States, orientated towards efficacy. The brand also has a female variant. Right Guard products are marketed with a strong emphasis on efficacy throughout the day and a mid-price positioning.

(33) Colgate-Palmolive includes brands such as Palmolive and Soft&Gentle in its deodorant range. Both brands are fragranced products although Palmolive deodorant is also sold under a skin-friendly proposition and the Soft&Gentle product range includes variants targeted at efficacy and skin care. Marketing of Palmolive and Soft&Gentle is directed at the female market and both brands occupy a mid-price positioning. In addition, the company is marketing deodorants under the Mennen brand in Central Europe.

(34) L’Oreal is active in a number of Union markets as a supplier of deodorants with brands such as Narta, Ushuaïa and Garnier Mineral. Narta is a mid-priced antiperspirant with a focus on efficacy and freshness. It targets the female market but also includes male variants. Ushuaïa is a fragrance deodorant prominently aiming at female customers in Belgium and France with a mid-low price positioning. L’Oreal's product range also

11 [...]*

8

includes Airness, which is a male deodorant with a similar positioning as Axe, and Garnier Mineral, recently launched in a number of Union markets with a skin-friendly brand positioning. In addition, L'Oreal has a licence from Colgate-Palmolive to use the Mennen brand in Belgium and France.

(35) Procter & Gamble markets the brands Mum, Secret, Gillette and Old Spice. Those brands currently have a very limited presence in the markets concerned except in Ireland and, to a lesser extent, in the United Kingdom and Portugal. Revlon supplies the Mitchum brand (male and female) in the United Kingdom and Ireland with an efficacy- orientated proposition and a premium-price positioning. Coty is active in a number of Union markets with its brand Adidas. Adidas is a mid-priced anti-perspirant focused on effective protection for highly active people, with a particular focus on sports. It has variants for women within its product range although the brand appeals more to male consumers. Coty has also recently launched the Playboy brand in several Member States, which is a fragranced deodorant targeted at young males, with a lower price positioning than Axe.

IV.2.1. Relevant Product Market

(36) The notifying party submits that there is only one market for all types of deodorants hence it is not appropriate to define relevant product markets more narrowly based on main differentiating factors such as gender or format.

(37) In previous decisions dealing with deodorants, the product market definition has been 12left open, without delineating the market according to further sub-segmentation . 13However, in the more recent Commission Decision, it was suggested that there may be a separate "men's care" market including deodorants for men. There was no need to reach a definitive conclusion in that case.

(38) Generally, deodorants display a mix of various differentiating features - most notably a differentiation according to gender and according to formats (spray, roll-on, stick) - but also features such as fragrances, propositions and functional benefits (like emphasis on skin care, efficacy, fragrance, "no white marks" features, antiperspirant function), and pricing positioning differ. Consistent with the previous indications about specific men's products, the market investigation in this case has confirmed the prominent role of gender segmentation for deodorants. Balancing supply and demand side factors, the conclusion of the investigation leads to a separation of the deodorants category into a male and a non-male market, while other distinctions (notably according to format) do not warrant such separation in terms of product market definition.

IV.2.1.1. Male and non-male deodorants

(39) In the Member States concerned, deodorants are essentially marketed as either male deodorants or as non-male deodorants (non-male may, in some Member States, include deodorants marketed or perceived as unisex, in addition to female). Deodorants targeted

12 Commission Decision of 15 July 2005, M.3732 - Procter & Gamble/Gillette; Commission Decision of 18 October 1995, IV/M.630 – Henkel/Schwarzkopf; Commission Decision of 23 February 1992, IV/M.186 – Henkel/Nobel.

13 Commission Decision of 20 April 2006, case COMP/M.4193 L'Oreal/The Body Shop.

9

specifically at male consumers tend to be marketed using dark colours and more masculine fragrances and marketing tends to be centred around masculine attributes deploying efficacy, sports or sex-appeal features. Non-male deodorants, on the other hand, typically use softer colours on packaging, feminine (or neutral) fragrances. Some offer skin caring benefits or specific added benefits such as "no white marks" or hair minimizing features, besides a basic efficacy proposition.

(40) The investigation has not confirmed the market definition proposed by the Parties whereby male deodorants form part of the same relevant product market as non-male deodorants. Instead, the investigation has shown that male and non-male deodorants are two distinct product markets.

Male deodorants are marketed differently from non-male deodorants and have different features

(41) Deodorants are branded goods where the brand plays a prominent role in competition between deodorant suppliers. Within the deodorants category, the male-branded deodorants are clearly distinct from non-male-branded deodorants.

(42) The brands used for male deodorants essentially fall into two categories. The first group are strongly male-oriented brands such as Axe/Lynx, Mennen, Adidas, Gillette, Williams, and L'Oreal Men expert. Some of those brands have obtained their strong masculine connotation from deployment in neighbouring personal care markets (for example Gillette or Williams from shaving). Those specific male brands represent typically [50-60]*%-[60-70]*% of the total male deodorant markets.

(43) The second type of male deodorants is the specific male variant of brands traditionally present in non-male variants such as Rexona for Men, Nivea for Men, Fa for Men, or the recently launched Unilever's Dove for Men. These male variants are clearly differentiated from their non-male counter-parts in terms of appeal, fragrance, marketing campaigns and packaging. Figure 1 shows how marketing aims to differentiate the product presentation by using, for example, darker or metallic colours and the addition of the word "men" on the product.

Figure 1 – pictures of male and non-male/female variants of some deodorants brands

10

(44) Although it is true that both male and non-male deodorants have similarities in terms of the core function of preventing odour and sweating, one has to consider that the perception of these kinds of branded consumer goods is heavily shaped by their marketing image. Deodorants are a category on which massive advertisement and marketing costs are spent; brand image and the proposition a product offer are key to creating a relationship with the actual product for the end-consumers and their expectations. Suppliers are building on this trend and are developing bespoke products specifically designed, composed and marketed towards male consumers.

(45) Consistent with the efforts of deodorant suppliers to clearly differentiate male deodorants from female/unisex deodorants, the market investigation shows that a great majority of retailers consider that there are significant differences between male and non-male deodorants; many of them pointing to differences in terms of packaging and to fragrance, and to different advertising/marketing campaigns, each focused on the 14respective target group.

Gender differentiation between male-deodorants and non-male-deodorants is strong across Member States and is increasing

(46) The trend is towards further gender differentiation, which is already significant in the deodorant category with male deodorants forming a large proportion of all deodorants. According to figures provided by Unilever, a significant proportion ([40-50]*%) of all deodorants sold in the EEA are male-specific deodorants.Looking at individual Member States which were analysed in more detail in the investigation, the proportion of male deodorants in the relevant markets varies between [30-40]*% in Spain to [50-60]*% in the United Kingdom and Ireland (see Table 3).

Table 2 - 2009 EEA sales of deodorants by gender – Source: Form CO. Turnover (EUR 000) Percentage of total Female [...]* [40-50]*% Male [...]* [40-50]*% Unisex/Other [...]* [10-20]*% Total 2 748 062 100%

Table 3 - 2009 sales of deodorants by gender per Member State – Source: Form CO. Non-male

Male Female Unisex

Ireland [50-60]*% [50-60]*%

United 16 Kingdom[50-60]*% [50-60]*%

14 Questionnaire to customers on deodorants sent on 23 April 2010, question 15.

15 Form CO, paragraph 6.109.

16 Total market size of all deodorants combined is larger than the sum of the gender segments, as the Parties were not able to provide the full data due to the limits of Nielsen database (see Note to Table 6.24 in the

Belgium [40-50]*% [50-60]*%

Denmark [40-50]*% [30-40]*% [10-20]*%

Netherlands [30-40]*% [40-50]*% [10-20]*%

17 Portugal[40-50]*% [50-60]*%

Spain [30-40]*% [10-20]*% [40-50]*%

(47) The trend towards greater gender differentiation (and male-specific products) was 18 mentioned by the majority of the respondents in the market investigation.Recent male-specific brand introductions confirm that trend; for example Unilever's Dove Men+Care, L'Oreal's Men expert brand, Procter & Gamble's Gillette, and Coty's Playboy, all of which are new brands exclusively targeted at male consumers. The Parties explain that "manufacturers strive for product differentiation by offering 19distinct products for men and women" as part of the drive to increase sales.

(48) Unilever, in its reply to the Statement of Objections, stated that it does not agree with the distinction between male and non-male deodorants. Unilever argues that a lot of brands have both male and female variants and puts forward that a third or more of the male market is made up of brands that are present in both male and non-male markets (like Nivea or Rexona) and that these brands take up half or more of the non-male market.

(49) That argument cannot be upheld. First, Unilever itself acknowledges that up to two-thirds of the male market is composed of specific male-oriented brands which have no presence or only a very small presence in the non-male market. Such a proportion is significant and has to be taken into account. Second, brands like Rexona or Nivea which are present on both gender markets are clearly split into male and female brand variants which are marketed differently, even if they have a common umbrella brand.

(50) Unilever also submits that differences in packaging and product associations do not necessarily imply separate markets, putting forward that roll-on and sprays are also packaged differently, and have different added benefits (such as leaving no white marks or products that care for the skin) are also marketed differently. Unilever submits that marketing alone cannot justify separate male/non-male markets, and that marketing varies even within a gender group (Rexona "Girl" variant has different packaging than Rexona woman/men).

(51) Unilever's arguments have to be put into perspective. It is not argued by the Commission that marketing differences alone justify separate markets. Separate 20markets result from absence of demand-side and supply-side substitutability. However, marketing/branding of branded goods like deodorants is an important factor in both the consumer choice and the suppliers' abilities to successfully put their products on the market. It is not contested that marketing varies even within a gender segment as the Rexona Girl example shows. However, Rexona Girl is a new niche product, trying to address specific consumers within a gender group (like Dove pro age for mature women). On the contrary, male deodorants are now a well-developed concept and an established category, which represents up to half of the total deodorant sales in some Member States.

Unisex deodorants are marketed distinctly from male deodorants

(52) As shown in Table 2 and Table 3, in some Member States – notably in Spain – there is a category of gender-neutral 'unisex' deodorant products which are distinct from male-specific products but are not marketed as female deodorants. Such unisex products typically take a neutral approach with respect to packaging, marketing and fragrances and sometimes may have a more clinical appeal – for example the non-male variants of Sanex in Spain are considered to be unisex.

(53) While unisex deodorants are popular in Spain (and to a lesser extent in some other Member States), the unisex segment is not evident in other national markets. In Belgium for example, none of the retailers classified any deodorant brand as being 'unisex', and Belgian retailers have instead indicated Sanex (presumably the non-male 21 variants) as belonging to the female-category.[reference to parties’ internal 22 23documents]or that [reference to parties’ internal documents] .

(54) In any event, even if it is more meaningful in some Member States than others to consider a unisex category, brands with more neutral proposition are seen either as female-oriented or unisex depending on the Member State, and even depending on the particular retailer or supplier within a Member State. This is demonstrated in detail in the part specific to Spain (see Section IV.2.3.2.6). Indeed, the boundaries between unisex and female deodorants are very unclear, and the attribution of a particular product into one or another category can be arbitrary. In contrast, male deodorants are 24always clearly distinct from their non-male (female or unisex) counter-parts.

The competitive landscape differs significantly for male deodorants and non-male deodorants, and the constraints on Unilever are clearly different

(55) The competitive landscape of the male and non-male deodorant market differs significantly. Male deodorants include a number of brands which are either not at all or marginally competing in the non-male market. The prime example is Unilever's Axe/Lynx brand as an exclusively male brand. Axe/Lynx is the clear market leader in a number of Member States, representing alone up to the half of the total male deodorant market (Ireland [50-60]*%, United Kingdom [40-50]*%, Belgium [40- 50]*%, the Netherlands [40-50]*%, Spain [30-40]*%). A similar argument can be made with Unilever's leading Dove brand, which has strong positions in a number of non-male markets, whereas it currently only has a very small presence in male market.

(56) Although a number of competitors are present in both male and non-male deodorants, their strength significantly varies between those two. Unilever is the leader in almost all the Member States analysed in the Statement of Objections (except Denmark) in both male and non-male markets. However, the nature and strength of the main competitors varies significantly between male and non-male markets. The market shares of Unilever and its three main competitors in the male and non-male markets are presented in Table 4.

Table 4: Comparison of market shares of Unilever and its three main competitors on male and non-male deodorants markets (value, 2009) - Source:Form CO. Male Non-male

Unilever inthree mainUnileverthree main competitors %competitors in % in % in %

Belgium [50-60]*% Beiersdorf: [20-30]*%[30-40]*% Henkel: [10-20]*%

Coty: [5-10]*%

Sara Lee [10-20]*%

L'Oreal: [5-10]*%

Beiersdorf: [10-20]*%

Denmark [10-20]*% Unicare: [20-30]*%[20-30]*% Sara Lee [20-30]*%

Saether A/S: [10- 20]*%

Unicare: [10-20]*%

Beiersdorf: [10-20]*%

Beiersdorf: [10-20]*%

Ireland [70-80]*% Beiersdorf: [5-10]*%[60-70]*% Beiersdorf: [10-20]*%

Colgate: [10-20]*%

Procter & Gamble: [0- 5]*%

Coty: [0-5]*% Sara Lee [5-10]*%

Netherlands [40-50]*% Beiersdorf: [20-30]*%[30-40]*% Beiersdorf: [10-20]*%

Coty: [0-5]*%

Sara Lee: [10-20]*%

Henkel: [0-5]*%

Henkel: [5-10]*%

Portugal [40-50]*% Beiersdorf: [30-40]*%[40-50]*% Beiersdorf: [20-30]*%

L'Oreal: [0-5]*%

L'Oreal: [10-20]*%

Conter: [0-5]*%

Sara Lee: [5-10]*%

Spain [50-60]*% Sara Lee [10-20]*%[20-30]*% Sara Lee [20-30]*%

Beiersdorf: [10-20]*%

Private labels: [20-30]*%

G. Puig: [5-10]*%

G.Puig: [5-10]*%

United [60-70]*% Henkel: [5-10]*% Kingdom

[50-60]*% Colgate: [5-10]*%

Coty: [5-10]*%

Revlon: [5-10]*%

Procter & Gamble:[5- 10]*%

Sara Lee [5-10]*%

(57) The strength of suppliers varies between the male and non-male markets. To take Belgium as an example, the strongest of Unilever's competitors in non-male deodorants - Henkel (Fa) and Sara Lee (Sanex) - both have greater than [10-20]*% market shares, whereas they have market shares below [0-5]*% in male deodorants. Coty (Adidas, Playboy) is the second largest competitor to Unilever in Belgian for male deodorants ([5-10]*%), but is almost absent in the non-male deodorant market (market share below [0-5]*%). Beiersdorf is the only strong competitor in both markets in Belgium.

(58) The differences between the composition and strength of the main competitors are also demonstrated in the other Member States. Generally, Colgate is stronger in the non-male market (with Palmolive and Soft&Gentle – which are rather female-oriented brands). The same applies to Henkel with Fa which is always much stronger in the non-male market than in the male market (except in the United Kingdom and Ireland where Henkel's main brand is Right Guard which is relatively strong in the male market). Coty has higher shares in male with Adidas and Playboy, its presence in the non-male deodorant market is very small. Procter&Gamble also primarily competes in the male market with Gillette and Old Spice (its female brand Secret in the United Kingdom and Ireland has a very small market share).

(59) Sara Lee (with Sanex) is consistently strong in non-male markets, taking the top-3 position for Unilever's competitors in all markets presented in Table 4. The only Member State where Sara Lee is a significant competitor in male market is Spain, where it also sells Williams (a male-only brand). Among the most significant international competitors, it is essentially only Beiersdorf who managed to develop

sizeable positions in both male and female deodorants, with Nivea and Nivea for Men. Some local brands also occasionally have a more material presence in both male and non-male markets (like L'Oreal's Narta and Narta for Men in Portugal, or L.Puig's brands in Spain).

(60) Given that the presence and strength of the main competitors and their brands differ significantly across the male and non-male deodorant markets, the competitive landscape and the constraints are thus significantly different for both markets.

(61) Unilever, in its reply to the Statement of Objections, argues that the differences in competitive landscape are overstated, arguing that most large competitors are present in both male and non-male markets. Also, it argues that some competitors are stronger in non-male markets because Axe takes up a very significant portion of the male market.

(62) These arguments cannot invalidate the conclusion that the competitive landscape and constraints are significantly different for male and non-male markets. Most of the suppliers have stronger positions in either one market or the other (the only exception from a large competitor being Beiersdorf), as determined by the composition of their brand portfolio. The large market share of Axe in the male market cannot explain alone the significantly different positions of competitors in male and non-male markets. The strong or weak presence in male or non-male market is due to the primary gender orientation and/or success of a particular producers' brand within its target group of consumers. This is best demonstrated by Sara Lee and the success of its non-male Sanex deodorants in a number of Member States, whereas Sanex for Men is not a strong brand in most of those Member States. In any event, the fact that the male market is largely dominated by Axe (without having any presence in non-male market) is itself proof that that the competitive landscape is significantly different in the two markets.

Gender differentiation [...]*

2526(63) [reference to parties’ internal documents]. ,

(64) [reference to parties’ internal documents]

27(65) [reference to parties’ internal documents]

Demand-side substitution

(66) The Commission Notice on the definition of relevant market for the purposes of 28 Community competition law (hereinafter the “Relevant Market Notice”)attributes a prominent role to the demand substitutability in the definition of the relevant product market. The market investigation provided a number of elements which allow for the

conclusion that male and non-male deodorants are not substitutable from a demand side point of view.

Retailers distinguish between male and non-male deodorants and present them separately in retail outlets

(67) A clear majority of retail customers across all Member States confirmed that they differentiate deodorants according to gender in the market investigation, with almost all confirming that they distinguish male-specific deodorants from female-specific or 29from unisex deodorants.

(68) Indeed, most retailers across Member States organise their shelf space so as to visually distinguish male from female deodorants (and unisex when relevant) on their 30 shelves.While some retailers have a separate "men's shop" with male-specific toiletries including male deodorants, others place male and female deodorants separately in different locations on the deodorant shelf, or even on different shelves. This is not disputed by Unilever.

(69) Such a gender-specific distinction of deodorants on the retailers' shelves indicates that most shoppers indeed look for a gender-specific deodorants, and then choose amongst the available deodorants for a gender-specific one. The prominent position of gender differentiation in the hierarchy of a consumers' decision of which deodorant to buy is 31also acknowledged by the Parties.

Price differences between male and female deodorants

(70) The investigation also points to significant differences in average price of male and non-male deodorants.

(71) A price comparison has been conducted on the basis of average prices of male and non-male deodorants for a number of national markets. To avoid combining prices of different formats (which indeed have different prices per litre or per kg), comparisons were made between male and non-male markets of the same format. Figure 2 presents the results for the most important formats (aerosol sprays and roll-on).

Figure 2: Comparison of historic average prices of male and non-male deodorants of a particular format (spray and roll-on) in Belgium, the Netherlands, Denmark, Spain, Portugal, the United Kingdom and 32Ireland

[...]*

(72) In almost all Member States, the average price of male deodorants is higher than non-male deodorants both in the spray and roll-on formats (with the exception of the Netherlands in the roll-on format). Such a difference in average prices is sometimes quite significant – in general the price gap is about [0-30%]*.

(73) In its reply to the Statement of Objections, Unilever argues that the pricing comparison is not very telling. It explains that average prices in the male market are higher because of the presence of male-focused brands like Axe, Gillette or Adidas which tend to have a premium price positioning, and due to the additional factor that Axe has a high price per litre price as it is sold in a smaller format. Unilever submits that brands which are present in both gender markets show no clear pattern when comparing prices of their male and non-male variants, so no real male-specific premium is noticeable with these brands. Apparent overall patterns are driven by the brand mix, rather then by any fundamental differences of competitive conditions (which would be manifested in all brands). Unilever also submits that promotions tend to take place across the entire brand and are not limited to just one gender-specific product.

(74) As Unilever points out, average prices are higher for male deodorants because of the brand mix (meaning that more premium and higher priced brands have a larger presence in the male deodorant market). This only confirms that the brand composition is indeed different for the two markets, and male consumers on average spend more on premium-priced products. It is true that a price premium for male deodorants] does not consistently exist for brands with a sizeable presence on both gender markets, but there are also price differences between average prices of male and non-male variants of such brands present in both markets. Unilever argues that where such difference in average prices occurs, it is also primarily driven by a different brand mix (for example premium variants launched for one gender). Again, this shows that the composition of variants is different for these brands (for example some female-branded products have premium-positioned variants like hair-minimizing, or variants designed for girls which do not exist in the male deodorant market.

(75) As demonstrated in Figure 2, a significant degree of co-movement of prices can be observed. However, there are also many examples where the price levels move in different directions (for example, roll-on deodorants in Belgium at the beginning of 2009, spray deodorants in Spain in late 2007 and others). While certain factors clearly influence both the prices of male deodorants and non-male deodorants in a similar way (such as production costs), the fact that prices may move in different directions shows that factors that are specific to either the male or female market also exist. Generally, the average price difference between male- and non-male deodorants is different in different Member States, but it does vary over time in the same Member State. This indicates that, in principle, a merger may create scope for increasing prices in one market without having the same effect in the other.

(76) Unilever contests the reliability of the analysis. It argues that a co-movement analysis which may be more relevant for market definition tends to show a significant co-movement of prices. Unilever criticises the fact that no statistical tests were included in the analysis as to the co-movement of prices and that only two examples are provided where prices have not moved in the same direction. Unilever argues that both of those examples are due to different brand mix and do not reflect the general difference between the competitive conditions. According to Unilever, the increase of

prices of male-deodorants in the roll-on segment in Belgium was influenced by the introduction of new premium variants (Nivea for Men's "Silver protect"), while in Spain, the increase in prices of male-deodorants in the spray segment was influenced by the launch of several premium Axe variants, coinciding with a period of a relative weakness of private label products in the male market, and a launch of premium male gift packs around Christmas.

(77) It is not excluded that the differences in price movements may coincide with launches of new premium products and the relative weakness of private label products (such as the examples of Belgium and Spain as argued by Unilever), so that the modified brand mix moves the average price of a given gender market in a different direction than the other market. However, this just shows that events which are specific to male or non-male markets and which have an influence on the average price level may occur separately for each market. The fact that average prices have moved in different directions at given times is also in line with the possibility that price increases may occur on only one market, independently from the other market.

Differences in the growth of the male and the non-male deodorant market

(78) Another feature that distinguishes male and non-male deodorants is the different market-growth. As can be seen from Table 5, the male deodorant market grew faster (or showed a lower decline) than the non-male deodorant market between 2008 and 2009 in all seven Member States where competition concerns were identified. The greatest divergences in growth rates were in Denmark (where the male market grew at [5-10%]* compared to sales of non-male-deodorants which grew at [0-5%]*), the Netherlands (male: [5-10%]*; non-male: [0-5%]*) and Portugal (male: [5-10%]*; non-male: [0-5%]*).

Table 5: Growth rates of male and non-male deodorant markets between 2008 and 2009 – Source: Form CO.

Male Non-male

Belgium [5-10%]* [5-10%]*

Denmark [5-10%]* [0-5%]*

Ireland -[10-20%]* -[10-20%]*

The Netherlands [5-10%]* [0-5%]*

Portugal [5-10%]* [0-5%]*

Spain [0-5%]* [0-5%]*

United Kingdom [0-5%]* [0-5%]*

(79) The different growth rates may also reveal differences in penetration of deodorants within male and within female consumer groups and differences in the frequency of

use of deodorants between genders, as suggested by some respondents in the market 33 34investigation [reference to parties’ internal documents].

(80) In its reply to the Statement of Objections, Unilever submits that different growth rates do not necessarily suggest separate markets, taking the example of plasma and LCD televisions where sales are growing in a different ways for each but both are considered to be part of the same market. Unilever also points to some Member States where differences in growth rates were not as significant as in others. Unilever submits that there are greater differences between the growth rates of Sanex and Dove products and that such differences did not affect the finding that those brands are close competitors.

(81) It is not argued that different growth rates are of themselves proof of separate markets, but they are rather seen as an additional element showing that the dynamics of the two markets are different. It is true that the differences in growth rates are not the same in all Member States – such differences are more significant in some then in others. The LCD/Plasma example suggested by Unilever in previous recital is misplaced, since those two products do not address distinct consumer groups, unlike male deodorants and non-male deodorants. The Dove/Sanex example given in previous recital is also misplaced. While Dove and Sanex are generally close competitors and compete for the same target group of consumers, male and non-male deodorants have different attributes as they target a distinct set of consumers. The growth of a particular brand (like Sanex) only reflects the success of that particular brand, and does not reflect a general market trend.

Cross-gender usage of deodorants

(82) Male deodorants are clearly designed for male consumers, and women do not typically use those products. As illustrated by the figures provided by Unilever, only a small fraction of male deodorant brands’ users are women in the five largest Member States (for which such data is available). According to those data, the percentage of female users of the main male-deodorant brands was generally in the range of [0-10]*%, with

37 some small exceptions.Thus, male deodorants are clearly not an option for female consumers who are likely to represent at least half of the total consumer population for deodorants. This means that male deodorants do not represent a significant direct competitive constraint to deodorants used by women, which is in line with the Parties’ argument that Unilever’s male-orientated Axe deodorant essentially does not compete with Sanex (and per analogy with other unisex or female brands).

(83) Within the non-male category, there are female deodorants, but also deodorants with a neutral, unisex proposition which in principle appeals to both sexes (the level of that appeal differs from Member State to Member State). A certain proportion of male consumers indeed prefer to use more neutral unisex deodorants to male-specific deodorants. This is particularly evident in Spain, where unisex deodorants are the most 38common. However, according to the figures provided by Unilever , up to approximately [60-70]*% of the non-male/unisex variants of Sanex in Spain are used by women. Other brands in Spain which are specifically non-male have an even 39 greater proportion of female users (roughly [70-90]*%).According to the same figures provided by Unilever, the percentages of female users is even higher (sometimes significantly) in other Member States, suggesting that even fewer men use non-male deodorants in those Members States.

(84) Based notably on those cross-gender figures, the Parties argue that no “hard lines” between male and female deodorants typically exist, and therefore all deodorants, irrespective of gender differentiation, should be considered as a single market. In its reply to the Statement of Objections, Unilever argues that up to 30% of non-male deodorant products are used by males, which Unilever argues is significant. Unilever notes that gender differentiation is not very developed in Spain, [...]*.

(85) [...]*.

(86) As regards the data on cross-gender submitted by Unilever, they have to be interpreted carefully. Those figures vary quite significantly according to the precise timeframe (for example, earlier figures in the Form CO for the United Kingdom showing that men account for [10-20]*% of Rexona for Women users and [20-30]*% of Nivea for

37 The percentage of female users of Axe was [0-5]*% in France, [0-5]*% in Germany, [0-5]*% in Spain, [0-5]*% in the United Kingdom and [10-20]*% in Italy (being an exceptional situation). For Nivea for Men, the percentage as [0-5]*% in France, [5-10]*% in Germany, [0-5]*% in Spain, [0-5]*% in the UK and [0-5]*% in Italy. For Rexona for Men, the percentage was [0-5]*% (France), [5-10]*% (Germany), [0-5]*% (Spain) and [0-5]*% (United Kingdom), data for Italy not available. See table 6.7 at page 54 of the Form CO. The subsequent submission by Unilever of 27 July 2010 provided more recent data for some additional brands in three Member States – in the United Kingdom the percentages would be Axe [0-5]*%, Nivea for Men [0-5]*%, Rightguard for Men [5-10]*%, Rexona for Men [5-10]*%, Vaseline for Men [5-10]*%; for France – Axe [0-5]*%, Nivea for Men [0-5]*%, Rexona for Men [0-5]*%, Narta for Men [5-10]*%, Sanex for Men [0-5]*%, Yves Rocher for Men [0-5]*%; for Spain – Axe [0-5]*%, Nivea for Men [0-5]*%, Rexona for Men [0-5]*%, Sanex for Men [0-5]*%.

38 [...]*

39 [...]*

21

Women users; subsequently received data show only [0-10]*% of male users for 40Rexona for Women and [0-10]*% for Nivea for Women).

(87) Furthermore, the figures show that the highest male usage ([20-30]*%) relate to Sanex in Spain, whereas other Spanish non-male deodorant brands have lower figures (in the 41range of [10-30]*%); in other Member States, those figures are generally within a 42 significantly lower range.Consequently, if one looks at the market in totality, the figures relating to cross-gender use are significantly lower. In addition, Sanex, as a relatively gender-neutral brand, has developed and launched a male variant (although it is generally not very successful). The decision of Sara Lee to develop a specific male Sanex deodorant demonstrates that the non-male /unisex Sanex deodorant (which in principle could be expected to appeal to both genders) was not enough to target the specific male deodorant market.

(88) Taking all these elements into account, the figures presented by Unilever which refer to the percentage of male users of certain non-male (unisex) brands are not sufficiently strong and robust so as to justify the existence of a separate deodorant market where male and non-male deodorants are be distinguished. In particular, cross-gender usage should be regarded also from the other side: men's deodorants are essentially used only by men and not by women (a fact not disputed by Unilever), which strongly suggests an absence of demand-side substitution for female consumers.

(89) The Parties also point to interaction indices showing a relatively high level of interaction between certain male and non-male deodorant brands in some cases (in particular, Axe is recorded as having a relatively high 'switching' with some female brands in the panel data). At the same time, the Parties explain that Axe is a very distant competitor to non-male deodorant brands such as Sanex, Rexona for women or Dove for women, a fact confirmed by the market investigation. However, the consumer panels used for the calculation of switching and interaction indices relate to purchases and not the actual use of the products. The purchaser may also buy

40 Data submitted by Unilever on 27 July 2010.

41 According to data submitted by Unilever on 27 July 2010 these figures for Spain are: Nivea for Women [20-30]*%, Rexona for Women [20-30]*%, Dove for Women [10-20]*%, Byly for Women [20-30]*%. According to earlier figures from the Form CO the figures for Spain are [20-30]*% for Nivea for Women and [20-30]*% for Rexona for Women.

42 United Kingdom - according to data submitted by Unilever on 27 July 2010: Dove for Women [5-10]*%, Nivea for Women [5-10]*%, Soft&Gentle [0-5]*%, Rightguard for Women [0-5]*%, Sanex for Women [20-30]*%, Rexona for Women 5-10%, Vaseline for Women [10-20]*%. According to earlier figures from the Form CO: Rexona for Women [10-20]*%, Nivea for Women [20-30]*%, Right Guard for Women [5-10]*%.

France: according to data submitted by Unilever on 27 July 2010: Dove for Women [5-10]*%, Nivea for Women [5-10]*%, Sanex for Women [10-20]*%, Rexona for Women [5-10]*%, Fa for Women [0-5]*%. Narta for Women [0-5]*%, Tahiti for Women [0-5]*%, Ushuaia for Women [0-5]*%, Yves Rocher for Women [0-5]*%. According to earlier figures from the Form CO: Rexona for Women [0-5]*%, Nivea for Women [0-5]*%, Narta for Women [0-5]*%.

Germany - according to earlier figures from the Form CO (no later figures provided): Rexona for Women [0-5]*%, Nivea for Women [20-30]*%, Fa for Women [10-20]*%. No figures provided for Italy.

22

43deodorants for other family members, meaning that the purchase switching patterns 44observed cannot be directly interpreted as sign of actual substitutability.

(90) The Parties also note that the distinction between unisex and female deodorants is not as clear, and in some cases it may be difficult to categorize a brand within the unisex or female segment. Indeed, in Spain for example, the investigation shows that the non-male Sanex deodorant (which is classified by the Parties as unisex based on Nielsen) competes significantly with non-male versions of Dove and Rexona (which were both classified as female). However, this fact is entirely in line with a product market definition based on a non-male market, where male deodorants are clearly separated from non-male deodorants.

(91) The Parties also imply that the gender distinction between deodorants is blurred because shoppers purchasing deodorants may be different from the end user (for example, mothers buying deodorant for their teenage children, or wives/girlfriends buying deodorants for their male partners). The Parties explain that in marketing strategies, producers may target the shoppers, by making male gift packs attractive to the shoppers who may not be male, for example. However, the fact that the purchaser or the decision-maker may not be the same person as the end-user of the product does not contradict the finding that from a demand-side, male deodorants are clearly distinct from non-male deodorants. Taking the gift packs as an example, when a wife/girlfriend buys a gift pack for her husband/boyfriend, that gift pack should be appealing to her as a purchaser, but she would clearly be buying it for the needs of the male consumer, and hence chooses between male deodorant gift-packs.

SSNIP tests of the gender segments using an economic merger simulation model

(92) In addition to the market investigation, an economic model based on Nielsen scanner data received from Unilever has been used. The precise methodology, limitations and results of this exercise are presented in the Annex, which also discusses Unilever's observations on the model. .

(93) The model's results tend also to support the qualitative findings of the investigation related to the market definition. In particular, a test based on a small but significant

43 This feature relates to consumer panels consisting of households (as the Parties acknowledge explicitly), but also to panels consisting of in individual purchasers, who still may be purchasing products for other users/family members

44 In its reply to the Statement of Objections, Unilever argues that it is not correct to explain the flaws of the data with 'false switching' recorded due to purchasing for other family members Unilever states that since the data compares purchasing behaviour at the buyer level over two consecutive years, the household effect can only trigger a significant spurious closeness between male and non-male deodorants if over the course of the two consecutive years there was significant change in the gender distribution of the households represented by individual buyers, which according to Unilever does not occur. Unilever's explanations do not seem to be correct. If buying habits within a family change, that change is also recorded as a switch. For example, when a teenage son starts to use deodorants, it is his mother who purchases those deodorants. There would also be a 'switch' recorded where the son starts buying deodorants on his own (instead of parents), or when he moves out of the home. The reason why this spurious effect mostly relates to Axe may indeed be explained by the fact that Axe is a very popular brand with teenage boys and it is a product bought by many families for their children.

45 See the Technical Annex for more details.

23

46 permanent increase in price ("SSNIP test")of the gender segments (male/non-male) has been run to see whether they can be separated in the context of anti-trust.

(94) The simulations show that the profits of a hypothetical monopolist of the male (non-male) segment would increase if the prices of all male (non-male) deodorants increased by 5%. These results support the conclusions of the market investigation. They confirm that a hypothetical monopolist in the non-male market would not be constrained by the male deodorant products, indicating that those products belong to a separate market.

Table 6: Percent change in profits due to a 5% SSNIP, separately for the different nests (segments) Segment

country Model Male Non-male

Belgium one-level 8 9

two-level 7.5 11.2

The Netherlands one-level 7.5 10.7

two-level 6.7 11.3

Spain one-level 23 26

two-level 22.2 24.6

The United Kingdom one-level 3.4 4.2

two-level 3.6 4.5

Note: Based on the estimates of Table 5 and Table 6 of the Technical Annex.

Supply-side substitution

(95) According to the Relevant Market Notice, supply-side substitution may be taken into account if suppliers are able “to switch production to the relevant products and market them in the short term.” Thus, supply-side substitution involves entry from existing adjacent activities at a low cost and without incurring irreversible investments and implies the ability to substitute production of one good for another at short notice in 47 response to relative price variationsOnly if these conditions are met will supply-side

46 A SSNIP test answers the question whether the parties' customers would switch to readily available substitutes in response to a hypothetical small (in the range 5 % to 10 %) but permanent relative price increase in the products being considered. If substitution were enough to make the price increase unprofitable because of the resulting loss of sales, additional substitutes are included in the relevant market. This would be done until the set of products and geographical areas is such that small, permanent increases in relative prices would be profitable (see paragraph 17 of the Relevant Market Notice).

47 According to the Relevant Market Notice (paragraph 24), potential competition is not taken into account when defining markets, since the conditions under which potential competition will actually represent an effective competitive constraint depend on the analysis of specific factors and circumstances related to the conditions of entry.

24

substitutability have an impact in terms of effectiveness and immediacy equivalent to the demand substitution effect.

(96) Supply-side substitution translates into market aggregation, that is, a broadening of market boundaries to include a larger group of products or geographical areas. However, the Relevant Market Notice also points out that aggregation makes sense only when production substitution among a group of products is found to be technologically feasible and economically viable for most, if not all, firms selling one or more of those products.

(97) The Parties submit that from a supply-side point of view, all deodorants belong to the same market. They argue that production technology and the know-how in order to produce male, female or unisex deodorants are very similar. They explain that male/female or unisex deodorants can be produced on the same production line and that all manufacturers produce or could produce the various gender deodorants. Only packaging and marketing differ slightly.

(98) According to the Parties, once produced, no significant cost or time delays arise in bringing different gender variants to market. They support their reasoning by noting that most male deodorant products started as female- or unisex deodorant products, and male variants were subsequently developed (with Adidas and Axe being the exceptions).

(99) In its Form CO, Unilever also detailed the time and costs needed to launch its last male variant – Dove for Men. The information provided shows that the launch of Dove for Men took [...]* between the initial project ([…]*) and the first launch […]* and had a total advertising and promotion (A&P) budget of EUR[…]* for deodorants 48 and skin cleansing products in the first year for a European-wide rollout..Unilever explains that with this launch, it was targeting a long term market share of […]* in male deodorant markets.

(100) The Parties argue that the launch of Dove for Men is not a representative case for the ease of switching supply between deodorants targeted for men and those for women mainly because [...]*.

(101) The Parties explain that ease of supply-side switching between different gender variants of deodorants can be evidenced by the success of deodorant products such as Nivea for Men, Rexona for Men and Sanex for Men. [...]*. Rexona for Men finally reached [0-5]*% share value in France and [0-5]*% in Germany after one year. [...]*

(102) The results of the investigation do not support the Parties’ view that male and non-male deodorants are supply-side substitutes for the purposes of relevant product market definition.

(103) The investigation reveals that an extension of a well known male/female/unisex deodorant brand into a different gender category would in principle be “possible” for the major deodorant suppliers, however, such an extension would still require significant time to prepare and launch the product, even if the existing production lines could be used to produce the products.

48 [...]*

25

(104) In the market investigation, the Parties' competitors suggest that the average time to launch a new product/variant would be approximately two years (between one year to 49 three years depending on the competitor and on the complexity of the project).Such ranges are in line with Unilever's submissions ([...]*).

(105) Such long time frames to launch the products demonstrates that the requirements for supply-side substitution as laid down in the Relevant Market Notice are not met. It is not the case that most, if not all, suppliers would be able to switch production to the relevant products and market them in the short term. Time frames of one to three years to extend an existing brand into a different gender cannot be regarded as sufficiently immediate for the purpose of supply-side substitution. This would equally hold true even if the time frame were only that of [...]*. This is all the more applicable in the context of fast-moving consumer goods such as deodorants, where there are significant costs and lead times (such as product development, testing, marketing strategy, advertising) before the products can actually be sold. Moreover, the significant investments and risks associated with such extensions render supply-side substitution unlikely.

(106) Finally, even if a number of unisex/female deodorants have their male variants, the success and competitive position of those male variants were developed gradually over time and are a consequence of significant development and commercial efforts, and of important investments. Therefore, a possible increase of production and sales of these male or female /unisex deodorants has to be treated in the competitive assessment as part of expansion/entry analysis, rather than supply-substitution.

(107) In its reply to the Statement of Objections, Unilever argues that most brands are already present in both markets. In Unilever's view, those brands such as Rexona or Nivea with an existing presence in both markets (making a third or more of male and half or more of non-male market) can switch production capacity or outsource to boost supplies in a different gender market with immediacy and effectiveness equivalent to demand substitution effect. Unilever submits that there are no barriers to switching supplies, and advertising/marketing/promotional budgets can be re-balanced between gender variants. Unilever gives the example of Beiersdorf's Nivea – the brand started to focus its advertising investments on male deodorants in 2009 and managed to increase its share of the male market, after it had invested more in the non-male market in the previous years. Unilever also points to L'Oreal first launching Garnier (a non-male deodorant brand), and subsequently launching L’Oreal Men expert brand, which is targeting the male market. .

(108) The Relevant Market Notice laid down strict conditions for supply-side substitution, which emphasize that suppliers should be able to switch production and market the products in the short term without incurring significant additional costs or risks in 50 response to small and permanent changes in relative prices.The Relevant Market Notice further specifies that different qualities of a product will be grouped into one product market, provided that most of the suppliers are able to offer and sell the

49 Questionnaire to competitors on deodorants sent on 25 June 2010, question 20.

50 Paragraph 20 of the Relevant Market Notice.

26

51 various qualities immediately and without the significant increases in costs.When supply-side substitutability would entail the need to adjust significantly existing tangible and intangible assets, or would entail additional investments, strategic 52decisions or time delays, it will not be considered at the stage of market definition. Contrary to the arguments pout forward by Unilever, most deodorant suppliers are not able to offer and sell both male and non-male deodorants immediately without incurring significant additional costs or risks.

(109) First, as noted by Unilever, brands that are present in both gender markets constitute a relatively small part of the male market – up to two thirds of the male market is made up of specific male deodorant brands with no material presence in the non-male market. Unilever's argument that expansion into a different gender market is a matter of rapidly switching capacities is therefore not relevant to those brands, as they would have to first undergo a timely and costly process of developing and launching a new gender variant.

(110) Second, even for brands present in both gender markets, an expansion of those brands in a particular segment in case of a small but significant price increase is not the result of a mere increase in production (as with commodities such as paper). Deodorants are branded goods whereby expansion builds upon important strategic decisions, investments and adjustment of marketing efforts, and occurs gradually rather than immediately (unlike winning an account from an important paper customer). Significant risks are also involved, as the sunk costs incurred may not always bring higher returns if the product is not well perceived by consumers. A deodorant brand has to convince consumers that it has the right product to fit the consumers' needs. In order to expand, suppliers invest in developing and marketing new innovations, product re-launches, and support them with intense marketing, advertisement and promotional activities. The Parties' presentation of supply-side substitution as easy and immediate is overly simplistic.

(111) This is also demonstrated by the example that the Parties offer – Nivea "switching" marketing expenditures from non-male to male deodorants between 2008 and 2009. The example is not just a mere (and immediate) switching of marketing expenses, but it coincides with an adjustment of the Nivea's male product range with the introduction of more premium equivalents and more importantly, with the introduction of a new premium product with an innovation feature – Nivea Silver Protect (as the Parties themselves explain). Such new strategic launches take a significant amount of time to prepare. The particular example of Nivea involves an increase of advertising spending/expenditure by about [...]* in Western Europe for male deodorants (with a corresponding decrease in advertising expenditure for non-male deodorants). The Parties argue that as a result of that "switch", Nivea for Men increased its market share in almost all markets. The annual increase ranged from as little as [0-5] percentage points in Portugal, [0-5] in the United Kingdom, [0-5] in the Netherlands and [0-5] in Belgium. The Parties omit to mention that in Spain, Beiersdorf's market share declined

51 Paragraph 22 of the Relevant Market Notice.

52 Paragraph 23 of the Relevant Market Notice. The Notice specifies that examples where supply-side substitution did not induce the Commission to enlarge the market definition are offered in the area of consumer products where suppliers are in principle able to produce the products, but there are other costs and lead times involved, such as those needed for advertising, product testing and distribution.

27

by [0-5] percentage point. The Nivea example thus rather proves that a shift in focus between the male and female markets is indeed not immediate but is rather a major strategic choice, involving significant investments and risks.

(112) Another example of a switch in market focus offered by the Parties is L'Oreal, which first introduced a female-oriented Garnier Mineral product, and is subsequently launching the Men Expert brand. Similarly, the example of L'Oreal has little to do with supply-side substitution, but is rather an example of a strategic decision involving an attempt to enter into the male and female markets separately. The Parties' argument that 'switching' marketing/advertisement and promotional budgets is proof of supply-side substitution is misconceived, as such 'switching' is simply a strategic decision to focus expansion or entry into a particular area (similar to switching advertising focus between Sanex bath & shower range and Sanex deodorant range).

(113) For the reasons specified in recitals (108)-(112) and in line with the Relevant Market Notice, such switching of a suppliers' focus between male and non-male deodorants – even if a particular brand has both male and female variants – will not be considered in the market definition.

Conclusion on gender differentiation

(114) It is thus concluded for the purposes of this case that male deodorants and non-male deodorants are separate relevant product markets.

IV.2.1.2. Deodorants by format

(115) Broadly speaking, deodorants tend to be supplied in contact forms (primarily roll-ons and sticks and occasionally wipes and creams) and non-contact forms (primarily aerosol sprays and occasionally pump sprays).

(116) The Parties submit that the appropriate product market definition for deodorants is a single (albeit differentiated) market for all types of deodorants. Considering a potential delineation of the market by format, the Parties recognize that demand-side substitutability arguments may not be sufficiently clear to conclude that all format types form part of a single product market. However, they submit that supply-side factors strongly suggest that that format is not a credible basis upon which to segment the deodorants product market.

Demand-side substitution

(117) The Parties explain that end-consumers seem to have a preference for one format over another, in particular between non-contact (aerosol sprays, pump sprays) and contact (roll-ons, sticks, creams and wipes) deodorants. They put forward that a proportion of consumers are so-called "repertoire purchasers", buying contact deodorants primarily for sports or for travelling (airline travel restrictions, for example, require use of a smaller format) and aerosol sprays primarily for use at home, but with some inter-changeability between uses.

(118) They admit that the evidence as to demand-side substitutability like switching behaviour or price difference may not be sufficiently clear to conclude that all format types form part of a single product market.

28

(119) The market investigation revealed that consumers tend to have purchasing preferences for one format over another. For many consumers, different formats may not be substitutable. The preference for one format over another can depend on a number of factors. Such factors could be health issues or consumers' skin types, perception of performance of a particular format, perception of environmentally-friendliness (sprays were considered to harm the ozone layer), the age of the user (sprays for younger people, roll-on for older), the price of a given format and preferences may also vary according to the gender of the user.

(120) A wide majority of the competitors nevertheless explain that end-customers perceive sprays, roll-ons and sticks deodorants as substitutes since all formats fulfil the same 53need. As a general pattern does not seem to exist with regard to the choice of the deodorant format, all the manufacturers offer their products in a range of format types (at least both contact and non-contact) in order to meet all customers' preferences and 54give them the possibility to choose the format that better suits their needs . For example, the recently launched Dove for Men + Care range includes aerosols, roll-ons and sticks within its deodorants product range.

(121) Although some brands may be stronger in one format, it is not necessarily due to difficulties in entering a different format. For instance, certain consumers associate the roll-on format with skin health, and the aerosol format is often associated with efficacy or fragrance. Depending on the segment of customers that a brand tries to target, a particular type of format may be chosen.

(122) Elements therefore exist which show that different formats are not substitutable for many end-consumers, however the mixed evidence does not fully support the conclusion that there is no demand-side substitutability between various deodorants formats.

Supply-side substitution

(123) From a supply-side point of view, the Parties explain that the production processes for deodorants of different formats are quite different and therefore production lines for one deodorant format cannot be used for another type of deodorant format.

(124) Aerosol deodorants are manufactured on a filling line. Aerosol cans are placed on the filling line and filled with the ingredients (including perfume, alcohol and antiperspirant). The cans are filled with liquefied gas and then undergo test baths (that is, they are transported through hot water to raise internal pressure and test for leaks) and are capped. The final product is wrapped, placed in trays, labelled/bar coded and palletised.

(125) Roll-on deodorants (and other contact formats), on the other hand, do not require gas or a test bath. Bottles/packs are placed on the production line and are filled with the product (which is mixed off-line). For example, in the case of roll-on deodorants, a roller ball is inserted in the packaging which is capped and labelled. The product is wrapped and placed in trays (as for aerosols).

53 See question 15 of questionnaire to competitors on deodorants sent on 24 April 2010.

54 See Annex 6.11 to the Form CO

29

(126) However, the Parties explain that a manufacturer can nonetheless switch production volumes between different formats easily and quickly by purchasing additional capacity from third party co-packers which manufacture all types of formats on behalf of manufacturers.

(127) According to the Parties, co-packers generally supply standardised packaging (aerosol cans, roll-ons) which can be adapted to the suppliers’ requirements by adjusting the colour and labelling for the brand. On that basis, a co-packer can supply a new format within [...]* months with an investment of under [...]* (for small volumes of less than five million units).

(128) [...]*. Once branded products suppliers like Unilever and Sara Lee set up the production capabilities with co-packers meeting their specific packaging requirements, production can subsequently be re-activated with even shorter lead times. Unilever thus explains that as a result, a manufacturer of one format of deodorant can readily switch to production of deodorants in alternative formats. In addition, there would be no significant costs or time delays in bringing different format variants to market.

(129) Indeed, although different formats need separate production lines and use different production processes, a majority of respondents indicated that they rely on third party manufacturers for the production of a new format, mainly for cost efficiencies and risk 55 management reasons.As one competitor notes, "outsourcing of parts of the production is common in the deodorant industry. However, essentially all major 56suppliers of deodorants (other than private label) have their own production."

(130) The market investigation thus confirms that the important presence of a relatively large number of third party co-manufacturers allows brand owners to easily avoid any issue linked to the production specific to each type of deodorant format, if they lacked spare capacities in-house.

(131) The investigation also confirmed that the time and costs necessary to deliver a new format of a brand, or to adapt to a demand increase of a certain format, are not obstacles for any of the brand owners willing to place such new formats on the market. The time-frames provided by the Parties for the supply of a new format by co-packers were broadly confirmed, and an existing manufacturing relationship between a brand owner and third-party co-packer (which is common in the industry) would make it even easier, several competitors mentioning that they would need one to three months 57for doing so with an existing co-packer.

(132) Production is thus not a major issue when switching into a different format. Moreover (and contrary to gender distinctions), such a switch would not entail any major costs or delays in terms of additional branding/marketing development, advertisement and promotional activities.

55 Question 8 of Questionnaire to competitors on deodorants sent on 24 April 2010.

56 Reply of Henkel to question 8 of Questionnaire to competitors on deodorants sent on 24 April 2010.

57 Question 8 of Questionnaire to competitors on deodorants sent on 24 April 2010.

30

(133) Indeed, a wide majority of competitors pointed out that they do not adjust brand/marketing propositions to formats, and that this is the same regardless of the 58 format.Thus, a new format of a particular existing brand can be placed on the market without any specific advertising/marketing efforts, as the particular brand proposition (including the specific variants of a brand, formulations, fragrances and added functionalities) already exists. An example of a particular range of deodorant products with the same fragrances, packaging design, appeal and marketing is shown in Figure 3.

Figure 3: Picture of a particular deodorant range of Rexona Men with multiple formats available (stick, spray and roll-on)

(134) Most major brands offer multiple formats (or the most prevalent formats in a given market). As Henkel notes, "Typically a manufacturer will offer all types of formats within each brand. Furthermore, the overall performance does not differ significantly 59 between the different formats."In addition, suppliers usually place different formats on the market simultaneously when launching a new brand range/variant.

(135) The marketing/branding and advertisement efforts are the main difference to gender distinction in the context of supply-side substitution. In order to be successful on the market, a new gender variant usually involves building a new and often quite distinct brand proposition targeting a different group of consumers. As explained in recitals (95) - (114)), the costs, risks, strategic decisions and time involved are significant due to the importance of marketing/branding, advertisement and promotional activities. It was also demonstrated that the majority of latest new launches of brands or brand variants are developed specifically for male or non-male markets (see recital (47)). This is different to formats, where new brands or brand variants usually cover a range of formats.

58 Question 18 of Questionnaire to competitors on deodorants sent on 24 April 2010.

59 Reply of Henkel to question 18 of Questionnaire to competitors on deodorants sent on 24 April 2010.

31

(136) Indeed, putting more deodorants in a particular format on the market (or adding another format) can be executed quickly enough, without incurring significant investments in marketing/advertisement or involving significant risk. Therefore, supply-side substitutability supports the conclusion that the markets should not be separated according to format.

Conclusion on different formats

(137) Taking all the elements into account, on balance, it is concluded that all deodorants belong to the same relevant product market regardless of format. Format will however be taken into account in the competitive assessment, in particular in the context of analysing the closeness of competition.

IV.2.1.3. Other Distinctions

(138) Deodorants can have other differentiating attributes although they are not as prominent as gender and format. Based on the investigation, neither of the further distinctions justifies a further split of the relevant product markets. Nevertheless, those further differentiating features are taken into account in the competitive assessment.

(139) As regards different variants or added benefits (such as fragrance orientation, skin friendly attributes, "no white marks"), they address different consumer preferences, but are always marketed under the umbrella of a male deodorant variant or a female deodorant variant, and represent the next level of differentiation after gender. One product often has more attributes (such as both an efficacy and a fragrance proposition) and the borders between these attributes can be fluid, so it is not always possible to draw clear lines between the variants. Competitors explain that these differentiating factors are considered to be the main benefits for consumers, but the vast majority of competitors give greater weight to segmenting deodorants by gender and format, including for the purposes of their internal reporting and market 60analysis.

(140) Such secondary functionality-oriented differentiation also applies to a possible distinction between deodorants and anti-perspirants (the latter reducing sweat in addition to regulating odour). Generally, antiperspirant capabilities are seen only as 61 one of many added benefits of deodorants.The Parties' agree with this position and explain that such a distinction plays a limited role in consumers’ decision-making process, which is also reflected by the fact that both deodorants and antiperspirants are invariably marketed in retail outlets under a common “deodorants” banner, and many 62 products combine both deodorant and antiperspirant qualities.The results of the investigation do not support a segmentation of the market between deodorants and anti-perspirants.

60 See replies to questions 13 and 30 of Questionnaire to competitors deodorants sent on 24 April 2010.

61 See for example the reply of Henkel to question 30 of the Questionnaire to competitors on deodorants sent on 24 April 2010, where Henkel refers to possible distinctions according to functionality, stating that those distinctions "could be fragrance, anti-transpiration, care or no residue, which are usually perceived as the main benefits of the product in question."

62 Form CO, paragraph 6.142.

32

(141) Private label deodorants are considered to be part of the same market as branded products in this Decision, as they compete with brands for the end-consumer.

IV.2.1.4. Conclusion on Relevant Product Market

(142) It is therefore concluded that the relevant product markets are the market for male deodorants and a market for non-male deodorants.

IV.2.2. Relevant Geographic Market

(143) The Parties submit that the relevant geographic market for deodorants is national.

63(144) In previous Decisions, the geographic markets for personal care products, including deodorants, have been analysed on a national basis. It has been noted that, even though many companies had started using brands on a European basis and there were international buying organisations, market shares and consumers’ preferences diverged among Member States, and there were significant retail price differences across Member States.

(145) The Parties agree with this analysis. Notably, the Parties argue that procurement and price negotiation takes place predominantly on a national basis, even for those customers that operate on a wider-than-national basis.

(146) The market investigation confirmed that the geographic market for deodorants remains national in scope. Customers and competitors across all Member States explain that prices and consumer preferences for brands, formats and gender varieties differ between Member States. Local brands still play an important role in several Member States. Moreover, almost all market participants confirmed that price negotiations and procurement are conducted at national level.

(147) For the purpose of this Decision, the deodorant markets are analysed on a national level.

IV.2.3. Compatibility with the Internal Market

(148) The market investigation in this case confirmed that the proposed transaction would significantly impede effective competition in a number of deodorant markets. The competitive assessment for deodorants is presented in two main parts (see Section IV.2.3.1 and Section IV.2.3.2. In the first part (IV.2.3.1), a general assessment is conducted, whereas a detailed and individual assessment on a market-by-market basis is provided in the second part (IV.2.3.2).

IV.2.3.1. General assessment

(149) The general arguments put forward by the notifying party in the notification and its responses to the Article 6.1(c) Decision as well as in its reply to the Statement of Objections are discussed in this section, followed by an analysis of certain elements common to all the national markets, including a description of the merger simulation

63 Commission Decision of 30 July 2003 in Case No. COMP/M.3149 - Procter & Gamble/Wella; Commission Decision of 15 July 2005 in Case No. M.3732 - Procter & Gamble/Gillette and Commission Decision of 30 July 2008 in Case No. M.5230 Capman/Litorina/Cederroth.

33

that has been used as well as general issues of potential countervailing buyer power and entry.

IV.2.3.1.1. Main arguments of the notifying party

(150) The notifying party submits that the transaction would not impede effective competition in the affected markets for deodorants. Unilever puts forward several core arguments at a general level and supports them with examples from certain Member 64States.

(151) First, Unilever submits that deodorant markets are characterised by a large number of brands and variants serving particular needs and tastes in terms of prices, packaging, format, general and secondary attributes. In such a differentiated product market, the Parties' market shares are not a true reflection of the competitive interaction between them and, according to the notifying party, overstate the effects on competition resulting from the transaction. Market shares in a differentiated product market are, according to the notifying party, mainly "providing a filter" by which markets 65requiring in-depth analysis can be identified .

(152) Second, the notifying party submits that Sara Lee’s brand, Sanex, with its female bias and its healthy skin proposition, does not compete closely with the brands representing a significant proportion of Unilever’s share of supply. This would most notably be the male-orientated brand Axe, which accounts for [30-50%]* of Unilever’s deodorant sales in all national markets concerned.

(153) Meaningful competitive interaction between Sanex and Unilever’s brands would hence be limited to Dove and Rexona. The notifying party submits however, that even for those brands, the differences are stronger than the similarities. Sanex shares a general skin caring proposition with Dove although Dove’s proposes moisturising to achieve beautiful skin whereas Sanex has a more healthy skin proposition aiming at avoiding skin irritation through natural ingredients. Also, Dove is on average more expensive than Sanex, which has a mid-price positioning. Sanex shares similar price positioning and efficacy attributes with Rexona, but Rexona does not exhibit the same healthy skin proposition offered by Sanex .

(154) Third, the notifying party puts forward that the deodorant market in the Union is highly competitive with a large number of multi-national and local suppliers across the spectrum, including in the skin care and efficacy segment where Sanex competes. Those suppliers own a large number of brands which will continue to spur the competitive rivalry in the affected markets post transaction.

(155) Fourth, the notifying party contends that retail customers exert strong countervailing buyer power on personal care manufacturers, which tend to be dependent on a handful of retail customers for the majority of their deodorant sales. In the affected markets, the notifying party submits that the retail segment is significantly concentrated and such a market structure enables retailers to exercise a strong competitive constraint on

64 Form CO, paragraph 6.40.

65 Form CO, paragraph 6.159.

34

suppliers in three different ways: as customers with the possibility to delist, as "gatekeeper" controlling the access to shelf-space, promotional slots and setting retail 66prices as well as competitor with their own private label brands.

(156) Finally, according to the notifying party, suppliers in adjacent product and geographic markets are well-positioned to expand into deodorants, exerting an additional 67competitive constraint. The notifying party puts forward that there are no significant barriers to entry and expansion in the deodorant markets. Whilst branding is normally considered to represent a barrier to entry/expansion in markets of this nature, the ability of suppliers active in adjacent markets to stretch their brands into neighbouring markets represents an often used starting point to successful market entry or expansion in the deodorant sector.

(157) According to the notifying party, there are several examples of recent brand extensions into deodorants, including Sanex's launch of Monsavon deodorants in France (from skin care), Procter & Gamble's entry in deodorants under the Gillette brands in United Kingdom, Italy and France (from shaving products) and Puig's extension of the Actovit/Lactovit range from skin care to deodorants (in Spain). Regarding extension of a deodorant line into adjacent geographic markets, the notifying party cites L'Oreal's Garnier Mineral from the Czech Republic, Poland, Hungary and Romania to 13 other European countries, Henkel's launch of the Right Guard brand in Germany in 2010 and Coty's entry into Spain with the Playboy brand in 2008.

IV.2.3.1.2. Likelihood of price increases

Elements to consider in the differentiated deodorant market

(158) In differentiated markets such as the market for deodorants, market shares, although providing an indication of the market power of the Parties may not fully reflect the competitive interaction. Indeed market shares may over or underestimate the effects of a transaction depending inter alia on the closeness of substitution between the merged products.

(159) The market investigation therefore included detailed questions (in questionnaires and follow-up interviews) to the five biggest retailers within each relevant national market as well as the six main multinational competitors. It also addressed questions to competitors with a more limited presence in national markets. Moreover, a large amount of internal documents of the Parties were reviewed in detail in relation to deodorant markets.

(160) This Decision firstly examines the elements pointing towards price rises before discussing countervailing factors like entry or potential buyer power of retailers in line with the framework for analysing non-coordinated effects in the Guidelines on the assessment of horizontal mergers under the Council regulation on the control of concentrations between undertakings (hereinafter referred to as “the Horizontal

66 Form CO, paragraph 6.99.

67 Form CO, paragraphs 6.79 to 6.97.

35

68Merger Guidelines”), before balancing all factors and reaching a conclusion on the impact of the transaction.

The merging firms are close competitors with regard to certain brands

(161) In assessing mergers in differentiated markets, the degree of substitutability between the merging firms’ products is an essential element in assessing the effects of the merger. The higher the degree of substitutability between the products of the Parties, the more likely it is that the merging firms will be able to raise prices significantly after the merger. To correctly assess the ability of the notifying party to raise prices, the analysis has to be considered at the level of brands, since the interaction between different brands of Unilever and Sara Lee respectively varies according to the brands.

(162) The notifying party argues that the Parties are not close competitors, underlining that Sara Lee's brands do not compete with the main brand of Unilever, namely Axe. A more detailed and refined analysis of the closeness of competition between Unilever and Sara Lee's brands is provided in the competitive assessment tailored to each Member State in Section IV.2.3.2. Two aspects, however, which are common to all national markets concerned, are addressed immediately as they were raised as a general argument by Unilever.

(163) First, it is not claimed that all of the Parties’ brands are close competitors with the core brand of Sara Lee, namely Sanex. Axe is clearly focused on the male market. In the non-male market, the closeness of competition analysis should consider the Unilever brands Dove and Rexona in particular, and Vaseline in certain Member States.

(164) Second, it is not argued that Unilever's brands are the only brands that have a comparable brand positioning to Sanex. Notably, Beiersdorf's brand, Nivea, also appears to be positioned close to Sanex. The fact that Sanex, Dove, Rexona and Nivea are relatively close competitors is in line with the experience of customers and competitors. 69 Respondents in the market investigation indicatedthat, in general, the Sanex brand is positioned in the medium to high segment of the product spectrum and competes directly with Unilever's brands Dove and Rexona and Beiersdorf's Nivea.

70 (165) [reference to parties’ internal documents][reference to parties’ internal 71documents]. For deodorants with specific attributes like skin care as well as the non-white mark sub-segment of the female deodorant segment, Sanex and Nivea are the only alternatives to Unilever's Dove and Rexona label.

68 OJ C31/5, 5.2.2004 .

69 Questionnaire to customers on deodorants sent on 23 April 2010, questions 27 and 30. Questionnaire to competitors on deodorants sent on 24 April 2010, question 56.

70 [...]*.

71 [...]*.

36

72 (166) The analysis of a number of interaction studies submitted with the Form COshows a high interaction between Unilever brands Vaseline (in Member States where that brand is available) Dove, Rexona and Sara Lee's brand, Sanex.

Sara Lee contributed to important innovations in the market

(167) The competition within the deodorant markets also has an innovation dimension. The investigation showed that Sara Lee was, via Sanex, contributing to the major innovations on the market, notably with its new 'natural' deodorants composition and a upside down roll-on deodorant.

(168) In the market investigation, a number of respondents in several Member States mentioned that Sanex has been able to grow in certain markets, in particular with its "Nature Protect" deodorants. The “Nature Protect” deodorant range launched by Sanex 73in 2008 is a deodorant with minerals (alum stones - Pierre d’Alum) as active ingredients instead of antiperspirants. The natural origin of the product gained success as many consumers were concerned about negative influence of certain ingredients such as aluminum which was said to be causing breast cancer.

(169) Another significant innovation stemming from Sara Lee is a deodorant with an upside-down roll-on device. Upside-down roll-on deodorant is characterised by a dispenser of viscous liquid comprising a container, a fitting connected to the container and a roller. This device enables the user to soften their underarms while keeping the flow of the product smooth and free-running. It avoids unwanted leakage and dryies of the surface of the roller when the dispenser is not in use. Therefore the dispenser is immediately 74ready for use whenever the consumer wants it .

(170) Sara Lee registered a patent in 2004 for its upside down roll-on deodorants and was the first supplier to launch that product on the market and was subsequently able to gain 75 market share in the non-male market with that innovation.Unilever introduced a similar product in 2008 and L'Oreal introduced one in 2009.

(171) Unilever does not deny that Sanex has contributed to innovations, but contested that Sara Lee "is uniquely innovative" and that the proposed transaction would change the current 76 dynamics of innovation in the market.According to Unilever, such a change would not occur because the market is characterised by a significant degree of creativity and Sanex 77innovations are not standout innovations, - they have all been replicated by other suppliers within a short period of time. As Unilever explains, the launch of Sanex

72 See Figures 6.7, 6.14 and 6.22 Form CO.

73 Alum is a mineral from volcanic rock that is found as a crystal in its natural state.

74 See Sara Lee International patent filling dated 10 May 2004, available in Appendix 2 of the L'Oreal Memorandum dated 04.05.2010.

75 The analysis of the non-male market in Belgium provides more details – see Section IV.2.3.2.1.1.

76 See the Reply to the Statement of Objections, paragraphs 5.14 and 5.15.

77 Parties' Reply the 6(1)(c) Decision, 14/06/2010, paragraphs 2-46 and 2-47 and annex 16.

37

NaturaProtect was followed quickly by the launch of other products by competitors, notably L'Oreal's Garnier Mineral, and Unilever's Rexona Mineral which, according to Unilever, have achieved higher market shares than Sanex NaturProtect in several 78Member States.

(172) Innovations such as upside-down roll-on deodorants and minerals as active ingredients are significant, as competitors quickly launch comparable products on the market, thus revealing that the innovation was important.

(173) The two Sara Lee innovations – upside-down roll-on and minerals as active ingredients - were indeed recognised as representing significant innovations on the market. At the same time, however, it is acknowledged that other companies like Unilever or Beiersdorf also introduce new concepts on the market like the “no white marks” deodorant (Unilever) or hair-minimizing deodorant which was first introduced on the market by 79 Beiersdorf.Customers and competitors both confirmed that innovation is an important aspect in the deodorant market in general and a necessary attribute for 80 success.Broadly speaking, the above-discussed innovations were also identified by 81third-party market reports provided by the Parties.

(174) It can therefore be concluded that Sara Lee has contributed to overall innovation on the market. While it is not alone, Sara Lee’s contributions are an important feature distinguishing it from other deodorant suppliers.

The merger simulation shows price increases

(175) The Statement of Objections used an economic model to predict the likely impact of 82 the Unilever/Sara Lee transaction on the deodorant markets.The model has two components. The demand side describes how consumers chose a deodorant product. The supply side describes how producers chose their prices. The two sides are interlinked and define a static market equilibrium.

(176) For the demand side, the so called one-level and two-level nested logit models were used. The one-level model postulates that consumers view the products with male proposition as more similar to each other than to non-male deodorant products, and non-male deodorant products are perceived to be more similar to each other than to male-deodorants. The two-level model is a refinement of the one-level structure. In the latter model, consumers perceive products with the skin friendly proposition as closer substitutes than others within the male or non-male segments.

(177) The strength of demand substitution within and across these segments and sub-segments can be estimated using standard econometric techniques. For this purpose, detailed product level scanner data by Nielsen were used for four Member States- Belgium, the Netherlands, Spain, and the United Kingdom. For each Member State, separate estimations were made, both with the one-level and two-level nested logit models.

(178) On the supply side, the model assumes that producers compete by setting their products' prices while viewing demand as described by the estimated model. This defines a standard, static Bertrand-Nash market equilibrium.

(179) The model can be used to simulate post-merger prices by assuming that after the merger, the merging brands are priced by the same firm, while they were competing with each other pre-merger. The elimination of competition between them creates an incentive to increase prices. The predicted price increase is obtained by comparing the model's post-merger market equilibrium to the prevailing pre-merger equilibrium.

(180) The model takes into account the distinction between male and non-male deodorants by placing them in different nests. However, the simulation takes account of any potential switching between male and non-male deodorant brands and produces a price prediction for both markets.

(181) The model's predicted price increase should be interpreted as a permanent shift in the price level. As a result of normal competitive interaction in the market, the observed prices can fluctuate around that price level. Even in the absence of mergers, the data might show significant price changes from period to period. These "regular" changes might continue to occur after the merger but around an elevated mean price level. The model's predictions attempt to quantify how much higher this mean price level would be as a result of the merger.

(182) According to the main results reported in the Statement of Objections, simulations 83 predict for the non-male segment price increases around 6% in Belgium,5%-6% in ,84 85 8687 the Netherlands,2% in Spain,and about 4% in the United Kingdom.The main source of these average price increases arise from the incentives to increase Sanex prices, which are predicted to go up substantially (over 10%). In Spain, however, the Sanex price increase is expected to be smaller (5%). Unilever's brands show a more moderate but still significant price increase. The results were also subjected to several 88robustness checks in the Statement of Objections.

83 With a 90% confidence interval of [2.6, 13.5]*%.

84 With a 90% confidence interval of [1.1, 9.8]*%.

85 With a 90% confidence interval of [0.2, 5.9]*%.

86 With a 90% confidence interval of [2.5, 6.0]*%.

87 The overall (male and non-male combined) price increases were around [5-10]*% [1.9, 9.8], [0-5]*% [1.7, 6.1], [0-5]*% [0.4, 6.2], and [0-5]*% [1.4, 3.8] for Belgium, the Netherlands, Spain, and the United Kingdom, respectively.

88 See Technical Annex for more details.

39

(183) The Statement of Objections emphasised that the model had several limitations due to 89 the necessary restrictions in its assumed structure.As explained further in the Technical Annex, such restrictions are standard in the economic literature of merger simulations. However, they constrain the results' possible scope of interpretation. The assessment of the results must be put in context with the other qualitative and quantitative analysis carried out during the investigation.

(184) Nevertheless, according to the Statement of Objections' conclusion, the modelling exercise does significantly increase the overall reliability of the assessment. The models' predictions are based on the processing of, and extrapolation from, tens of thousands of case relevant observations, which themselves represent aggregations of millions of consumer transactions. The value of the econometric estimations and merger simulations lies in the attempt to take on board the information content of this massive set of consumer data.

90 (185) Unilever criticises the model on several counts.First, it emphasises that the restrictions the nested logit model puts on the data make it unsuitable for evaluating the transaction's likely effects. Second, it argues that the chosen estimation methodology (instrumental variables estimation) is not appropriate as the available instrumental variables are weak. Third, the model's limitation in describing the vertical relationship between retailers and producers reduces its capability of predicting the price effects of the merger.

90 (185) Unilever criticises the model on several counts.First, it emphasises that the restrictions the nested logit model puts on the data make it unsuitable for evaluating the transaction's likely effects. Second, it argues that the chosen estimation methodology (instrumental variables estimation) is not appropriate as the available instrumental variables are weak. Third, the model's limitation in describing the vertical relationship between retailers and producers reduces its capability of predicting the price effects of the merger.

(186) The assessment of Unilever's critique has involved carrying out further robustness checks, as well as reemphasising the Statement of Objections' stand on the proper, cautious interpretation of the results. The robustness checks have included modelling refinements (along the lines suggested by Unilever) and re-estimations, implementation of weak instrument robust parameter tests, and cost calibration exercises. The results have shown that (i) the estimation/identification methodology, within the framework of the models used, is reliable; and (ii) if the results change due to the modelling refinements the direction of change is upwards: the predicted price increases are higher than those of the Statement of Objections.

(186) The assessment of Unilever's critique has involved carrying out further robustness checks, as well as reemphasising the Statement of Objections' stand on the proper, cautious interpretation of the results. The robustness checks have included modelling refinements (along the lines suggested by Unilever) and re-estimations, implementation of weak instrument robust parameter tests, and cost calibration exercises. The results have shown that (i) the estimation/identification methodology, within the framework of the models used, is reliable; and (ii) if the results change due to the modelling refinements the direction of change is upwards: the predicted price increases are higher than those of the Statement of Objections.

(187) Overall, the final assessment of the modelling is that the estimated price effects are robust and most likely conservative in the sense that they do not overpredict the likely price effects of the transaction. Still, because of the statistical and modelling uncertainty, which is inevitable in such exercises, the interpretation has to be cautious, and the results should be nested into the collection of the other qualitative and quantitative evidences that is available.

(187) Overall, the final assessment of the modelling is that the estimated price effects are robust and most likely conservative in the sense that they do not overpredict the likely price effects of the transaction. Still, because of the statistical and modelling uncertainty, which is inevitable in such exercises, the interpretation has to be cautious, and the results should be nested into the collection of the other qualitative and quantitative evidences that is available.

Table 7: Overview of estimated price increases

Table 7: Overview of estimated price increases

89These limitations of the model involve the implied substitution patterns (see section 3.3 of the Technical Annex), as well as the form of competition (no entry, fixed retailer markup; see section 4.4 of the Technical Annex).

89 These limitations of the model involve the implied substitution patterns (see section 3.3 of the Technical Annex), as well as the form of competition (no entry, fixed retailer markup; see section 4.4 of the Technical Annex).

90See Annex 4 of the Reply to the Statement of Objections and section 7 of the Technical Annex.

90 See Annex 4 of the Reply to the Statement of Objections and section 7 of the Technical Annex.

40

40

Overall Genders Brands

Overall Genders Brands

male non-AXDOVE REXONA SANEX IMPULSE VASELINE male E

male non-AXDOVE REXONA SANEX IMPULSE VASELINE male E

Belgium 4.2 1.2 6.2 0.5 10.3 3.5 18.2

Belgium 4.2 1.2 6.2 0.5 10.3 3.5 18.2

Spain 2.1 2.3 2 1.7 3.4 3.4 5.3

Spain 2.1 2.3 2 1.7 3.4 3.4 5.3

The Netherlands 3.8 1.1 5.7 0.7 10.2 2.8 20.6 10.9

The Netherlands 3.8 1.1 5.7 0.7 10.2 2.8 20.6 10.9

The United Kingdom 2.5 1 4 0.7 2.8 2.5 30.7 1.9 2.7

The United Kingdom 2.5 1 4 0.7 2.8 2.5 30.7 1.9 2.7

Note: see the Technical Annex for the confidence intervals and more technical details.

Note: see the Technical Annex for the confidence intervals and more technical details.

Unilever expects [...]*

Unilever expects [...]*

91(188) [reference to parties’ internal documents] .

91(188) [reference to parties’ internal documents] .

92(189) [reference to parties’ internal documents] . [reference to parties’ internal 93 94documents]. [reference to parties’ internal documents]

92(189) [reference to parties’ internal documents] . [reference to parties’ internal 93 94documents]. [reference to parties’ internal documents]

Figure 4: Unilever's profitability across categories

Figure 4: Unilever's profitability across categories

[...]*

[...]*

9596(190) [...]*.[...]*.

95 96(190) [...]*.[...]*.

(191) [...]*.

(191) [...]*.

9798(192) [...]*

9798(192) [...]*

99100(193) [...]*.[...]*.

99 100(193) [...]*.[...]*.

91[reference to parties’ internal documents]

92[reference to parties’ internal documents]

91 [reference to parties’ internal documents]

93[reference to parties’ internal documents]

92 [reference to parties’ internal documents]

94[reference to parties’ internal documents]

93 [reference to parties’ internal documents]

95[...]*.

94 [reference to parties’ internal documents]

96[...]*.

95 [...]*.

97[...]*

96 [...]*.

98[...]*.

97 [...]*

99[...]*.

98 [...]*.

100[...]*.

99 [...]*.

41

100 [...]*.

(194) [...]*.

41

(195) [...]*.

(194) [...]*.

(196) [...]*.

(195) [...]*.

101(197) [...]*.

(196) [...]*.

(198) [...]*.

101(197) [...]*.

IV.2.3.1.3. Lack of countervailing buyer power

(198) [...]*.

The notifying party’s arguments

IV.2.3.1.3. Lack of countervailing buyer power

(199) The notifying party argues that any attempt to raise prices would be countered by the buyer power of retailers. Such countervailing buyer power would result from their double role as not only customers but also those controlling market access and competitors supplying private label products.

The notifying party’s arguments

(200) The notifying party explains that retailers because of this double role are able to acquire products from manufacturers at competitive prices, enabling them to make satisfactory margins. If retailers dislike the terms or conditions proposed by manufacturers, they can renegotiate them by using their possibility to grant access to shelves for the products of the Parties. They can threaten to delist, even in relation to stronger brands. This would be possible either through delisting of a supplier's secondary brand, refusal to stock new variants of the leading brand or reducing the number of stock keeping units ("SKUs") that are displayed of the leading brand.

(199) The notifying party argues that any attempt to raise prices would be countered by the buyer power of retailers. Such countervailing buyer power would result from their double role as not only customers but also those controlling market access and competitors supplying private label products.

(201) The notifying party also argues that retailers can determine the competitive success of a brand that is listed in a number of ways, such as deciding to increase or decrease shelf space, determining shelf location, pricing strategy and promotional opportunities 102and deciding whether to distribute a supplier's "new innovations". In this context, great attention is given to the number of SKUs of a given supplier which will be carried, how many facing each SKU has and where on the shelves this SKU will be placed (top, bottom or eye level). This is an area where competitors seek to persuade retailers to optimise the presentation and merchandising of the products through category management activities. [...]*

(200) The notifying party explains that retailers because of this double role are able to acquire products from manufacturers at competitive prices, enabling them to make satisfactory margins. If retailers dislike the terms or conditions proposed by manufacturers, they can renegotiate them by using their possibility to grant access to shelves for the products of the Parties. They can threaten to delist, even in relation to stronger brands. This would be possible either through delisting of a supplier's secondary brand, refusal to stock new variants of the leading brand or reducing the number of stock keeping units ("SKUs") that are displayed of the leading brand.

(202) The notifying party puts forward that retailers act as competitors of manufacturers through sales of private label products which have two main advantages from the supermarkets' point of view. Supplying own label products allows retailers to drive consumers' loyalty to their stores and improve volume sales. In addition, the notifying party submits that it gives retailers sharper visibility on manufacturing costs and an ability to exert some pricing pressure on suppliers of the deodorant category.

(201) The notifying party also argues that retailers can determine the competitive success of a brand that is listed in a number of ways, such as deciding to increase or decrease shelf space, determining shelf location, pricing strategy and promotional opportunities 102and deciding whether to distribute a supplier's "new innovations". In this context, great attention is given to the number of SKUs of a given supplier which will be carried, how many facing each SKU has and where on the shelves this SKU will be placed (top, bottom or eye level). This is an area where competitors seek to persuade retailers to optimise the presentation and merchandising of the products through category management activities. [...]*

(203) Unilever has, during the procedure, underlined that analysis of retailer buyer power should take into account all the elements mentioned in the Horizontal Merger 103 104 Guidelines. The Horizontal Merger Guidelinesdescribe countervailing buyer power as "the bargaining strength that the buyer has vis-à-vis the seller in commercial negotiations due to its size, its commercial significance to the seller and its ability to switch to alternative suppliers". They indicate a number of non-exclusive different ways in which buyer power could arise. This includes (i) the ability to resort, within a reasonable timeframe, to alternative sources of supply should the supplier decide to increase prices or to deteriorate quality of the conditions of delivery; (ii) the ability to threaten to vertically integrate into the upstream market; (iii) the ability to sponsor upstream promotion or entry for example by persuading a potential entrant to enter by placing large orders to this company, or (iv) the ability to refuse to buy other products manufactured by the supplier or delaying purchases, particularly for durable goods.

(202) The notifying party puts forward that retailers act as competitors of manufacturers through sales of private label products which have two main advantages from the

(204) The notifying party submitted that in previous decisions like Procter & 105Gamble/Gillette, but also in statements of national competition authorities within the 106Union, buyer power has been established.

supermarkets' point of view. Supplying own label products allows retailers to drive consumers' loyalty to their stores and improve volume sales. In addition, the notifying party submits that it gives retailers sharper visibility on manufacturing costs and an ability to exert some pricing pressure on suppliers of the deodorant category.

(205) In a situation where the combined entity holds high market shares with a significant advance compared to its main competitor and the merger eliminates one of the strongest competitive constraints on Unilever's brands, the proposed merger is likely to provide Unilever with the incentives to unilaterally increase prices or reduce promotional pressure and thereby cause consumer harm. Therefore, it has to be carefully analysed whether buyer power is capable of being a mitigating factor in a case which results in anti-competitive effects in several deodorants markets.

(206) However, such an analysis of buyer power has to be carried out on a case by case basis and cannot rely on previous decisions, in particular if those references are not related to 107the deodorant market or a neighbouring personal care market. Unilever fails to

(203) Unilever has, during the procedure, underlined that analysis of retailer buyer power should take into account all the elements mentioned in the Horizontal Merger 103 104 Guidelines. The Horizontal Merger Guidelinesdescribe countervailing buyer power as "the bargaining strength that the buyer has vis-à-vis the seller in commercial negotiations due to its size, its commercial significance to the seller and its ability to switch to alternative suppliers". They indicate a number of non-exclusive different ways in which buyer power could arise. This includes (i) the ability to resort, within a reasonable timeframe, to alternative sources of supply should the supplier decide to increase prices or to deteriorate quality of the conditions of delivery; (ii) the ability to threaten to vertically integrate into the upstream market; (iii) the ability to sponsor upstream promotion or entry for example by persuading a potential entrant to enter by placing large orders to this company, or (iv) the ability to refuse to buy other products manufactured by the supplier or delaying purchases, particularly for durable goods.

103 Unilever’s reply to the request for information dated 25 June 2010, paragraph 7-4 as well as Reply to the Statement of Objections Section 6.

(204) The notifying party submitted that in previous decisions like Procter & 105Gamble/Gillette, but also in statements of national competition authorities within the 106Union, buyer power has been established.

104 Horizontal Merger Guidelines paragraphs 64 to 67.

(205) In a situation where the combined entity holds high market shares with a significant advance compared to its main competitor and the merger eliminates one of the strongest competitive constraints on Unilever's brands, the proposed merger is likely to provide Unilever with the incentives to unilaterally increase prices or reduce promotional pressure and thereby cause consumer harm. Therefore, it has to be carefully analysed whether buyer power is capable of being a mitigating factor in a case which results in anti-competitive effects in several deodorants markets.

105 Commission Decision of 15 July 2005 in Case No. M.3732 - Procter & Gamble/Gillette.

(206) However, such an analysis of buyer power has to be carried out on a case by case basis and cannot rely on previous decisions, in particular if those references are not related to 107the deodorant market or a neighbouring personal care market. Unilever fails to

106 See Reply to the Article 6.1(c) Decision paragraph 2.34 as well as the Reply to the Statement of Objections, paragraphs 6.5 – 6.7. Reference is made to Ireland (food and grocery markets), the United Kingdom (grocery markets), the Netherlands (savoury snacks), France (no specific market), Nordic countries (food markets) and Spain (food market).

explain why in its view a generalisation of case specific findings with different market 108 structures compared to the present transaction would be suitable.Moreover, several of the documents mentioned by Unilever provide a more nuanced view and indicate that in addition to buyer power also the market power of branded suppliers has to be 109taken into account.

107 The Parties refer to paragraph 24 of the Procter & Gamble/Gillette Decision dealing with oral care products. There it is stated that "customers' buyer power is an important mitigating factor for potential competition problems brought about by the merger". While this factor was taken into account in the assessment of the horizontal overlap, it was ultimately argued that it was not sufficient to counteract the potential adverse effects of the merger. Later in the Decision, reference is made to buyer power, but in the context of potential conglomerate effects.

(207) The analysis of buyer power must address the question whether and to what extent the buyers would be in a position to resist such a price increase. If the buyer is able to withstand price increases through shifting of supplies, upstream vertical integration or delistings, the exercise of bargaining power is likely to prevent the merged entity to increase prices post-merger. If, on the contrary, retailers are not in a position to switch to sufficient alternatives or delist because some of the supplier's products are "must-have", that is, the retailers must keep on the shelves in order not to lose substantial sales, it is unlikely that countervailing buyer power will counteract anticompetitive effects of the merger.

43

(208) The arguments of the notifying party are addressed in the context of the key elements mentioned in the Horizontal Merger Guidelines. The respective margins of Unilever are used as an indication of the bargaining strength of the retailers and the notifying party. Before addressing the ability and incentive of retailers to resist price rises, it is also important to assess Unilever's strong position on the deodorant markets.

explain why in its view a generalisation of case specific findings with different market 108 structures compared to the present transaction would be suitable.Moreover, several of the documents mentioned by Unilever provide a more nuanced view and indicate that in addition to buyer power also the market power of branded suppliers has to be 109taken into account.

Unilever is already the leading supplier in deodorants

(207) The analysis of buyer power must address the question whether and to what extent the buyers would be in a position to resist such a price increase. If the buyer is able to withstand price increases through shifting of supplies, upstream vertical integration or delistings, the exercise of bargaining power is likely to prevent the merged entity to increase prices post-merger. If, on the contrary, retailers are not in a position to switch to sufficient alternatives or delist because some of the supplier's products are "must-have", that is, the retailers must keep on the shelves in order not to lose substantial sales, it is unlikely that countervailing buyer power will counteract anticompetitive effects of the merger.

(209) Even prior to the proposed transaction, Unilever was the leader in deodorants in several Member States and has a unique portfolio of brands including Axe (Lynx in the United Kingdom and Ireland), Rexona (Sure in the United Kingdom and Ireland) and Dove. This rather unique position is not only reflected in the respective market shares, but also in the weighted distribution Unilever is able to achieve across Member 110States, which is only matched by a few competitors.

(208) The arguments of the notifying party are addressed in the context of the key elements mentioned in the Horizontal Merger Guidelines. The respective margins of Unilever are used as an indication of the bargaining strength of the retailers and the notifying party. Before addressing the ability and incentive of retailers to resist price rises, it is also important to assess Unilever's strong position on the deodorant markets.

(210) In addition, if category management takes place in Member States for deodorants, it is 111 Unilever that acts as category manager or advisor.Although it is true that most retailers do not think category management actually leads to favouring own products as retailers verify the recommendations given by the category manager, this just demonstrates the importance of Unilever on the market and puts the company in a unique position vis-à-vis all its competitors, reinforcing its general degree of influence on the retailers and on the market.

Unilever is already the leading supplier in deodorants

108 In that respect, Unilever seems to ignore the particularities of the case at hand. At the same time, Unilever accuses the Commission in the Reply to the Statement of Objections that it had ignored specific characteristics of national markets, see paragraph 4.29.

(209) Even prior to the proposed transaction, Unilever was the leader in deodorants in several Member States and has a unique portfolio of brands including Axe (Lynx in the United Kingdom and Ireland), Rexona (Sure in the United Kingdom and Ireland) and Dove. This rather unique position is not only reflected in the respective market shares, but also in the weighted distribution Unilever is able to achieve across Member States, which is only matched by a few competitors.

109 See for example the Competition Commission, The supply of groceries in the UK: market investigation, 30 April 2008, paragraph 9.21: "we found that buyer power of even the largest grocery retailers may be offset by the market power possessed by suppliers of the most prominent branded goods".

(210) In addition, if category management takes place in Member States for deodorants, it is 111 Unilever that acts as category manager or advisor.Although it is true that most retailers do not think category management actually leads to favouring own products as retailers verify the recommendations given by the category manager, this just demonstrates the importance of Unilever on the market and puts the company in a unique position vis-à-vis all its competitors, reinforcing its general degree of influence on the retailers and on the market.

110 For details see Table 17 for Belgium, Table 26 for Ireland, Table 34 for Spain, Table 49 for the Netherlands, Table 54 for Portugal, Figure 56 for the United Kingdom.

108 In that respect, Unilever seems to ignore the particularities of the case at hand. At the same time, Unilever accuses the Commission in the Reply to the Statement of Objections that it had ignored specific characteristics of national markets, see paragraph 4.29.

111 Unilever acts as a category manager for several retailers in the United Kingdom, Ireland and Spain – see Annex 8.15 Form CO.

109 See for example the Competition Commission, The supply of groceries in the UK: market investigation, 30 April 2008, paragraph 9.21: "we found that buyer power of even the largest grocery retailers may be offset by the market power possessed by suppliers of the most prominent branded goods".

44

110 For details see Table 17 for Belgium, Table 26 for Ireland, Table 34 for Spain, Table 49 for the Netherlands, Table 54 for Portugal, Figure 56 for the United Kingdom.

(211) Unilever’s average gross margins across its brands are usually above [...]*% in each 112 Member State and often reach [...]*% in some Member States.While it is true that Unilever's business model requires that a significant part of the margin is spent on establishing the brand, once a strong brand has been established, it provides the 113supplier with bargaining power vis-à-vis its customers.

111 Unilever acts as a category manager for several retailers in the United Kingdom, Ireland and Spain – see Annex 8.15 Form CO.

(212) In conclusion, Unilever has already obtained a sizeable position within the deodorant market, which allows it to have a better bargaining position compared to its competitors. This position would be further enhanced by the proposed transaction and cannot be countered by retailers as their bargaining position is further weakened.

44

Retailers may pass-on price increases to end consumers

(211) Unilever’s average gross margins across its brands are usually above [...]*% in each Member State and often reach [...]*% in some Member States.While it is true that Unilever's business model requires that a significant part of the margin is spent on establishing the brand, once a strong brand has been established, it provides the 113supplier with bargaining power vis-à-vis its customers.

(213) An assessment of retailer buyer power focuses on the ability of retailers as customers to resist price rises. At the same time, the Horizontal Merger Guidelines explicitly refer to situations in which "it may be important to pay particular attention to the incentives of 114 buyers to utilise their buyer power".In this case, there are indications that even if one were to accept a certain purchasing power of retailers, it is not evident that in all cases a price increase of Unilever's products would, as such, contradict the economic interests of retailers.

(212) In conclusion, Unilever has already obtained a sizeable position within the deodorant market, which allows it to have a better bargaining position compared to its competitors. This position would be further enhanced by the proposed transaction and cannot be countered by retailers as their bargaining position is further weakened.

(214) Retailers and suppliers stated in the market investigation that key issues in negotiations for a retailer usually relate to the level of investments provided by suppliers (such as promotional support), expected sales growth, and the profit made by the retailers on selling the products.

Retailers may pass-on price increases to end consumers

(215) During such negotiations, the main focus of the retailers in price discussions is to achieve an advantage vis-à-vis the other retailers (or to avoid suffering a disadvantage) 115 and to secure their margins, rather than the actual price itself.[...]* Price increases have been accepted in the past by retailers if those appeared justified, usually by increases of raw material prices, re-launches or improvements of products (when a higher price could be justified by an increased perceived value for the consumer), or by an increased advertisement support.

(213) An assessment of retailer buyer power focuses on the ability of retailers as customers to resist price rises. At the same time, the Horizontal Merger Guidelines explicitly refer to situations in which "it may be important to pay particular attention to the incentives of 114 buyers to utilise their buyer power".In this case, there are indications that even if one were to accept a certain purchasing power of retailers, it is not evident that in all cases a price increase of Unilever's products would, as such, contradict the economic interests of retailers.

(216) Indeed, retailers across the Member States concerned generally confirmed that they 116 would accept price increases if applied generally in the market.As an Irish retailer puts it- "(a general price increase) would be amenable if all retailers were to implement

(214) Retailers and suppliers stated in the market investigation that key issues in negotiations for a retailer usually relate to the level of investments provided by suppliers (such as promotional support), expected sales growth, and the profit made by the retailers on selling the products.

117the price increase ". Likewise, a Spanish retailer indicated that it "would pass on the 118price increase as [the retailer's] competitive position would be the same".

(215) During such negotiations, the main focus of the retailers in price discussions is to achieve an advantage vis-à-vis the other retailers (or to avoid suffering a disadvantage) 115 and to secure their margins, rather than the actual price itself.[...]* Price increases have been accepted in the past by retailers if those appeared justified, usually by increases of raw material prices, re-launches or improvements of products (when a higher price could be justified by an increased perceived value for the consumer), or by an increased advertisement support.

(216) Indeed, retailers across the Member States concerned generally confirmed that they 116 would accept price increases if applied generally in the market.As an Irish retailer puts it- "(a general price increase) would be amenable if all retailers were to implement

(217) It is not excluded that a general increase in the list prices of certain products (such as certain Sanex SKUs), applied to all retailers simultaneously, could indeed be successfully implemented by Unilever post-merger, if such increases were managed in a way as to ensure that the relative competitive positions of retailers would not be affected.

117the price increase ". Likewise, a Spanish retailer indicated that it "would pass on the 118price increase as [the retailer's] competitive position would be the same".

(218) Unilever has argued that such a hypothesis, while theoretically logical, is not directly relevant to the question at hand and submitted several observations: [...]* Even if justifications have been provided, Unilever claims that retailers would still request compensation with rebates and refers to the response of [retailer], [retailer] and [retailer] indicating that an increase in list prices would not be acceptable without compensation.

(217) It is not excluded that a general increase in the list prices of certain products (such as certain Sanex SKUs), applied to all retailers simultaneously, could indeed be successfully implemented by Unilever post-merger, if such increases were managed in a way as to ensure that the relative competitive positions of retailers would not be affected.

(219) In addition, Unilever referred to the market investigation and argued that retailers expressed doubts on how they can be sure that Unilever indeed would uniformly increase prices across all customers – retailers therefore face uncertainty on whether their competitors would pass-on the price increase or not.

(218) Unilever has argued that such a hypothesis, while theoretically logical, is not directly relevant to the question at hand and submitted several observations: [...]* Even if justifications have been provided, Unilever claims that retailers would still request compensation with rebates and refers to the response of [retailer], [retailer] and [retailer] indicating that an increase in list prices would not be acceptable without compensation.

(220) Since retailers' ultimate aim is to remain competitive vis-à-vis other retailers, they would rather reduce their margins than lose customers. Thus, competition among retailers ensures that suppliers are under pressure to provide discounts or more promotions.

(219) In addition, Unilever referred to the market investigation and argued that retailers expressed doubts on how they can be sure that Unilever indeed would uniformly increase prices across all customers – retailers therefore face uncertainty on whether their competitors would pass-on the price increase or not.

(221) Unilever’s reference to the current situation in the United Kingdom and Ireland seems to be misplaced as it neglects the structural change in the market caused by the proposed transaction which would be applicable to all retailers active in the market. It transpires from the market investigation that retailers usually pass on price increases when they anticipate that it is not an idiosyncratic price increase, but rather a general one applicable 119to the entire market like for example an increase in raw materials.

(220) Since retailers' ultimate aim is to remain competitive vis-à-vis other retailers, they would rather reduce their margins than lose customers. Thus, competition among retailers ensures that suppliers are under pressure to provide discounts or more promotions.

(222) While several retailers confirmed that they could never be sure that their competitors would receive the same supply conditions, it was also indicated that retailers try to resolve the uncertainty by observing the pricing behaviour of the market before deciding 120on a pass-on.

(221) Unilever’s reference to the current situation in the United Kingdom and Ireland seems to be misplaced as it neglects the structural change in the market caused by the proposed transaction which would be applicable to all retailers active in the market. It transpires from the market investigation that retailers usually pass on price increases when they anticipate that it is not an idiosyncratic price increase, but rather a general one applicable 119to the entire market like for example an increase in raw materials.

(223) In light of these answers, it is concluded that the proposed transaction may allow Unilever, as the leading supplier often controlling more than half of the total market, to implement a price increase post-merger. First, the merger would result in a structural change in the market common to all retailers, therefore retailers have more certainty that the requested price increase they are facing is similar to the one for all competitors. Second, a general price increase would leave the retailers target margin and its relative

(222) While several retailers confirmed that they could never be sure that their competitors would receive the same supply conditions, it was also indicated that retailers try to resolve the uncertainty by observing the pricing behaviour of the market before deciding 120on a pass-on.

competitive position unchanged. Third, the price increase could be complemented by minor product modifications or repositioning, providing an additional context for the 121general price increase.

(223) In light of these answers, it is concluded that the proposed transaction may allow Unilever, as the leading supplier often controlling more than half of the total market, to implement a price increase post-merger. First, the merger would result in a structural change in the market common to all retailers, therefore retailers have more certainty that the requested price increase they are facing is similar to the one for all competitors. Second, a general price increase would leave the retailers target margin and its relative

(224) In any case, it cannot be concluded that retailers have the necessary buyer power to be able to counter likely price increases following the proposed merger.

competitive position unchanged. Third, the price increase could be complemented by minor product modifications or repositioning, providing an additional context for the 121general price increase.

Even if large retailers were to have some bargaining power, this is not the case for all retailers

(224) In any case, it cannot be concluded that retailers have the necessary buyer power to be able to counter likely price increases following the proposed merger.

(225) The countervailing buyer power defence rests on the presumption that powerful buyers will be able to protect themselves – and thus, ultimately, also final consumers – against adverse changes in the terms of supply following the notified merger. The size of the buyers is seen as one of the elements that can contribute to buyer power in the Horizontal Merger Guidelines: “It is more likely that large and sophisticated customers will possess this kind of countervailing buyer power than smaller firms in a 122 fragmented industry”.However, even if some buyers were to have some bargaining power, it has been found that not all buyers would have the same degree of it.

Even if large retailers were to have some bargaining power, this is not the case for all retailers

(226) In all the national markets for which competition concerns have been identified, the degree of concentration on the retail side varies as can be seen in Table 8. In all Member States, the difference between the leading retailers is often significant and in each Member State a number of smaller retailers are buying from the Parties.

(225) The countervailing buyer power defence rests on the presumption that powerful buyers will be able to protect themselves – and thus, ultimately, also final consumers – against adverse changes in the terms of supply following the notified merger. The size of the buyers is seen as one of the elements that can contribute to buyer power in the Horizontal Merger Guidelines: “It is more likely that large and sophisticated customers will possess this kind of countervailing buyer power than smaller firms in a 122 fragmented industry”.However, even if some buyers were to have some bargaining power, it has been found that not all buyers would have the same degree of it.

Table 8: Percentage of deodorant value sales accounted for by major retailers – Source: Form CO Customer 1 Customer 2 Customer 3 Customer 4 Customer 5 Customer 6 Combined

(226) In all the national markets for which competition concerns have been identified, the degree of concentration on the retail side varies as can be seen in Table 8. In all Member States, the difference between the leading retailers is often significant and in each Member State a number of smaller retailers are buying from the Parties.

BelgiumNational[30-40]* [20-30]* [10-20]* -- -- -- [60-70]* sales

Table 8: Percentage of deodorant value sales accounted for by major retailers – Source: Form CO Customer 1 Customer 2 Customer 3 Customer 4 Customer 5 Customer 6 Combined

Unilever[30-40]* [20-30]* [10-20]* -- -- -- [70-80]* sales

BelgiumNational[30-40]* [20-30]* [10-20]* -- -- -- [60-70]* sales

NetherlandsNational[20-30]* [10-20]* [10-20]* [5-10]* -- -- [60-70]* sales

Unilever[30-40]* [20-30]* [10-20]* -- -- -- [70-80]* sales

Unilever[30-40]* [30-40]* [10-20]* [10-20]* [5-10]* -- [90-100]* sales

NetherlandsNational[20-30]* [10-20]* [10-20]* [5-10]* -- -- [60-70]* sales

National[20-30]* [10-20]* [10-20]* [5-10]* [10-20]* [5-10]* [80-90]* sales

Unilever[30-40]* [30-40]* [10-20]* [10-20]* [5-10]* -- [90-100]* sales

United Kingdom

National[20-30]* [10-20]* [10-20]* [5-10]* [10-20]* [5-10]* [80-90]* sales

Unilever[30-40]* [10-20]* [10-20]* [5-10]* [10-20]* [5-10]* [80-90]* sales

United Kingdom

Ireland National[30-40]* [20-30]* [10-20]* [5-10]* [5-10]* -- [80-90]* sales

Unilever[30-40]* [10-20]* [10-20]* [5-10]* [10-20]* [5-10]* [80-90]* sales

121 [...]*

Ireland National[30-40]* [20-30]* [10-20]* [5-10]* [5-10]* -- [80-90]* sales

122 Horizontal Merger Guidelines, paragraph 65. See in addition paragraph 67 Horizontal Merger Guidelines: "Countervailing buyer power cannot be found to sufficiently off-set potential adverse effects of a merger if it only ensures that a particular segment of customers [...]* is shielded from significantly higher prices or deteriorated conditions after the merger."

121[...]*

122Horizontal Merger Guidelines, paragraph 65. See in addition paragraph 67 Horizontal Merger Guidelines: "Countervailing buyer power cannot be found to sufficiently off-set potential adverse effects of a merger if it only ensures that a particular segment of customers [...]* is shielded from significantly higher prices or deteriorated conditions after the merger."

47

47

Unilever[30-40]* [20-30]* [10-20]* [0-5]* [5-10]* -- [80-90]* sales

Unilever[30-40]* [20-30]* [10-20]* [0-5]* [5-10]* -- [80-90]* sales

National[10-20]* [10-20]* [5-10]* [5-10]* -- -- [40-50]* sales

National[10-20]* [10-20]* [5-10]* [5-10]* -- -- [40-50]* sales

Spain

Spain

Unilever[10-20]* [20-30]* [5-10]* [5-10]* -- -- [50-60]* sales

Unilever[10-20]* [20-30]* [5-10]* [5-10]* -- -- [50-60]* sales

PortugalNational[30-40]* [10-20]* [10-20]* [5-10]* -- -- [70-80]* sales

PortugalNational[30-40]* [10-20]* [10-20]* [5-10]* -- -- [70-80]* sales

Unilever[20-30]* [10-20]* [5-10]* [5-10]* -- -- [60-70]* sales

Unilever[20-30]* [10-20]* [5-10]* [5-10]* -- -- [60-70]* sales

DenmarkNational[30-40]* [10-20]* [5-10]* [20-30]* [10-20]* [0-5]* [90-100]* sales

DenmarkNational[30-40]* [10-20]* [5-10]* [20-30]* [10-20]* [0-5]* [90-100]* sales

Unilever[20-30]* [20-30]* [10-20]* [20-30]* [5-10]* [0-5]* [90-100]* sales

Unilever[20-30]* [20-30]* [10-20]* [20-30]* [5-10]* [0-5]* [90-100]* sales

(227) In Belgium for example, the three largest retailers account for [60-70%]* of the total national deodorants sales, but the leading retailer accounts for [30-40%]* - almost three times as large as the third leading retailer. The size of retailers varies significantly in the Netherlands as well. The four largest retailers account for [60-70%]* of the total national deodorants sales. While the leading retailer represents [20-30]*% of the sales, the fourth largest only accounts for [5-10%]* of the sales. In Spain, the degree of concentration on the retailer side is significantly lower compared to the other Member States.

(227) In Belgium for example, the three largest retailers account for [60-70%]* of the total national deodorants sales, but the leading retailer accounts for [30-40%]* - almost three times as large as the third leading retailer. The size of retailers varies significantly in the Netherlands as well. The four largest retailers account for [60-70%]* of the total national deodorants sales. While the leading retailer represents [20-30]*% of the sales, the fourth largest only accounts for [5-10%]* of the sales. In Spain, the degree of concentration on the retailer side is significantly lower compared to the other Member States.

(228) Although varying size is already an indication that not all retailers have the same importance as a potential distributor for Unilever to get access to the market, Unilever 123 re-iterated in several submissionsthat it is not correct to claim that countervailing buyer power is limited to the larger one or two customers in the market such that the benefits will not felt by customers more generally.

(228) Although varying size is already an indication that not all retailers have the same importance as a potential distributor for Unilever to get access to the market, Unilever 123 re-iterated in several submissionsthat it is not correct to claim that countervailing buyer power is limited to the larger one or two customers in the market such that the benefits will not felt by customers more generally.

(229) In particular, Unilever argued that even small retailers with a market share of 5% represent a significant volume and value of Unilever's deodorants sales, which can not be distributed through other channels in case of a delisting. Such a loss can only be ignored if one assumes that customers would switch retailers if they were not able to find their preferred brand. However, according to Unilever, no examples are available showing that customers would indeed behave in such a way. To the contrary, examples of delistings in the United Kingdom and Spain would suggest the opposite.

(229) In particular, Unilever argued that even small retailers with a market share of 5% represent a significant volume and value of Unilever's deodorants sales, which can not be distributed through other channels in case of a delisting. Such a loss can only be ignored if one assumes that customers would switch retailers if they were not able to find their preferred brand. However, according to Unilever, no examples are available showing that customers would indeed behave in such a way. To the contrary, examples of delistings in the United Kingdom and Spain would suggest the opposite.

(230) Moreover, Unilever submitted, the intense competition among retailers provides sufficient incentives for small retailers to form buying groups or employing a different business model. In particular, buying groups should be considered as "formidable 124negotiators".

(230) Moreover, Unilever submitted, the intense competition among retailers provides sufficient incentives for small retailers to form buying groups or employing a different business model. In particular, buying groups should be considered as "formidable 124negotiators".

123 Reply to the Article 6.1(c) Decision, paragraphs 2.35 to 2.39 and Reply to the Statement of Objections, paragraphs 6.16 to 6.25.

123 Reply to the Article 6.1(c) Decision, paragraphs 2.35 to 2.39 and Reply to the Statement of Objections, paragraphs 6.16 to 6.25.

124 See the Reply to the Statement of Objections, paragraph 6.23.

124 See the Reply to the Statement of Objections, paragraph 6.23.

48

48

(231) [...]*

(231) [...]*

(232) The arguments put forward by the notifying party are not sufficiently convincing to conclude that smaller retailers have the same bargaining power as large ones. Unilever suggests that size does not matter when determining bargaining power (and therefore ultimately buyer power) of a retailer, which is a questionable assumption. The smaller the buyer, the easier it would be for a supplier to sell the delisted products to other buyers: while it would be possible for Unilever to shift 5% of its sales to other retailers very easily for example, shifting 30% of the entire volume would only be possible if a significant incentive were to be granted lowering prices and profits. Thus, while a [...]*% loss in the United Kingdom might, as Unilever argued, be equivalent to its sales in Denmark or Ireland, this comparison is meaningless when assessing Unilever's position in the United Kingdom. It is not clear why Unilever's economic dependency on an individual retailer is determined by the retailer's size in proportion to another national market in which Unilever is active. It can rather be argued that for a supplier with a global presence in deodorants like Unilever, dependency on a retailer in a particular Member State is very unlikely as each retailer represents only a minor part of the notifying party's deodorant business.

(232) The arguments put forward by the notifying party are not sufficiently convincing to conclude that smaller retailers have the same bargaining power as large ones. Unilever suggests that size does not matter when determining bargaining power (and therefore ultimately buyer power) of a retailer, which is a questionable assumption. The smaller the buyer, the easier it would be for a supplier to sell the delisted products to other buyers: while it would be possible for Unilever to shift 5% of its sales to other retailers very easily for example, shifting 30% of the entire volume would only be possible if a significant incentive were to be granted lowering prices and profits. Thus, while a [...]*% loss in the United Kingdom might, as Unilever argued, be equivalent to its sales in Denmark or Ireland, this comparison is meaningless when assessing Unilever's position in the United Kingdom. It is not clear why Unilever's economic dependency on an individual retailer is determined by the retailer's size in proportion to another national market in which Unilever is active. It can rather be argued that for a supplier with a global presence in deodorants like Unilever, dependency on a retailer in a particular Member State is very unlikely as each retailer represents only a minor part of the notifying party's deodorant business.

(233) In addition, comparing smaller retailers' share of Unilever' national sales with the notifying party's overall market share in a Member State shows that the relationship between those retailers and Unilever is rather asymmetric. While smaller retailers usually represent less than 10% of Unilever's deodorant sales, Unilever would, post transaction, account for at least [40-50]*% of national non-male-specific deodorant sales across Belgium, Denmark, Ireland, the Netherlands, Spain, Portugal and the United Kingdom. Consequently, replacing Unilever and its well-known brands is more difficult for a retailer than it is for Unilever to find alternative ways of distribution for a relative small share of its sales. Indeed, a number of retailers confirmed in their reply to the market investigation that Unilever has a unique position within the deodorant category as customers expect its brands Axe, Rexona and Dove to be on the shelves, and indicated that a substantial amount of their customers would switch to other retailers if these brands were no longer available.

(233) In addition, comparing smaller retailers' share of Unilever' national sales with the notifying party's overall market share in a Member State shows that the relationship between those retailers and Unilever is rather asymmetric. While smaller retailers usually represent less than 10% of Unilever's deodorant sales, Unilever would, post transaction, account for at least [40-50]*% of national non-male-specific deodorant sales across Belgium, Denmark, Ireland, the Netherlands, Spain, Portugal and the United Kingdom. Consequently, replacing Unilever and its well-known brands is more difficult for a retailer than it is for Unilever to find alternative ways of distribution for a relative small share of its sales. Indeed, a number of retailers confirmed in their reply to the market investigation that Unilever has a unique position within the deodorant category as customers expect its brands Axe, Rexona and Dove to be on the shelves, and indicated that a substantial amount of their customers would switch to other retailers if these brands were no longer available.

(234) As one retailer based in the United Kingdom explains with regard to the balance of power vis-à-vis Unilever: "retailers do have good negotiation powers but on deodorants this isn't weighted in our favour as Unilever has key brands that 126 customers want and this therefore strengthens their position".Another retailer explains that "Customers expect certain products to be stocked by retailers (including most obviously the Unilever brands), and for there to be a wide variety of choice. Therefore retailers are required to ensure these brands are listed or face the prospect 127of losing sales to competitors."

(234) As one retailer based in the United Kingdom explains with regard to the balance of power vis-à-vis Unilever: "retailers do have good negotiation powers but on deodorants this isn't weighted in our favour as Unilever has key brands that 126 customers want and this therefore strengthens their position".Another retailer explains that "Customers expect certain products to be stocked by retailers (including most obviously the Unilever brands), and for there to be a wide variety of choice. Therefore retailers are required to ensure these brands are listed or face the prospect 127of losing sales to competitors."

125 See replies to Questionnaire to customers on deodorants sent on 25 June 2010, question 17. In all seven Member States for which competition concerns have been identified in this Decision, at least half of the responding retailers indicated that a significant number of their customers would switch to another retailer if particular brands were not available.

125 See replies to Questionnaire to customers on deodorants sent on 25 June 2010, question 17. In all seven Member States for which competition concerns have been identified in this Decision, at least half of the responding retailers indicated that a significant number of their customers would switch to another retailer if particular brands were not available.

126 Reply of [retailer]'s to Questionnaire to customers on deodorants sent on 25 June 2010, question 16

126 Reply of [retailer]'s to Questionnaire to customers on deodorants sent on 25 June 2010, question 16

127 Reply of an anonymous British retailer to Questionnaire to customers on deodorants sent on 25 June 2010, question 16.

127 Reply of an anonymous British retailer to Questionnaire to customers on deodorants sent on 25 June 2010, question 16.

49

(235) Moreover, Unilever has not been able to demonstrate that delistings of smaller retailers lead to a significant decline in its output. The Reply to the Statement of Objections refers to the Reply to the Article 6.1(c) Decision, in which examples of delistings, [...]* were put forward to show Unilever’s loss of turnover with these companies as a result of a delisting. Although these retailers are not small in size, the delistings have been analysed in more detail.

129 128(236) [...]*. ,While it is correct to say that Unilever lost turnover as a consequence of delisting with that particular customer, a detailed analysis of the [...]* market before and after the delisting at [...]* indicates that Unilever’s claim that the delisting was the reason for its declining market share in [...]* cannot be supported.

(237) Comparing Unilever’s sales before and after the delisting [...]* with its sales at all other retailers, it can be concluded that the value sales [...]* do not significantly differ from the evolution of the rest of the market during that period [...]* indicating that a decline in sales at [...]* could also be explained by other factors prevalent in the 130overall market. [...]*

Table 9: Comparison of Unilever's deodorant sales in [a specific retailer] and outside [the specific retailer] – own calculations based on wholesale value data – September 2008 = 100.

[...]*

(238) The second example is linked to different products [...]*, and therefore not applicable 131 to the deodorants market.The third case provides an indication that customers would switch retailers if several Unilever brands were not available at the same time. 133132 [...]*.[...]*

(239) With the exception of Superunie, none of the buying groups mentioned by Unilever have been included in the list of the six main customers provided by Unilever for the purposes of the market investigation. This tends to show that none of these buyers have the same 134 size and purchasing capacity as the Parties' main retail customers.In addition, Unilever only submits a general description on different forms of alliances like buying groups, franchises or symbol groups.What becomes clear from Unilever's

128 [...]*

129 [...]*

130 [...]*

131 [...]*

132 [...]*

133 [...]*

134 Indeed, most buying groups represent between [...]* of Unilever's national deodorant sales. The only exceptions can be found in Spain, the Netherlands and Denmark, where alliances account for [...]*

135 See paragraphs 6.20-6.24 of the Reply to the Statement of Objections.

50

submission is that most of these organisations lack the clear hierarchical structure of a central organisation as large retailers like [retailer], [retailer] and others enjoy. Therefore, central elements and pre-conditions (like for example same product range, planogram, promotion programme) for potential bargaining power seem to lack from the outset. [...]*. Based on the information available about the scope of the negotiations of buyer groups compared to retailers, it cannot be concluded that buyer groups are in the same position of strength as large retailers.

136 (240) Unilever also provided a comparison between Unilever’s gross margin (as a percentage of its turnover) on deodorant sales to its five largest customers in Member States where concerns have been identified with Unilever’s average gross margin on sales to all its other customers in all countries.

(241) [...]*. These variations seem to suggest that Unilever promotes brands differently across retailers and might even be able to use retailers strategically.

(242) As to the argument that competition among retailers will equalise the gross margins Unilever achieves across retailers and ultimately ensure that end-customers are shielded from adverse effects of the transaction, Unilever explained in the Reply to the Statement of Objections that buyer power of larger retailers will be transmitted to smaller ones as follows: "The very fact that a smaller retailer expects a larger retailer to get a lower price improves the small retailer's bargaining power since it has less to lose from disagreement. The retailer has less to lose from delisting the products since it would earn less from these products anyway if they are sold at a lower price in the lager stores."

(243) An important assumption for this mechanism to work is that customers were to switch stores in case of a price difference. However, Unilever fails to explain why customers would switch in case of a price difference between retailers if they were – according to Unilever – not willing to switch in case their preferred brand would not be available in a supermarket. Absence of switching would ultimately result in a lack of bargaining power on behalf of smaller retailers. Even if switching were to occur, it is not clear from Unilever's submission how the feedback mechanism in the bargaining between different retailers would ultimately result in similar conditions across all retailers independent of their size or business model.

(244) It seems therefore valid to conclude that the equal margins across retailers are - contrary to Unilever's submission - not an indication that all Unilever's customers have strong bargaining position. They only show that all customers, small or large, have 137roughly the same bargaining position (or lack thereof) vis-à-vis Unilever. [...]* .

(245) In light of the absence of convincing arguments why smaller retailers would have market power and the [...]* margins achieved by Unilever in these markets for small and large customers (between [...]*% and [...]*%), it is logical to conclude that Unilever holds significant selling power in relation to both large and small customers.

Ability to integrate vertically as an element of buyer power - private label deodorants

136 [...]*.

137 [...]*

51

(246) The deodorant market is primarily driven by branded products. The share of private label in deodorants is remarkably low when it is compared to many other consumer 138 goods markets.This is a strong indicator that customers are not in a position to credibly threaten to vertically integrate into the upstream market in the sense of the Horizontal Merger Guidelines.

(247) Although Unilever agrees with the analysis that the penetration of private label in the 139deodorant category is lower than in other consumer goods markets, this should not be seen as an indication that retailers would not be able to vertically integrate. Moreover, the share of private label in the overall deodorant market exceeds Sanex share either in value or in volume.

140 (248) Furthermore, Unilever arguesthat price competition between brands is so intense that there is no need for retailers to add private label to alter their bargaining position in relation to branded suppliers. However, even a small share of private label allows the retailer to increase competition among suppliers for shelf space as well as gaining 141insights into the underlying cost structure.

(249) Finally, Unilever points to replies of retailers in the market investigation stating that they have the intention to increase their private label offers during the next 1-2 142years.

(250) Reviewing the notifying party's arguments, it is noted that competition concerns have mainly been identified in the non-male deodorants markets in several Member States. In those non-male markets, the private label share in value never exceeds the one of Sanex. In Member States such as Denmark, Sanex is ten times larger than private label. In the Spanish male deodorants market, Sara Lee also has a higher value share than private label brands.

(251) Manufacturers active across the EEA perceive private labels as small and limited 143 competitors, with hardly any impact on their pricing– unless the brand is positioned at the entry level, which is not the case for Unilever brands, which are positioned in the medium to high-end segment. Although competitors confirmed Unilever's submission that private label deodorants allow retailers to have insights into the suppliers' production costs, a majority indicated as well that private label will continue

138 See Decision in case M.5644 Kraft Foods /Cadbury of 6 January 2010 where the share of private label for chocolate tablets was [10-20]*% in UK and Ireland and [20-30]*% in Portugal and France. See also decision in case M.4533 SCA/P&G of 5 September 2007 where private labels were found to "predominate" in Western Europe in toilet papers, handkerchiefs and kitchen towels.

139 See paragraph 6.26 of the Reply to the Statement of Objections.

140 See paragraph 6.28 of the Reply to the Statement of Objections.

141 See paragraph 6.29 of the Reply to the Statement of Objections.

142 See paragraph 6.30 of the Reply to the Statement of Objections.

143 The exception is Spain where private label has a market share of [20-30]*% in the non-male market and local competitors feel the impact of retailers' own brands on their pricing – see Questionnaire to competitors on deodorants sent on 24 April 2010, question 52.

52

to have a limited presence because deodorants require a significant amount of investment in advertisement, brand awareness and promotion, which makes it difficult 144for private label to prosper.

(252) This view was confirmed by customer replies in the market investigation. Even in the current economic situation, customers do not appear interested to switch to less 145 expensive private label products to any significant extent.Both customers and competitors confirmed that the private label option is not flagged by retailers in the 146negotiations with branded good suppliers.

(253) When asked why private label penetration is particular low for deodorants, a number of reasons were provided. Several retailers referred to the importance of brands and the need to invest into advertisement and marketing since the product is very "personal and sensitive". Private label brands could not compete on that level, in particular in those Member States were according to a few retailers prices for branded products are 147 low.Several retailers indicated that they expect to introduce or re-launch private 148label deodorants, but on a small scale.

(254) Based on low degree of substitutability between branded and private label deodorants among end-consumers, it does not appear that supermarket chains can credibly threaten to source private label in negotiating prices for branded products. It is therefore concluded that upstream vertical integration as an alternative to sourcing branded products is not an important source of buyer power for retailers in the deodorant 149market .

Delisting of brands by retailers

(255) With respect to delisting, a distinction should be made between "regular" delisting and non-regular delisting. Regular delisting is decided annually by retailers in the framework of a category review and they tend to eliminate products with a poor track record in 150terms of performance. This type of delisting is often the result of discussions between the retailer and the category manager or advisor who is frequently being asked to identify

144 Questionnaire to competitors on deodorants sent on 24 April 2010, questions 48, 49 and 51.

145 Questionnaire to customers on deodorants sent on 25 June 2010, question 23.

146 Questionnaire to customers on deodorants sent on 23 April 2010, question 35 and Questionnaire to competitors on deodorants sent on 24 April 2010, question 51.

147 See remark by a British retailer in its response to Questionnaire to customers on deodorants sent on 23 April 2010, question 32(d) as well as an Irish retailer in its response to Questionnaire to customers on deodorants sent on 25 June 2010, question 23.

148 See replies to Questionnaires to customers on deodorants sent on 25 June 2010, question 23. The replies for Spain deviate from those of the other Member States as private label in Spain has already achieved a relatively high presence. For a detailed analysis of private labels see the country specific analysis in section IV.2.3.2.2.

149 The merger simulations fully take into account the competitive pressure that private label exercise on the branded deodorants when consumers make their choice in the supermarket.

150 Form CO, page 387.

53

underperforming SKUs for delisting, on the basis of objective criteria defined by the retailers. In that role, a category advisor must make recommendations on an objective and transparent basis. As Unilever rightly points out "Recommendations to delist SKUs must be backed up by sales or performance data. Thus even when Unilever is category adviser it may recommend to delist its own SKUs if they are underperforming" .

(256) Retailers have unanimously confirmed that they undertake on a regular basis (one or several "windows" per year) this category review which might lead to delistings of 152underperforming SKUs). For instance, a Danish retailer explained that "we have 5-6 annual so called windows, where new products are listed and bad selling items are de-listed. In 2010 this happens in week 2+10+22+36+44. The general principle is one in = one out, which means that when a new item is listed another is de-listed, thereby ensuring that we do not overfill the shops". When asked about the reasons used to determine which SKU should go out, [retailer] submitted that it is linked to "Poor sales figure, bad development in profitability or simply because another item is seen as more valuable in turnover and/or margin".

(257) Likewise, a French retailer submitted that listing new items is likely to lead to a removal of some products from the shelves "Due to shelves capacity's limits, adding new references result on a delisting of other references (from the same supplier). This kind of assortment strategy is fully shared with the suppliers named before, who react in a 153constructive way given that our objective is to optimize our category growth plan. ".

154 (258) Unilever agrees with the distinction of delisting into regular and non-regular. It submits that non-regular delisting is a rare event at the end of an escalating level of threats available to the retailer and refers to examples presented to the Commission in the course of the investigation, which would clearly show that retailers have buyer power.

(259) Unilever's examples, however, are often not put into context. Sometimes they refer only to a threat of delisting without specifying whether the threat was executed or Unilever gave in. Moreover, they do not clearly distinguish between a regular delisting because of 155poor performance and a delisting during negotiations.

(260) In addition, Unilever and retailers negotiate on a bi-lateral basis, thus, both sides take into account "threats" during the bargaining and anticipate such steps in their preparation – therefore, just looking at the steps taken by retailers during this bargaining procedure

151 Form CO, page 388.

152 Questionnaire to customers on deodorants sent on 23 April 2010, question 49.

153 "Les gondoles et notre capacité linéaire n'étant pas extensibles, ce rajout de référence éventuel se fait au détriment d'autres produits (pour un même fournisseur). Ce type de stratégie d'assortiment est entièrement partagé avec les fournisseurs cités, qui réagissent dans un esprit parfaitement constructif puisque l'objectif est d'optimiser notre plan de croissance catégoriel."

154 See paragraph 6.31 of the Reply to the Statement of Objections.

155 [...]*

54

gives the wrong impression as it ignores not only Unilever's possibilities, but also the 156ultimate objective for each party during the negotiations.

157(261) [...]*

(262) The overall portfolio of products a supplier is selling to a retailer is important if disagreement between the branded good supplier and a retailer about pricing arise. [...]*

158159(263) [...]* ,

(264) [...]*

(265) As a result, there may be financial consequences for the retailer when it decides whether to promote sales of a given Unilever SKU or brand or not. These consequences might go beyond the specific products in question, and could spill over to other deodorant brands. [...]*

(266) In the Reply to the Statement of Objections, Unilever refers to several statements made by competitors, which, according to the notifying party, confirms that retailers 160 have buyer power.A closer look at those references shows that several competitors submitted important qualification on when retailers have buyer power. Beiersdorf, for example, states that "the smaller the manufacturer is, the more easily the retailer may exert its buyer-power [...]* in relation to deodorants, it is Unilever where the buyer-power faces limits.

Henkel adds that buyer power is "Diminished when the manufacturer has a very large overall product portfolio, a significant market share for a certain product category or a "must have" brand [...]*. Due to all these factors, retailers are less likely to delist a Unilever brand.

L'Oreal also refers to the size of the supplier and refers in particular to Unilever with its important brand as an example 161where such a buyer power would be limited.

(267) The relevance of Unilever's size and the number of brands in its deodorant portfolio is also reflected in customers' replies. The ability for a retailer to switch to other suppliers and the ability to deny access to final customers are closely linked. Several retailers in the Netherlands, the United Kingdom and Spain have identified Unilever's brands for deodorants as must-have brands ("brands with few relevant alternatives which need to 162be kept on the shelves to avoid loss of substantial sales"). Delisting top brands with

156 [...]*

157 [...]*

158 See Annex 8-11 Form CO.

159 Unilever has a standard published national list price for its products with standard terms and conditions regarding matters such as payment terms, minimum order quantity and order and delivery conditions. These list prices are the starting point for negotiations with retailers on discounts and rebates

160 See paragraph 6.33 of the Reply to the Statement of Objections, where Unilever refers to Henkel, Beiersdorf, Procter & Gamble, Colgate-Palmolive and L'Oreal.

161 See replies to Questionnaire to competitors on deodorants sent on 25 June 2010, question 5.

162 Questionnaire to customers on deodorants sent on 23 April 2010, question 53.

55

high consumer recognition would be detrimental for them. Moreover, Unilever stands out due to the strength of its overall brand portfolio. Delisting important Unilever brands such as Axe, Dove or Rexona would clearly endanger retailers' turnover in the deodorant 163category.

(268) In its reply to the Article 6(1) (c) Decision as well as in its Reply to the Statement of 164Objections, Unilever rejected the suggestion that any of its brands are must-have brands. Unilever argues that while a retailer delisting is a lost profit that Unilever cannot recover, retailers can replace Unilever products with those of rivals, thereby largely compensating them for any loss. On the other hand, it would be hard for Unilever to find alternative outlets to market for the product not sold through the retailer. It argues that consumers are not likely to switch stores to the detriment of the retailer in the event that they cannot find their usual brand of deodorant in any particular store.

(269) To substantiate its argument – which is contrary to the feedback received during the 165 166 market investigation- Unilever provided brand loyalty figuresfor the Parties and their competitors' brands. This information was provided by GFK/Europanel which defined loyalty as "the proportion of a brand buyer's category purchasing that goes to the brand within a given 12 month period". As an example, if a shopper spends EUR 10 on deodorants in a given year, EUR 3.5 of which is spent on Dove, then that shopper's loyalty to Dove is recorded at 35%.

(270) The figures provided by the notifying party do not necessarily mean that brand loyalty would be particularly low in the deodorant markets or that customers of the Parties' brands would be less loyal toward their brands than consumers of competitors' brands. In the United Kingdom, the highest loyalty rates in value are for Rexona/Sure ([...]* ) and Axe/Lynx ([...]*) whereas it is [...]* for Gillette and [...]* for Soft & Gentle and Nivea. In the Netherlands, Rexona ([...]*), Nivea ([...]*), Sanex ([...]*) and Axe ([...]* ) have the highest loyalty rates in value. A similar picture arises in Belgium and in Spain. Finally in Ireland, Axe/Lynx ([...]*) and Palmolive ([...]*) have very loyal consumers followed by Rexona/Sure ([...]*). In general, loyalty rate for Dove is

163 See replies to Questionnaire to customers on deodorants sent on 25 June 2010, question 17. Almost all retailers across the eight Member States covered by in-depth investigation indicated that their customers would either switch their deodorant purchases or their entire purchases to another retailer in case one of the core brands of the Parties would not be available in their stores.

164 Reply to the Article 6(1)(c) Decision, paragraphs 2.10 to 2.19 and the Reply to the Statement of Objections, paragraphs 6.35 to 6.41.

165 See replies to Questionnaire to customers on deodorants sent on 25 June 2010, question 17.

166 Reply from Unilever to a RFI dated 25 June 2010, paragraphs 4-1 to 4-7 and Annex 4.

167 The figure will therefore be influenced by the price as well as the sales volume of the product: more expensive products will count more towards the shopper's spend in that particular year. Moreover, market shares have also to be taken into account when analysing these purchasing patterns since it is more likely in the case of a brand with a high market share that a household which has initially purchased this brand, purchases the same brand in all its subsequent purchases. Finally, these figures bring information regarding the purchasing patterns of buyers of the products (for example a housewife purchasing deodorants for the whole family) but they are of limited relevance to assess the strength of the attachment of the end-consumer (the deodorant user) towards the brand.

56

(299) In assessing buyer power pre-merger, it is interesting to compare the margins Unilever obtains with the margins retailers derive with respect to the products at stake. In a negotiation context, the relative strength of the two sides should be reflected in the benefits that each party (the supplier and the retailer) manages to earn on the specific product group over which they are bargaining.

(300) Unilever has argued that comparing supplier and retailer margins would be a misconception of the margin that is relevant for the assessment of relative bargaining 187 power.According to Unilever, differences in accounting operating margins for a downstream and an upstream business directly reflects differences in business models between the two brands – with a supplier/manufacturer incurring high investments in innovation and brand development, and a distribution/retail business focused on achieving high volumes. Instead Unilever has provided data on Unilever's and retailers' "share of surplus", where the surplus is defined as the sum of Unilever's surplus and retailers' surplus. Unilever's surplus [...]*, while the retailers' surplus is calculated as net retail price (net of value added tax, any discounts, buy-one-get-one free offers and so forth), less wholesale cost (which is the wholesale price paid to Unilever, net of any rebates, promotional support).

(301) For the year 2008, Unilever concluded that its share if total margin was [...]* depending 188 on the Member Stateconfirming in Unilever's view that retailers have material bargaining power.

(302) However, the results of the surplus split cannot directly be transferred into a measurement of bargaining power as many other variables might have an impact on the surplus which can be achieved. Factors having an impact are the concentration at the supplier level, the degree of concentration at the retail level, the intensity of competition among market participants as well as country specific factors (purchasing power, elasticity of demand, business cycle). In particular, the price to end-consumers is subject to regular changes and is only determined after the negotiations between a supplier and a retailer. Unilever's lower share of the surplus in [...]* for example could be the result of its lower market share of around [20-30]*% and the presence of close competitor like Sara Lee, while higher margins in [...]* are potentially a result of Unilever's market share.

(303) Even if the share of surplus figures were an indication of bargaining power, it is not sufficient that buyer power exists prior to the merger, it must also exist and remain 189 effective following the merger.However, the main effect of the merger - as also 190 Unilever concedes in the Reply to the Statement of Objections– removes a

187 Reply from Unilever to a RFI dated 25 June 2010, paragraph 7-12 and paragraph 6.58 of the Reply to the Statement of Objections.

188 Unilever provided data for the seven Member States covered in the Statement of Objections. Its share of the total surplus was [60-70]*% in Belgium, [30-40]*% in Denmark, [30-40]*% in Ireland, [50-60]*% in the Netherlands, [40-50]*% in Spain, [40-50]*% in the United Kingdom and [30-40]*% in Portugal.

189 Horizontal Merger Guidelines, paragraph 67.

190 See Annex 7 of the Reply to the Statement of Objections where it is stated in paragraph 36 that "the only relevant effect of the merger is therefore a reduction in the number of good substitutes [...]*. This worsens

substitute to Unilever's brands and therefore shifts bargaining power from the retailer to the supplier that is Unilever. While Unilever tries to play down this effect by arguing that in-store switching is more prevalent than moving shops, it fails to provide any evidence that goes beyond mere speculation – referring to literature from 1976 and 1985 not linked to the deodorant market and the delisting examples of [...]*, which as described in recital (235) - (238) are not supportive either. In addition, a majority of respondents indicated in the market investigation that Unilever's bargaining power will increase post-transaction allowing it to get better access to shelf-space, promotional 191slots and higher margins.

(304) In conclusion, the surplus split analysis is not sufficient to support the claim that buyer power would exist post-merger.

General conclusion on countervailing buyer power

(305) Having assessed all the elements it is concluded that there is no indication that retailers would be in a position to resist a general price increase of Unilever post-merger.

IV.2.3.1.4. Sufficient entry unlikely to occur

(306) For entry to be considered a sufficient competitive constraint on the Parties, it has to be demonstrated that entry is likely, timely and sufficient to deter or defeat any 192 potential anti-competitive effects of the merger.In the this case, the market investigation revealed that entry into the deodorant markets is not only time consuming, but even if entry took place in the past it usually happened - with the exception of L'Oreal's entry with Garnier Mineral - on a small scale, insufficient to counteract the adverse effects on competition that will arise by the removal of a sizeable competitor like Sara Lee with its brand Sanex.

(307) Entry into the deodorant market can generally occur in three ways: entry of a new player not yet active in deodorants or a neighbouring personal care market; entry from a neighbouring personal care with an existing brand into deodorants or expansion of an existing male/non-male deodorant brand into the other gender deodorant market. Most of the current deodorant suppliers have entered deodorants from a neighbouring 193 personal care marketand indeed such a type of entry is viewed by Unilever "as the 194most likely source of new entry".

the inside option of the retailer compared to the inside option of the supplier and therefore increases the relative bargaining power of the supplier".

191 See reply to Questionnaire to customers on deodorants sent on 25 June 2010, question 7 and reply to Questionnaire to competitors on deodorants sent on 25 June 2010, question 6.

192 See Horizontal Merger Guidelines, paragraph 68.

193 See Form CO, paragraph 6.82: "Dove (from soaps), Vaseline (from petroleum jelly), Radox (from bath & shower products), Nivea (from moisturising creams), Lactovit (from body milks), Palmolive (from soaps), Narta (from fragrances), Gillette (from shavings), Garnier (from skin and hair care) and Sanex (from shower gel) are all examples of this."

194 See the Reply to the Statement of Objections, paragraph 7.2.

61

(346) As illustrated in Table 11, for Austria, the Czech Republic, Hungary, Italy, Latvia, Lithuania, Slovakia and Sweden, the overlap is either insignificant (below [0-5]*%) or the combined market share of the Parties as well as the concentration levels are within the ranges identified in the Horizontal Merger Guidelines for which it is unlikely that 234competition concerns would arise.

Table 11: Market shares for selective Member States – male, non-male and overall deodorants – 2009 – Source: Form CO

229 See document "M.5658 Unilever / Sara Lee – L'Oreal subsidiaries' comments on the SO", 14 September 2010, page 5 and 9.

230 See document "M.5658 Unilever / Sara Lee – L'Oreal subsidiaries' comments on the SO", 14 September 2010, page 11.

231 See reply to Questionnaire to competitors on deodorants sent on 24 April 2010, question 45.

232 [...]*

233 [...]*

234 See paragraphs 18-20 of the Horizontal Merger Guidelines.

69

Country Market Unilever Sara Lee Combined Post-merger Herfindahl- Hirschmann Index (HHI)

Change in HHI

Overall [20-30]*% [0-5]*% [20-30]*% [1000-1500]* [70-80]*

Czech Republic

Male [10-20]*% [0-5]*% [10-20]*% [1500-2000]* [50-60]*

Non-male [20-30]*% [0-5]*% [20-30]*% [1000-1500]* [80-90]*

Overall [10-20]*% [0-5]*% [10-20]*% [1500-2000]* [30-40]*

Italy

Male [10-20]*% [0-5]*% [20-30]*% [2500-3000]* [40-50]*

Non-male [10-20]*% [0-5]*% [10-20]*% [1500-2000]* [20-30]*

Overall [20-30]*% [0-5]*% [20-30]*% [1500-2000]* [0-10]*

Latvia

Male [10-20]*% [0-5]*% [10-20]*% [1000-1500]* [0-10]*

Non-male [20-30]*% [0-5]*% [20-30]*% [1500-2000]* [10-20]*

Overall [10-20]*% [0-5]*% [10-20]*% [1000-1500]* [70-80]*

Lithuania

Male [5-10]*% [0-5]*% [10-20]*% -- --

Non-male [20-30]*% [0-5]*% [20-30]*% [1500-2000]* [120-130]*

Overall [20-30]*% [0-5]*% [30-40]*% [1500-2000]* [50-60]*

Hungary

Male [20-30]*% [0-5]*% [20-30]*% [1500-2000]* [10-20]*

Non-male [30-40]*% [0-5]*% [30-40]*% [1500-2000]* [90-100]*

[0-10]* Overall [30-40]*% [0-5]*% [30-40]*% [2000-2500]*

Austria

Male [40-50]*% < [0-[40-50]*% [2500-3000]* [0-10]* 5]*%

Non-male [20-30]*% [0-5]*% [20-30]*% [2000-2500]* [0-10]*

[1500-2000]* [20-30]* Overall [20-30]*% [0-5]*% [20-30]*%

Slovakia

Male [10-20]*% [0-5]*% [20-30]*% [1500-2000]* [20-30]*

Non-male [20-30]*% [0-5]*% [20-30]*% [1500-2000]* [20-30]*

[3500-4000]* [30-40]* Overall [50-60]*% [0-5]*% [50-60]*%

Sweden

Male [50-60]*% [0-5]*% [50-60]*% [4000-4500]* [0-10]*

Non-male [40-50]*% [0-5]*% [40-50]*% [3000-3500]* [60-70]*

70

(347) For Bulgaria and Malta, market shares are only available for the overall deodorant market and not for the potential male and non-male deodorant markets. In Bulgaria, the combined market share of the Parties post-transaction would be [20-30]*% with an increment coming from Sara Lee of [0-5]*%, while in Malta Sara Lee achieved a market share of [5-10]*% and Unilever had [10-20]*% resulting in a combined share 235of [20-30]*%.

(348) In addition, in all these Member States sizeable competitors are present. Therefore, it can be concluded that the concentration is not likely to lead to a significant impediment of effective competition as regards the deodorant markets (overall, male or non-male) in Austria, Bulgaria, Czech Republic, Hungary, Italy, Latvia, Lithuania, Malta, Slovakia and Sweden. The remaining Member States are analysed in more detail in sections.

IV.2.3.2.1. Belgium

IV.2.3.2.1.1. Non-male deodorant market

(349) A number of factors indicative of significant non-coordinated effects are present in the non-male deodorant market in Belgium.

Merging firms have high market shares

(350) According to the Parties, the non-male deodorant market in Belgium had a total value of EUR 35.2 million 2009, [90-100]*% of which is covered by supplier brands and the remaining [5-10]*% by private labels. The non-male market in Belgium has experienced an increase in value of [5-10]*% between 2008 and 2009.

(351) Unilever is active in the non-male deodorant market in Belgium with its brands Rexona and Dove, while Sara Lee serves the Belgian non-male deodorant market with its brand Sanex.

Table 12: Market shares in the non-male segment for deodorants, in value, 2009, Belgium – Source: Form CO.

2009

000 € %

UNILEVER [...]* 30-40

- DOVE [...]* 10-20

- REXONA [...]* 20-30

SARA LEE [...]* 10-20

- SANEX [...]* 10-20

COMBINED [...]* 50-60

BEIERSDORF [...]* 10-20

COLGATE [...]* 0-5

COTY [...]* 0-5

HENKEL [...]* 10-20

L'OREAL [...]* 0-5

235 See Form CO, page 66.

71

PRIVATE LABEL [...]* 5-10

OTHERS [...]* 0-5

TOTAL MARKET 35 207 100.0

(352) The Parties achieve a combined market share of [50-60]*% which has increased by roughly [0-5] percentage points between 2008 and 2009. It exceeds by nearly three times the share of its nearest competitors Henkel with its brand Fa ([10-20]*%) and Beiersdorf with its brand Nivea ([10-20]*%). Other competitors with a more limited share of the market are private labels ([5-10]*%) and L'Oreal with its brands Narta and Ushuaia ([0-5]*%)

(353) The most successful supplier in the market is Sara Lee with its brand Sanex, whose sales increased by [10-20]*% (compared to the overall market growth of [5-10]*%). 236237[reference to parties’ internal documents] , .

(354) Unilever grew by [10-20]*%, mainly resulting from the Dove brand, which shares a similar skin caring positioning.

(355) The competitors Beiersdorf (+[10-20]*%) and Henkel (+[10-20]*%) have experienced similar increases of their sales as Unilever. By contrast, L'Oreal sales (Narta and Ushuaïa) have decreased by [20-30]*% in 2009. The impact of the introduction of Garnier Mineral in January 2010 in Belgium is discussed separately in recitals 441-455.

(356) Such high market shares suggest that the merged entity is likely to enjoy significant market power in the absence of any mitigating factors. Furthermore, the overlap between the Parties is significant with Unilever holding a [30-40]*% market share pre-merger (with [20-30]*% for Rexona, the leading non-male brand, and [10-20]*% for Dove) and Sara Lee a [10-20]*% share. This is indicative of strong non-coordinated effects.

(357) In the market investigation, several respondents expected that the merger would lead 238to anticompetitive effects for deodorants products in Belgium.

(358) In its reply to the Statement of Objections, Unilever submitted that several respondents to the market investigation did not specifically predict a price increase for 239deodorants, although they identified potential anticompetitive effects. In particular, Unilever pointed out to L'Oreal's response which did not, according to Unilever, have any clear view as to whether price increases are likely post-merger.

(359) In fact, L'Oreal clearly submitted that a price increase is a likely outcome due to the 240significant change in the structure of the Belgian non-male deodorant market. As

236 [reference to parties’ internal documents]

237 [reference to parties’ internal documents]

238 See replies to Questionnaire to customers on deodorants sent on 23 April 2010, questions 63 to 65.

239 Reply to the Statement of Objections, paragraph 8-3.

240 Questionnaire to customers on deodorants sent on 25 June 2010 ,reply to Question 13 provided by L'Oreal.

72

pointed out by L'Oreal, "(A price increase) is more likely to happen if a supplier has an undisputable leadership and can convince the retailers that he can alone significantly change the price perception of a category through a general price increase of its own brand", which is, as reiterated by L'Oreal throughout the questionnaire, precisely the situation of Unilever post-transaction. Consequently the takeover of Sanex by Unilever "could possibly lead to less promotional pressure and as a consequence high prices for the consumers". Likewise, a Belgian retailer submitted that the merger would lead to "less players=> less promotion=> less 241advantage for the customer"

Merging firms are close competitors

(360) An important aspect for assessing unilateral effects arising from the proposed merger is the degree of substitutability between the Parties' non-male deodorants. The higher the degree of substitutability between their products, the more likely it is that the Parties will be able to significantly raise prices to retailers for non-male deodorants.

(361) All retailers active in Belgium explained that the Parties are close competitors. In particular, with respect to Dove and Sanex, those retailers explained that they share 242 comparable attributes and are therefore directly competing in the market.A majority of retailers considered that Dove, Sanex and Nivea are similarly positioned as 243umbrella skin-caring brands. One retailer takes the view that Sanex and Rexona are 244close competitors because both brands have an efficacy claim for sports activities . By eliminating a close competitor, the merger would eliminate the primary source of competitive pressure that prevails in the non-male deodorant market.

(362) With respect to brands of the main competitors, a distinction should be made between Nivea (Beiersdorf) and Fa (Henkel). Nivea deodorants are sold with a skin-caring proposition which is broadly comparable with the attributes of Dove and Sanex. Its price positioning (mid-range to premium) is also relatively akin to Dove and Sanex.

(363) Fa, on the other hand, is a brand positioned at the entry level to mid-range which is marketed under a fragrance based/feminine proposition and it does not share the skin-caring proposition of Dove and Sanex.

(364) One retailer explained in that regard that "In the current situation (pre-transaction), Sanex products are directly competing with Dove and the existence of Sanex keeps a certain balance. With this competitor disappearing, there are not much alternatives for consumers of deodorants for sensitive skin. The closest competitor that will be left 245is Nivea, which still is perceived differently by the consumer". Likewise, another

241 See reply from [retailer] to Questionnaire to customers on deodorants sent on 25 June 2010, question 5 b).

242 Questionnaire to customers on deodorants sent on 23 April 2010, questions 27 and 30.

243 See replies by [retailer] and [retailer] to Questionnaire to customers on deodorants sent on 23 April 2010, questions 27 and 30.

244 See replies by [retailer] to Questionnaire to customers on deodorants sent on 23 April 2010, questions 27 and 30.

245 See reply from [retailer] to Second phase Questionnaire to customers on deodorants sent on 25 June 2010, question 5 c).

73

Belgian retailer, referring to Dove and Sanex put forward that in the women segment only two suppliers have enough weight in terms of volume and range.

246.246(365) In its reply to the Statement of Objections, Unilever acknowledged that Dove and Sanex are competing brands since they share a skin-friendly platform. Unilever nevertheless considered that the closeness of competition between both brands is moderated by the higher price positioning of Dove compared to Sanex. Unilever also claimed that the Statement of Objections failed to take into account the constraints 247imposed by competing brands such as Nivea, Fa and Narta. According to Unilever, these brands have comparable weight and depth to Sanex. They argue that Nivea shares similar attributes and positioning with Sanex and Fa has a comparable price positioning.

(366) With respect to the price positioning of Dove compared to Sanex, Unilever itself downplays in the Form CO the relevance of the pricing analysis in Belgium. Unilever considers that these pricing results, based on Nielsen data, need to be treated with some caution given the frequent use of discount coupons, which are not captured in 248the Nielsen pricing data. Approximately 9 out of 10 promotions for deodorants in Belgium are driven by coupons and the average discount on a coupon is roughly 10-20%.

(367) Moreover, the higher price positioning of Dove compared to Sanex is mainly the result of recent price increases of Dove whereas Sanex's prices have not followed the same trend. In the skin segment, Figure 7 and Figure 8 show that Dove (as well as Rexona) increased its prices per litre quite substantially in 2008 and 2009 whilst Sanex remained constant (in sprays) or increased moderately (in roll-ons). This evolution 249 [reference to parties’ internal documents]and is acknowledged by Unilever in the 250reply to the Statement of Objections. That widening of the price gap does not show that Dove and Sanex are distant competitors but rather that Sanex has pursued an independent pricing strategy in Belgium, which will be abandoned post-transaction.

Figure 7: Evolution of price per litre of non-male roll-on deodorants in Belgium between April 2006 and March 2009

[...]*

Figure 8: Evolution of price per litre of non-male spray deodorants in Belgium between April 2006 and March 2009

246 Reply to the Statement of Objections, paragraphs 8-4 to 8-6.

247 As Narta has a [0-5]*% market share in the non-male deodorant market in Belgium (reply to the statement objections, paragraph 8-6), the competitive constraint that this brand exerts on the merging parties appears to be particularly limited

248 Form CO, paragraph 6-183

249 [reference to parties’ internal documents]

250 In footnote 68 of the Reply to the Statement of Objections, where Unilever states that Dove "has seen an increase in average prices since 2007".

74

[...]*

(368) It is not contested that Nivea is in a cluster of close competitors together with other brands with skin caring proposition - Dove and Sanex. It is not argued that Sanex is Dove's closest competitor but rather that Dove and Sanex are close competitors and that therefore a potential price increase is more likely to be sustainable if those two brands merge. However, as mentioned in recital (364), one retailer explains that Nivea is perceived differently by consumers than Dove and Sanex.

(369) Unilever did not substantiate why Fa should be considered as a close competitor to Sanex or Dove in its reply to the Statement of Objections, except in relation to price. The available evidence in the market investigation suggests the opposite. None of the customers that responded to the market investigation identified Fa as a close competitor of either Dove or Sanex. One respondent considered that Fa is a close competitor of 251 Rexona due to their particular focus on efficacyand another one submitted that Fa's 252strength lies in fragrances whilst Dove is positioned in skin care. Unilever describes Fa as "having a large number of trendy and inspiring fragrances with colourful and 253feminine packaging and a tagline of "feel-good freshness", a positioning which is significantly different from skin caring propositions carried by Dove or Sanex.

(370) This closeness of competition between Dove and Sanex is also reflected in an interaction 254analysis submitted with the Form CO, conducted by GFK/Europanel, which shows a high interaction between Dove and Sanex, although the interaction with Rexona is less strong. The interaction index is also high between Sanex and Nivea, suggesting that Dove, Sanex and Nivea are close competitors.

255.255(371) On 26 July 2010, Unilever provided an updated analysis of this interaction data . This analysis focuses on the competitive dynamics between Rexona, Dove and Sanex.

(372) Europanel calculates these interaction indices using the data on households/individual purchases from the consumer panels that it operates in each country. The Europanel interaction indices are based on the actual purchasing behaviour of a large sample of purchasers in each country and hence give insights, according to Unilever, into purchasers' actual behaviour in each country.

(373) Europanel calculates an interaction index for each pair of brands covered by this analysis. The index is the ratio of actual switching between the two brands to the expected level of switching if buyers switched in proportion to market shares. Therefore,

251 See reply by [retailer] to Questionnaire to customers on deodorants sent on 23 April 2010, question 30.

252 See reply by [retailer] to Questionnaire to customers on deodorants sent on 23 April 2010, question 30.

253 Form CO, paragraph 6-63

254 See Figure 6.7 Form CO.

255 "M.5658 Unilever/Sara Lee Body Care: Supplementary submission regarding interaction indices" submitted on 26 July 2010.

75

if the index is higher than 100, then the actual switching is greater than expected, which indicated that these brands are close substitutes. Conversely, if the index is less than 100, then it would indicate that these brands are not close substitutes.

(374) After having examined the data, it is considered that these results should be interpreted cautiously. Since an individual customer may purchase deodorants also for other family members, the switching patterns observed cannot be directly interpreted as a sign of actual substitutability. A female customer who is tracked as switching from, say Sanex for women to Sanex for Men, or another male brand may in reality have simply been purchasing a deodorant for her son or her partner without having changed her actual usage pattern.

(375) Although contaminated by these factors, the switching data may however contain information about actual substitutability and hence closeness of competition. Each brand may in a given period lose customers to (all) other brands and simultaneously win customers from all other brands. The sum of lost and won customers between two brands, the total gross switching is thus an indicator how closely the two brands are competing. The interaction indices presented in Unilever's submission of 26 July 2010 are reproduced in Figure 9.

Figure 9: Interaction index for non-male Sanex in Belgium

[...]*

(376) The interaction indices presented by Unilever are a comparison of the actual gross consumer switching of the brand (here, Sanex for Women) with the "expected" switching based on the proportions of the competing brands' market shares in the total 256 deodorant category (combining male and non-male).A high interaction of Sanex with a given brand is not in itself indicative of significant switching figures between Sanex and this brand because this high index can be brought about by the small share of the brand. In this example, Adidas (Coty) has a high interaction index with Sanex for Women ([150-170]), but it market share is quite small. When the market share is taken into account, the actual consumer switching as observed in the panels is significantly lower than the interaction index alone would suggest.

(377) Therefore, while interaction indices provide some insights about the closeness of brands, the gross switching information provides a more illustrative picture about the actual switching patterns of consumers.

(378) Upon request, the Parties submitted the underlying data used by GFK/Europanel containing the actual switching data. Based on the same data as used for the depicted interaction index, Figure 10 shows, in descending order for each brand, their share of total gross switching to and from Sanex for Women in 2009. Consistent with the method used by GFK/Europanel, the total gross switching is compared with the proportions of the competing brands' market shares in the total deodorant category ("value of competitors").

256 The market shares are calculated on the basis of the brands' sales within the panel. This may to some extent vary from the market shares based on Nielsen data, but in general provides a rather consistent picture with those.

76

Figure 10: Gross switching for non-male Sanex in Belgium

[...]*

(379) Figure 10 shows more consistent results for the interaction with Sanex for Women. While Nivea for Women ([10-20]*%) has the highest level of gross switching with Sanex for Women, Dove for Women ([10-20]*%) Rexona for Women ([10-20]*%) and Fa for Women ([10-20]*%) also interact significantly with Sanex.

257.257(380) In its reply to the Statement of Objections, Unilever considered that these switching or interaction data suggest that Nivea is the closest competitors to Sanex, with interaction noticeably higher than Dove. The share of gross switching to and from Sanex accounted by Nivea ([10-20]*%) is deemed by Unilever as "disproportionately high" compared to its value share and would therefore show that Nivea is the closest competitor to Sanex.

(381) These data indicate that Dove and Nivea are close competitors to Sanex whilst Fa and Rexona are more distant. Due to the complexity of the assessment, it is not appropriate in this particular case to label one particular brand as "the closest" competitor for the purpose of drawing conclusions about the competitive constraints exercised by some individual brands. The switching data confirms in general the results of the market investigation: Dove, Sanex and Nivea are close competitors in the Belgian non-male market.

(382) The gross switching data show that of all the Sanex users that switched to or from a 258 non-Sanex female brand, [40-50]*% switched to or from a Unilever brand.These figures confirm that the risk of losing sales to Unilever appear to be an important constraint on Sanex prior to the merger. Such switches would not constrain Sanex after the transaction since the switches would no longer be considered lost to the merged entity.

259(383) Unilever arguesthat the majority ([50-60]*%) of gross switching by Sanex purchasers is to or from non-Unilever brands, including the brand that interacts the most with Sanex (Nivea) and two of the top four closest brands to Sanex (Fa and Nivea). This does not, however, mean that Unilever brands are not constraining Sanex. The indication from the consumer panel data that more than [40-50]*% of all Sanex switching within the non-male market in Belgium occurs with Unilever's brands means that Unilever is indeed a very significant constraint on Sanex. Even if the other half of switching is to or from other brands, the merger instantly removes nearly half of the total competitive constraint previously exercised on Sanex. This increases the possibility for a price increase of Sanex post-merger.

257 Reply to the Statement of Objections, paragraph 8-6 (iv) a).

258 This figure is calculated taking into accounts only non-male brands. When certain brands were not clearly defined as male and non-male (for example private label, or the category of 'others'), the non-male value was calculated according to the proportions of male and non-male turnovers which this particular supplier(s) or brands achieve in the respective country (turnover split taken from market share tables provided by Parties – Annex 7.1 of the Form CO).

259 Reply to the Statement of Objections, paragraph 8-10.

77

(384) These switching data therefore confirm the results of the market investigation and indicate that the most direct competitors to Sanex for Women are Nivea and Dove with Fa and Rexona being more distant. However, the relatively high level of gross switching with Axe Bodyspray and Rexona for Men indicates that the data must be interpreted with caution.

Assessment on the basis of format and product attributes

(385) More detailed data about the SKUs sold at brand level in Belgium have also been analyzed. These data provide information by product attributes, in particular format (crème, roll-on, spray, wipes, vapo) but also other characteristics (skin friendly, fragrance, anti-perspirant or AP, efficacy, nowhite-marks, girl or others). The sub-260segmentation is based on a classification provided by Unilever .

(386) This refinement can in general help to identify in more detail in which sub-segments the Parties brands are positioned and show where a significant overlap exists.

(387) Table 13 shows the characteristics of the products in the non-male market in general. A similar analysis is also presented on the basis of formats. Finally, an analysis which combines price points and product characteristics is presented.

Table 13: Belgium: Deodorants: 2009: sub-segment shares value – non-male.

Non-MaleSkin Care Fragrance Antiper-Efficacy No WhiteGirl Without 2009 [...]* spirant Marks Characte- ristic

Rexona [5-10]*% [20-30]*% [60-70]*% [30-40]*% [20-30]*% [40-50]*%

Dove [30-40]*% [20-30]*% [20-30]*% [20-30]*%

Sanex [20-30]*% [20-30]*% [0-5]*% [20-30]*% [30-40]*%

Combined [60-70]*% [60-70]*% [60-70]*% [90-100]*%

Nivea [30-40]*% [20-30]*% [10-20]*%

Ushuaia [5-10]*%

Narta [5-10]*% [5-10]*%

Fa [80-90]*% [5-10]*% [70-80]*%

Others [0-5]*% [5-10]*% [5-10]*% [5-10]*% [0-5]*% [0-5]*%

Share of [40-50]* [20-30]* [60-70]* [20-30]* [10-20]* [5-10]* [10-20]*% non-male 261segment

260 This classification is also used for calculation of the predicted price increases in the merger simulation

261 The sum of percentage of each segment is superior to 100% because one given SKU can have several characteristics (AP, skin care, no white marks, etc).

78

Turnover of13.9 7.0 20.5 7.7 3.7 3.0 4.9 segment in million EUR

(388) As can be seen from Table 13, Sanex is only present in three sub-segments (with a very small presence in efficacy) skin, AP and "no white marks"]. Firstly, Dove appears to be a close competitor in that it is present in exactly the same three sub-segments as Sanex, and only in these three sub-segments. Also, three of the four brands of the other competitors are absent from these three sub-segments, with the exception of Narta which has a small presence in the no white marks sub-segment. While Nivea is present in the skin care and AP sub-segments, Fa has a particular strength in the fragrance and girl sub-segments (where the Parties' brands are hardly present, except Rexona in the girl sub-segment) and Ushuaia can only be found in the fragrance sub-262segment.

(389) To complement the analysis of closeness of competition, the non-male market has been segmented by format. This assessment focused on two specific formats, namely sprays and roll-ons which account for 73% and 23% of all non-male deodorants sold in Belgium, respectively.

(390) Further segmenting the non-male market according to format does not alter the analysis regarding the overall non-male market. Market presence and market shares in sprays and roll-ons are broadly comparable to the overall market shares although Nivea and Sanex appear to be stronger in roll-ons than in sprays.

Table 14: Belgium: Deodorants: 2009: sub-segment shares value – non-male sprays.

Non-MaleSkin Care Fragrance Anti-Efficacy No WhiteGirl Without spray 2009 Perspirant Marks Characte- [...]* ristics

[5-10]*% [20-30]*% [70-80]*% [30-40]*% [20-30]*% [50-60]*%Rexona

[30-40]*% [20-30]*% [20-30]*% [20-30]*%Dove

[10-20]*% [10-20]*% [0-5]*% [20-30]*% [20-30]*%Sanex

[60-70]*% [60-70]*% [70-80]*% [90-100]*% [20-Combined 30]*%

[90-100]*%

[30-40]*% [20-30]*%Nivea

[10-20]*%Ushuaia

[80-90]*% [70-80]*%Fa

[5-10]*% [0-5]*%Narta

262 Nivea has a small presence in the no-white mark segment, however, it is well below [0-5]*%.

79

[0-5]*% [5-10]*% [5-10]*% [10-20]*% [0-5]*% [0-5]*%Others

[30-40]*% [20-30]*% [50-60]*% [20-30]*% [10-20]*% [10-20]*% [10-20]*%

Share of non-male spray 263segment

Turnover of segment in million EUR

9.1 5.8 13.9 5.5 2.5 2.6 4.2

Table 15: Belgium: Deodorants: 2009: sub-segment shares value – non-male Roll-on.

Non-MaleSkinFra- Anti-Efficacy No Girl Without Roll-on 2009CaregrancePerspirant WhiteCharacteristics [...]* Marks

[0-5]*% [10-20]*% [40-50]*% [30-40]*% [20-30]*%Rexona

[10-20]*% [5-10]*% [20-30]*% [10-20]*%Dove

[30-40]*% [30-40]*% [30-40]*% [80-90]*%Sanex

[40-50]*% [50-60]*% [40-50]*% [90-100]*% [20-30]*% [90-100]*%Combined

[50-60]*% [30-40]*% [30-40]*%Nivea

Ushuaïa

[90-100]*% [10-20]*% [70-80]*Fa

[5-10]*% [0-5]*Narta

[0-5]*% [0-5]*% [0-5]*% [0-5]*Others

[50-60]*% [10-20]*% [70-80]*% [20-30]*% [10-20]*% [5-10]*% [5-10]*%Share of non- male roll-on 264segment

4.3 1.3 6.0 2.0 1.2 4.0 0.7Turnover of segment in million EUR

(391) Finally, these product characteristics have been combined with average prices for the last three years. Figure 11 and Figure 12 show the price positioning of the main brands

263 The sum of percentage of each segment is superior to 100% because one given SKU can have several characteristics (AP, skin care, no white marks, etc).

264 The sum of percentage of each segment is superior to 100% because one given SKU can have several characteristics (AP, skin care, no white marks, etc).

80

in (i) the skin care sub-segment and (ii) the no white-marks sub-segment for spray and roll-ons respectively.

Figure 11: Average price points for Belgian non-male deodorants, skin care segment (spray deodorants) for 2007, 2008 and 2009

[...]*

Figure 12: Average price point for Belgian non-male deodorants, nowhite marks (spray deodorants) in 2007, 2008 and 2009.

[...]*

(392) Combining the product characteristics with average prices indicates that Sanex has not been able to achieve similar prices as the other brands in the spray skin sub-segment during the last three years. While Rexona, Dove and Nivea increased their average prices by [...]* or more, Sanex increased much less. In the no white marks sub-segment it seems that Nivea and Dove have repositioned prices upwards to the high end of the market with Sanex remaining at a price point at least [...]* below the leading brands.

(393) In the roll-on market, Sanex and Rexona's price positioning were broadly comparable until 2008 but as already noticed in the spray sub-segment, Sanex (or Nivea in this case) has not been able to achieve similar price increases as Dove and Rexona in the skin care sub-segment over the last three years. This price increase has not impacted Rexona and Dove's sales which have increased by [...]* in value, in line with market average. Sanex is now positioned at the low end of the market.

Figure 13: Average price point for Belgian non-male deodorants, skin care segment – (roll-on deodorants) in 2007, 2008 and 2009.

[...]*

(394) In the no white marks sub-segment, Sanex, Dove and Rexona have a relatively similar price positioning whereas Narta is [...]* more expensive as shown in Figure 14.

Figure 14: Average price point of Belgian non-male deodorants, no white marks segment (roll-on) in 2007, 2008 and 2009.

[...]*

(395) The analysis of the sub-segments basically confirms the qualitative evidence collected during the first phase market investigation, namely that Sanex appeals to non-male customers sensitive to soft skin and no white marks attributes. Dove is the closest competing brand to Sanex, Nivea is also close for skin care attributes.

Demand estimation and merger simulation

81

(429) Dove, Rexona, Sanex, Nivea, and Fa are the brands with the highest weighted distribution. Post-transaction, Unilever will control three out of these five brands with the highest weighted distribution rate in the non-male market. Moreover, Unilever controls Axe with a very strong position on the deodorants for male market, accounting for [40-50]*% of the sales. The transaction would therefore reinforce Unilever's position as unavoidable trading partner for the retailers.

88

292(430) Unilever considers that the notion of "unavoidable trading partner" is ill-founded since, according to Unilever, "must-have" brands are an ill-defined concept in the context of relative bargaining power between manufacturers and retailers. Unilever puts forward that there are some brands that retailers are keen to have on shelf due to their share of sales but similarly manufacturers are keen to have a trading partner in each of the retailers, otherwise those volumes are likely to be switched to alternative brands (rather than alternative retailers).

(431) Unilever's arguments cannot be accepted. It is not contested that retailers need some brands as much as manufacturers need an access to the market. However, the balance of power between retailers and manufacturers depends upon the market share of the brand. Unilever controls more than half of the total market and possesses the most successful brands with high notoriety and customer base in Belgium. As retailers indicate, end-consumers expect to find these strong and popular brands on the shelves, and retailers cannot ignore this. The combination of powerful brands, which are considered by retailers as necessary, in the hands of a single supplier is a major shift in the structure of the Belgian market.

293(432) Unilever also argues that the available responses to the market investigation seem to indicate that retailers in Belgium have bargaining power. Unilever points to the responses of one retailer submitting that there are no must-have brands as well as responses indicating that retailers would not accept price increases from the merged entity.

(433) However, the investigation does not support Unilever’s arguments. The largest Belgian retailer clearly submitted that a substantial amount of customers would switch to other retailers if they were not stock one of their brands, and this would have a negative commercial impact on them. The retailer quoted by Unilever considers that three options were possible (major switches to other retailers if some brands are absent, switch of deodorant purchases to other retailers, switch to other deodorants in the same store) depending on the customers but notes that “Customer is very attached to the brand” .

(434) Based on its experience, [...]* also explained that a substantial amount of customers would switch to other retailers if we were not to stock one or more of Unilever's brands. The [...]* case provides an indication that customers would switch retailers if several Unilever brands were not available at the same time. According to a survey [...]*, roughly 30% of [...]* customers indicated that they switched supermarkets to get their preferred brands.

292Reply to the Statement of Objections, paragraph 8-29.

293Reply to the Statement of Objections, paragraph 8-27.

294See reply from [retailer] to second phase Questionnaire to customers on deodorants sent on 25 June 2010, question 17.

295See reply from [retailer] to Questionnaire to customers on deodorants sent on 25 June 2010, question 17.

296See Recital (275) for further details.

89

(435) "Non-regular" delistings (which are not linked to removal of underperforming products) are exceptional given that retailers fear that such behaviour would have negative impact on their turnover and profitability in the light of the significance of the top brands in their assortment. Unilever cited the example of [...]*, [...]*. However, as explained in detail in recital (276), such delistings took place in the context of extremely tough negotiations between Unilever and [...]*, which resulted in the conclusion of a contract whereby [...]* might be forced to continue listing Unilever SKUs [...]*.

(436) The combination of the Parties' strong brands in the hands of a single supplier would shift bargaining power (if any) from the retailer to the supplier, by severely reducing the set of credible alternatives that the retailer can choose from to replace Unilever's products if it so chooses.?. In the post-merger market structure, Unilever would thus become an unavoidable trading partner in deodorants in Belgium in relation to price setting, promotions, new listings and placement on shelves.

(437) Furthermore, it is not sufficient that buyer power exists prior to the merger, it must also exist and remain effective following the merger. This is because a merger of two suppliers may reduce buyer power if it thereby removes a credible alternative. In this case, the merger will remove one of the two main suppliers considered to be close competitors in the market, since the market investigation did not reveal that there is a set of credible alternative suppliers which could replace Sanex in terms of product characteristics and innovation capabilities. It will also bring together three of the five main brands in the deodorant market, which has a direct negative impact on the ability and incentives of retailers to use their alleged countervailing buyer power.

(438) Therefore, it is concluded that countervailing buyer power post-merger cannot be considered as a mitigating factor which is likely to off-set any of the potential adverse effects of the proposed transaction.

Sufficient entry unlikely to occur

(439) The market investigation has revealed that entering the deodorant market – either from a neighbouring personal care market or as a new entrant is difficult. Even if the technology required and the investment to manufacture deodorants could be considered as "accessible", high barriers to entry exist in the form of significant advertisement and promotion (A&P) expenditure to create brand awareness and get access to shelf space from retailers.

(440) In addition to the barriers to entry from the creation of brand awareness and access to shelf space, a potential entrant would have to take into account the likely reaction of a powerful incumbent like Unilever. In Belgium, Unilever's leading position with market shares above 50% indicates that Unilever has a strong incentive to prevent entry of new brands or expansion of existing ones as it would be through its leading position the one suffering most from such activities.

(441) [...]*, Unilever's increased market share with the addition of another brand like Sanex strengthens its incentive to fight entry.

(442) Also, the combination of an unrivalled number of main brands in the hands of a single supplier increases the possibility for Unilever to adjust its products accordingly, preventing the competitor's product from having a unique product proposition. With one additional brand, Unilever's ability to slightly reposition its products to squeeze

90

the new entrant will also be improved. Consequently, the addition of a new brand to Unilever's already broad portfolio could increase barriers to entry.

(443) Apart from Garnier Mineral, the elements gathered during the market investigation showed that there has been no significant entry in the Belgian non-male deodorant market since L'Oreal's entry in 2007 with Narta. L'Oreal's sales have been decline recently ([20-30]*% between 2008 and 2009) and Narta’s market share is now only [0-5]*%. Hence, this entry was not very successful.

(444) Despite the fact that the non-male deodorant market is growing, the market investigation revealed that retailers carefully select new products to be listed. Retailers in Belgium do not appear to respond to what the notifying party describes as 297 consumers' continuing desire to be "surprised and delighted"by an innovative set of new deodorants.

(445) Belgian retailers explained this very clearly. One retailer submitted that it is not possible to list the entire assortment of a new entrant, thereby diminishing its chances of success. “Due to shelf space limitations, we are not able to follow the entire assortment of all suppliers. In case of innovations, we first evaluate the potential of the considered product/range – cf whether it will make the category grow, rather than cannibalizing other products turnover and whether it will increase penetration in the 298category”. The leading retailer also confirmed that it is unable and unwilling to list the entire product range of a new supplier in its stores. Another retailer cites the example of Garnier Mineral which requested the listing of 10 SKUs; this retailer 299accepted to list only a few of them. In the case of Dove for Men, retailers indicated 300[...]* .

(446) One competitor explains that "Generally retailers do not favour new suppliers to stimulate competition. Space is limited, so any new supplier has to take the place of another product. Listing of new products only take place if the chances of adding value to the overall category value are very high or if huge investments are granted to 301 the retailers to buy distribution."A second competitor stated that new suppliers can be favoured in only two instances: in case of a niche product; or if the supplier 302demonstrates that the retailer will significantly increase its margin.

(447) Indeed, apart from re-launches of existing products and brand extensions (such as adding new variants to a brand), the only significant new entry in the Belgian non-male deodorant market in recent years was Garnier Mineral in January 2010. Unilever,

297Form CO, paragraph 6-196.

298See reply by [retailer] to questionnaire to customers on deodorants sent on 23 April 2010, question 56.

299See reply by [retailer] to questionnaire to customers on deodorants sent on 23 April 2010, question 57.

300See replies to questionnaire to customers on deodorants sent on 25 June 2010, question 8.

301See reply from Beiersdorf to Questionnaire to competitors on deodorants sent on 25 June 2010, question 23.

302See reply from Henkel to questionnaire to competitors on deodorants sent on 25 June 2010, question 23.

91

in the Form CO emphasizes the Garnier entry as it is carried out by a major international supplier L'Oreal with a high level of investment supporting the launch. Indeed, Garnier represents the most significant entry of a new competitor in recent years.

(448) As it only took place in January 2010, it is too early to conclude on the outcome of Garnier’s entry in the long term. This is also the opinion of all the retailers which 303responded to the market investigation .

(449) Unilever reacted immediately to the Garnier Mineral entry by launching Rexona Mineral in Belgium in February 2010. [...]*

(450) Unilever submits, however, that Garnier Mineral provides an actual example of the ability of competitors to enter the non-male segment, which should not be dismissed despite its recent launch. The launch shows every sign of being ‘significant’, according to Unilever.

(451) Information has been available for 8 months since the product was launched in 304Belgium with 12 SKUs. Based on the most recent data, in the first six months, Garnier Mineral achieved wide distribution of [80-90]*%, accounting for a [0-5]*% share of shelf and a [0-5]*% share of the non-male segment. According to Unilever, 305 the brand has shown consistent growth in market share since its launch.Available monthly Nielsen data show significant fluctuations of Garnier Mineral' market share and do not allow to identify any noticeable trend.

(452) Unilever also mentions the recent entry of Gillette which reached a market share of [0-3065]*% in the male segment (week 28 2010) after its launch “in 2010”. The very different competitive landscapes between the male market and the non-male market is described and it is established that the factors favouring or hindering entry do not have the same impact in the male and non-male markets. It is therefore impossible to draw any conclusion for the non-male market regarding the introduction, successful or not, of a new brand in the male market. It is even more difficult to determine whether the launch of Gillette was a success, given it was only launched on 1 July 2010 and that a market share on week 28 reflects purchases made during the week of the launch or the week after, when customers demonstrate high curiosity for the new product.

(453) Consequently, it is unlikely that new entry will occur in the Belgian non-male market to the extent that it will counteract the adverse effects of the merger.

Overall conclusion

(454) It is therefore concluded that the notified concentration is likely to significantly impede effective competition on the market for non-male deodorants in Belgium.

303See replies to questionnaire to customers on deodorants sent on 23 April 2010, question 57.

304Garnier proposed 10 SKUs to be listed by [retailer].

305Reply to the Statement of Objections, paragraph 8-34.

306Reply to the Statement of Objections, paragraph 8-36.

92

IV.2.3.2.1.2. Male deodorant market

(455) According to the Parties, the Belgian male deodorant market had a total value of EUR 26.8 million in 2009. It experienced an increase in value of [5-10]*% between 2008 and 2009. Unilever is active in the male deodorant market mainly with its brand Axe, followed by Rexona for Men. In addition, Unilever has launched in 2009/2010 a new deodorants series Dove for Men. Sara Lee serves the Belgian male deodorant market with its brand Sanex and has only minor sales with its Williams brand.

Table 18: Belgium: Deodorants - 2009 market shares and value, male market – Source: Form CO.

Men

000 € %

UNILEVER [...]* 50-60

- AXE [...]* 40-50

- DOVE [...]* 0-5

- FABERGE [...]* 0-5

- REXONA [...]* 5-10

- ZZ OTHERS [...]* 0-5

SARA LEE [...]* 0-5

- SANEX [...]* 0-5

- WILLIAMS [...]* 0-5

- ZZ OTHERS [...]* 0-5

COMBINED [...]* 50-60

BEIERSDORF [...]* 20-30

COLGATE [...]* 0-5

COTY [...]* 5-10

HENKEL [...]* 0-5

L'OREAL [...]* 5-10

PRIVATE LABEL [...]* 0-5

OTHERS [...]* 0-5

TOTAL 26 809 100.0

(456) Post-merger, the Parties would achieve a combined market share of [50-60]*% (Unilever [50-60]*%, Sara Lee [0-5]*%), followed by Beiersdorf ([20-30]*%), Coty ([5-10]*%) and L'Oreal ([5-10]*%). In spite of this high market share, the proposed transaction will not result in a significant impediment of effective competition for several reasons.

(457) First, Sara Lee has a limited presence in the market. Its main brand Sanex for Men accounted for [0-5]*% market share and was declining compared to 2008, while the market was growing.

(458) Second, Sanex for Men is not perceived as a close competitor to Unilever's main brand Axe, which accounts for more than [80-90]*% of Unilever's presence in the male market. While Sanex for Men is perceived as a brand with skin care attributes, retailers

93

consider Axe to occupy a unique position within male deodorants because of its brand 307 recognition, orientation towards teenagers and young men as well as fragrance.This is also confirmed by average pricing figures showing that Axe is usually [...]* more 308expensive than Sanex.

(459) Third, in the male deodorant market several other competitors are active with a similar or significantly higher presence that Sara Lee. Beiersdorf with its Nivea for Men range achieved a market share of [20-30]*% and is in general perceived as a close competitor to Unilever's Rexona for Men brand. Coty, with its Adidas brand, has a similar position as Axe. In addition, L'Oreal and Henkel are well-known suppliers in the deodorant market positioned with their brands at similar price points as Sanex for Men.

(460) Based on these elements, it can therefore be concluded that the concentration is not likely to lead to a significant impediment of effective competition as regards the market for male deodorants in Belgium.

IV.2.3.2.2. Denmark

IV.2.3.2.2.1. Non-male deodorant market

Preliminary remarks on market definition for Denmark

(461) In the reply to the Statement of Objections, Unilever put forward that there is a single, albeit differentiated, market for all types of deodorants in Denmark. Alternatively, a distinction could be made between male, female and unisex deodorants. In the latter case, no competition concerns should, according to Unilever, arise as there would be no overlap between Unilever's "female" brands Dove and Rexona and the "unisex" brands of Sara Lee (Sanex and Neutral).

(462) The market definition for deodorants is discussed in Section IV.2.1.1. Male deodorants form a distinct market from non-male deodorants. In this section, specific features of the Danish market are discussed.

(463) First, it is acknowledged that in Denmark the Nielsen classification on which market share data is based distinguishes between three categories, "male", "female" and 309 "family". However, data classifications cannot determine market definitions.At the same time, the classification of GFK on which the switching analysis is based distinguishes between "for men" on the one hand and "for woman + Unisex" on the other hand.

(464) Responses in the market investigation show the difficulties in distinguishing between unisex and female brands. When asked to list the brands considered to be male, female and unisex respectively, one retailer simply deleted the reference to a "unisex" category

310and stated that deodorants "are placed in the male category or female category". Sanex was included in the "female" category, together with Rexona, Nivea, Palmolive and Dove. Another retailer indicated that Sanex is positioned within the "unisex" deodorants, but also within the "female" products, together with Rexona, Dove, Palmolive or Nivea. It explained that unisex deodorants are placed on the shelves with 311"female" deodorants.

(465) The largest Danish retailer refers to Sanex together with Rexona and Palmolive as 312 "unisex" products.However, a hypermarket owned by the same group explained that Sanex deodorants are placed in the shelves together with Rexona, Dove and Nivea in 313 the "damenduft"/fragrance for women shelf.

314(466) [reference to parties’ internal documents]

(467) It is concluded that a specific market definition for the Danish market which would distinguish between unisex and female deodorants is not supported by the market investigation.

Non-coordinated effects in the non-male deodorant market

(468) A number of factors indicative of significant non-coordinated effects are present in the non-male deodorant market in Denmark.

Market structure

(469) According to the Parties, the non-male deodorant market in Denmark had a total value of EUR 27.2 million in 2009. It is a relatively fragmented market with very limited presence of private label products. A specific feature is the presence of prestige brands in the major retail outlets, which is not common in other Member States.

(470) Sara Lee is the market leader in the non-male deodorant market with [20-30]*% market share. It is present with Sanex as well as Neutral. Sanex is a particularly strong brand in Denmark.

(471) Moreover, Sanex is a growing force: It has seen its sales increasing by almost [0-10]*% in value (between 2008 and 2009), compared to an overall growth rate of [0-10]*% in

the non-male deodorant market in the same period. [reference to parties’ internal 315 documents].

(472) Neutral has a [5-10]*% market share and has been growing by almost [20-30]*% in value. Its relatively strong market position adds an important element to Sara Lee's overall strength.

(473) Unilever is the number two player with a market share of [20-30]*%. It is present with its core brands Dove and Rexona. Unilever sales were rather stable between 2008 and 2009 (+[0-5]*%).

(474) Beiersdorf with its Nivea brand is the number three supplier. It has, however lost some market share lately as sales fell by [0-5]*% between 2008 and 2009.

(475) Colgate-Palmolive has been growing in Denmark ([10-20]*% between 2008 and 2009). Colgate-Palmolive's market presence, however, is still very modest (their value market shares increased from [0-5]*% in 2008 to [0-5]*% in 2009).

(476) L'Oreal's introduction of Garnier Minerals only occurred in 2010 so no market share was achieved in 2009. Garnier Mineral will be discussed in more detail in recitals (553).

(477) The transaction would bring together the two leading players in the Danish market 316controlling post-transaction four out of the five biggest brands.

(478) Table 19 shows the fragmented market structure in Denmark and the very limited presence of private label products ([0-5]*%). It also highlights the presence of the prestige brands.

317Table 19: Market shares in the non-male deodorant market in value, 2009, Denmark

2009

000€ %

UNILEVER [...]* 20-30

- DOVE [...]* 5-10

- REXONA [...]* 10-20

SARA LEE [...]* 20-30

315[reference to parties’ internal documents]

316Beiersdorf is the only competitor with a brand (Nivea) that achieves a turnover higher than Neutral The competitors Unicare and E.Tjellessen have higher market shares than Neutral, but this results from the combined turnover of several brands. According to Unilever’s internal documents, Unicare’s brand adidas [...]*. However, considering that adidas’s sales are made predominantly in the male deodorant market, where Unicare holds a market share almost three times its market share in the non-male deodorant market, it can reasonably be assumed that no Unicare brand has a market presence at least similar to Neutral ([5-10]*%).

317The non-male deodorant market in Denmark includes deodorants for women and family. The Parties submitted data for a third category, "other". However, as the Parties are not active within this third category, the market shares for the non-male deodorants market were calculated without taking into account this "other" category. For the sake of the completeness it should be noted that there are only three competitors active within this category, namely L'Oreal (2009 sales of EUR [...]*), SAETHER (2009 sales of EUR [...]*) and United Drugstores (2009 sales of EUR [...]*).

96

- NEUTRAL [...]* 5-10

- SANEX [...]* 10-20

COMBINED [...]* 40-50

BDF [...]* 10-20

Unicare [...]* 10-20

E.Tjellesen [...]* 5-10

Lancome [...]* 0-5

Colgate [...]* 0-5

SBI [...]* 0-5

Elisabeth Arden [...]* 0-5

SAETHER [...]* 0-5

Cederroth [...]* 0-5

Nygaard Cosmetics

[...]* 0-5

L'OREAL [...]* 0.5

PRIVATE LABELS

[...]* 0-5

United Drugstores [...]* 0-5

OTHERS [...]* 0-5

TOTAL 27 291 100.0

(479) A number of the prestige brands are present in Denmark through distributors. Unicare, the largest distributor with focus on the prestige segment, has a relatively high market share. However, almost all of its brands (the exception is Adidas with [0-5]*% market 318 share) achieve market shares below [0-5]*%.

(480) Saether supplies a range of premium/prestige deodorants such as Boss. Elisabeth Arden is present in the prestige segment with brands such as Chloe and Elizabeth Arden. 319Lancome, primarily supplies Biotherm and Lancome.

(481) Although Unicare grew strongly between 2008 and 2009 in comparison with the general market, many of the smaller suppliers suffered a loss of sales. For example, E.Tjellesen (which supplies primarily Gosh and Nothing) saw its sales decreasing by [20-30]*%. Lancome fell by [0-10]*% and SBI (which supplies Vanderbilt) by [0-10]*%.

Prestige brands – mass market brands

(482) The Parties argue that the "prestige brands" can compete on price with mainstream 320 brands as they are typically sold on deep price promotion.Nevertheless, the price

97

difference between these brands and the "mass market brands" such as the brands of 321 the Parties remains significant.

(483) The significant price differences are illustrated by Figure 15 and Figure 16 showing the average selling price per brands in Denmark for contact and non-contact formats. While the Parties' brands are positioned at an average price similar to competitors like Beiersdorf (Nivea), Coty (Adidas) and Colgate-Palmolive, brands like Hummel, Kappa or Vanderbilt are priced almost twice as high:

322Figure 15: Denmark Deodorant Prices Weighted by Volume, Contact (Source: Form CO) [...]*

323Figure 16: Denmark Deodorant Prices Weighted by Volume, Non-Contact (Source Form CO) [...]*

(484) In the reply to the Statement of Objections, Unilever stresses that the prices of prestige 324 brands have fallen significantly relative to mass market brands.According to Unilever, there was a steady decline in average retail price for prestige brands between 2005 and 2009.

(485) It may well be the case that the price gap between prestige and mass market brands has narrowed. However, it is difficult to draw firm conclusions on the basis of the analysis provided by Unilever. First, Unilever does not provide any detail as to which criteria were used to distinguish between prestige and mass market brands for this price 325 evolution analysis.Also, the brands taken into account are not the same throughout that period. Finally, the data includes some surprising information, notably a spectacularly price volatility of Gucci, a brand distributed by Saether. It price varied between [...]* per unit in 2005, [...]* per unit in 2006, [...]* per unit in 2007, [...]* per unit in 2008 and [...]* per unit in 2009.

(486) The Parties also argue that the growth of Unicare's sales by volume ([10-20]*%) is significantly stronger for 2009 than its growth in value terms ([10-20]*%), suggesting that the average price paid for the prestige brands is falling over time. Again, the information should be interpreted with caution and cannot automatically be attributed to a deliberate strategy on the part of Unicare. It appears that parallel imports of prestige

321 [...]*

322 See Figure 6.28, Form CO. Size of the circle represents volumes of SKUs sold in each EUR 0.5 average price band. Red spot indicates weighted average selling price for the brand. This data is price per unit due to the nature of the data available.

323 See Figure 6.29, Form CO.

324 See Unilever's Reply to the Statement of Objections, paragraph 14.11.

325 For example, Dolce and Gabbana, a brand supplied by Unicare, was included within the "mass market" brands. Also, regarding Unicare's brands, it is not explained why brands such as Naomi Campbell ([...]*), Gant ([...]*) and Moschino ([...]*), were considered as mass market brands, while Lamborghini ([...]*) was included within the prestige brands. Neither is it explained why Gosh and Nothing, two brands supplied by E.Tellejen should be considered as "prestige brands".

98

(487) Moreover, the largest supplier of the prestige brands considers itself particularly disadvantaged by its position, given that it "is depending on the brand owner, and cannot freely decide, for example, when to launch a new SKU's. A distributor cannot 327"follow" the trend of the market as easily as, for example, Unilever".

326 See non confidential minutes of a conference call with Unicare on 13 September 2010.

327 See non confidential minutes of a conference call with Unicare on 13 September 2010.

328329330(488) [reference to parties’ internal documents] , ,

328 [reference to parties’ internal documents]

329 [reference to parties’ internal documents]

330 [reference to parties’ internal documents]

(489) Finally, it has not been argued that prestige brands belong to a separate market. Despite 331their particular position on the market, the sales of prestige brands are included in the 332 calculation of the market shares.Due to their high price, the effect on the market shares are particular significant when these are measured by value. This effect on market shares can be seen by looking at the volume figures. Calculating market shares for the non-male market based on volume results in a combined market share of [50-60]*% for the Parties in 2009 (Unilever [20-30]*%, Sara Lee [20-30]*%) compared to [40-50]*% 333measured in value.

331 [...]*

332 If a market were defined excluding prestige brands, the market share of the Parties would be dramatically higher. Although a detailed breakdown of the non-male deodorant market in the different brands is not available, a rough calculation is possible assuming that the relative share of the highest price segment is similar for the non-male deodorant market compared to the overall market, i.e. represents a value share of roughly [20-30]*%. Recalculating the market shares of the Parties by taking out this prestige segment provides the following estimates: Unilever [20-30]*%, Sara Lee [30-40]*% and combined market share of [60-70]*%.

333 Comparing value and volume market shares for the overall deodorant market in 2009 of Unilever with those of Unicare and Saether further illustrates this point. Unilever has a value market share of [10-20]*%, Unicare is close with [10-20]*% and Saether [5-10]*%, the difference between Unilever and its competitors becomes significantly wider when market shares are measured in volume: Unilever has [20-30]*%, followed by Unicare with [10-20]*% and Saether [0-5]*%.

99

(490) Sara Lee is the number one deodorant supplier in Denmark, and has been one of the most successful suppliers in terms of market expansion in recent years. This performance is even more significant if considering that Sara Lee's customer base was already large before this expansion.

(491) Furthermore, the transaction would bring together the two leading suppliers as well as four of the five leading brands in an otherwise fragmented market. The overlap between the Parties is significant with Unilever holding a [20-30]*% market share pre-merger and Sara Lee a [20-30]*% share. Such a market structure is indicative of non-coordinated effects.

Unilever and Sara Lee's brands are close competitors

(492) In assessing the likely non-co-ordinated effects of the transaction in the differentiated non-male deodorants market, it is necessary to analyze whether the brands of the Parties are close competitors.

(493) The closeness of competition between Unilever and Sara Lee brands is analysed based on responses in the market investigation as well as internal documents and switching data. The strength of different brands when looking at market segments according to formats is also analysed, to complement the assessment.

(494) The Parties argue that a number of other significant suppliers are close competitors to 334 their brands.It is not contested that the Parties face rivalry from competitor brands such as Nivea. However, the interaction between the Parties' brands is significantly stronger than expected based on market shares.

(495) First, the responses of retailers in the market investigation show a somewhat mixed picture. Only one of the three responding retailers clearly identified Sanex as the 335 closest competitor to Unilever brands.However, they all mentioned Rexona as a 336 close competitor to Sanex in terms of price and product range.Two of them also pointed to the closeness of competition between these brands in terms of brand 337 recognition.Two retailers consider Dove and Sanex as close competitors in terms of 338price and product range.

339340(496) [reference to parties’ internal documents] ,

341342(497) [reference to parties’ internal documents] ,

334 In the Reply to the Statement of Objection, Unilever argues that Nivea has a similar skin care proposition to Dove and Rexona. Reference is also made to Palmolive, a brand which has a natural/skin care proposition and would be "particularly strong in roll-on formats". Garnier Mineral is also mentioned by Unilever as competing with Dove ("Garnier Mineral has similar skin care and "natural" attributes as Dove") and Rexona, in particular with Rexona Natural Mineral and Dove Go Fresh.

335 See reply from [retailer] to Questionnaire to customers on deodorants sent on 23 April 2010, question 30.

336 See replies to Questionnaire to customers on deodorants sent on 23 April 2010, question 29.

337 See replies from [retailer] and [retailer] to Questionnaire to customers on deodorants sent on 23 April 2010, question 29.

338 See replies from [retailer] and [retailer] to Questionnaire to customers on deodorants sent on 23 April 2010, question 29.

339 [reference to parties’ internal documents]

340 [reference to parties’ internal documents]

100

(498) In order to assess the closeness of competition between the brands, Unilever provided interaction indices. Figure 17 shows the interaction index for Sanex for women 343(2009).

Figure 17: Interaction index for non-male Sanex in Denmark

[...]*

(499) Unilever notes that, according to the index, Sanex interacts with many brands. Although it interacts with Dove ([150-170]*) and Rexona ([120-140]*), Sanex also has an equal or higher level of interaction with several third parties' brands, notably Palmolive for Men ([270-290]*), Palmolive women/unisex ([120-140]*), Hugo Boss for Men ([120-140]*) as well as Nivea women/unisex ([120-140]*).

(500) The interaction index shows some results that appear rather counterintuitive such as high interaction with certain male brands. Sanex for women has a very high interaction index with Palmolive for Men ([270-290]*), as well as Sanex for Men ([250-270]*).

(501) The interaction indices presented by Unilever are a comparison of the actual gross consumer switching of the brand (here, Sanex for Women) with the "expected" switching based on the proportions of the competing brands' market shares in the total 344 deodorant category (combining male and non-male deodorants).A high interaction of Sanex index with a given brand is not in itself indicative of significant switching figures between Sanex and this brand because this high index can be brought about by the small share of the brand. When the market share is taken into account, the actual consumer switching as observed in the panels is significantly lower than the interaction index alone would suggest.

(502) Therefore, while interaction indices provide some insights about the closeness of competition between brands, the gross switching information provides a more illustrative picture about the actual switching patterns of consumers.

(503) Figure 18 shows in descending order for each brand their share of total gross switching to and from Sanex for Women in 2009.

341 [reference to parties’ internal documents]

342 [reference to parties’ internal documents]

343 Each brand may in a given period lose customers to (all) other brands and simultaneously win customers from all other brands. The sum of lost and won customers between two brands, the total gross switching is an indicator of how closely the two brands are competing. The interaction indices presented by the Parties are a comparison of the actual gross consumer switching of the brand (here, Sanex for Woman) with the "expected" switching based on the proportions of the competing brands' market shares in the total deodorant category (combining male and non-male deodorants). The market shares are calculated on the basis of the brands' sales within the panel. This may to some extent vary from the market shares based on Nielsen data, but in general provides a rather consistent picture with those.

344 The market shares are calculated on the basis of the brands' sales within the panel. This may to some extent vary from the market shares based on Nielsen data, but in general provides a rather consistent picture with those.

101

(529) Post transaction, the Parties would own the two most sold brands in Denmark (Rexona and Sanex). Moreover, the transaction will bring together four of the five main brands in the deodorant market, which has a direct impact on the ability and incentives of retailers to use their alleged countervailing buyer power. Combination of the Parties strong brands in Denmark in the hands of a single supplier would reduce the set of credible alternatives that the retailer can choose from to replace Unilever's products if he has the willingness to do so. In the Danish market structure, it would be very difficult for the retailer to construct a deodorant shelf without Unilever's products. If the retailer is not willing to entirely eliminate Unilever’s products from the shelves, even a partial delisting of important Unilever brands such as Axe, Dove or Rexona would clearly endanger the retailer's turnover in the deodorant category.

(530) The market investigation indicates that "must have" deodorant brands exist, although views of the retailers differ to some extent on the brands that a retailer would necessarily have to include in its shelves.

(531) When asked if some brands/SKUs have such a significant presence on the market that they cannot credibly be delisted, retailers indicate that "some brands are significant and would be missed if we delist them".Moreover, the majority of retailers indicated that their customers would either switch their deodorant purchases or their entire purchases to another retailer in case one of the core brands of the Parties would not be available in their stores.

(532) The leading retailer in Denmark identifies both Sanex and Nivea as "must have" brands.Another retailer as well identifies Sanex as a "must have" brand both for

"unisex" deodorant and for "female" deodorant. Other brands are also cited, like Rexona and Garnier Mineral or Neutral and private label products.

(533) In its reply to the Statement of Objections, Unilever points out that one retailer only mentioned its own private label brands as a "must-have" brand, and considered that no brand/SKU is so significant that they cannot credibly threat to delist them. However, when questioned about the possibility of finding alternative products if not able to stock either Party's products, the answer is less clear: "yes, both Sanex and Rexona would cause great loss in turnover for us if they were still sold in other channels".

(534) The strong position of Unilever and Sara Lee as deodorant suppliers is also reflected in their presence in the retail channels, which is illustrated by the data on weighted distribution. Weighted distribution represents the share of sales accounted for by all the stores in which the brand is stocked. It is therefore a measure of the retail exposure given to a brand. Table 22 shows that Unilever, Sara Lee and Beiersdorf managed to secure widespread distribution compared to their competitors in Denmark .

Table 22: Deodorants weighted distribution, Denmark (excluding the hard-discount distribution)

21 20 YTDYTDJUNEJUNE MAT2009201020092010

T. AXE […]* […]* […]* […]* […]*

T. DOVE. […]* […]* […]* […]* […]*

T. REXONA […]* […]* […]* […]* […]*

T. NIVEA […]* […]* […]* […]* […]*

T. SANEX […]* […]* […]* […]* […]*

T. NEUTRAL […]* […]* […]* […]* […]*

T. WILLIAMS […]* […]* […]* […]* […]*

T. BOSS […]* […]* […]* […]* […]*

T. ADIDAS […]* […]* […]* […]* […]*

T. VAN GILS […]* […]* […]* […]* […]*

T. HUMMEL […]* […]* […]* […]* […]*

T. VANDERBILT […]* […]* […]* […]* […]*

T. PALMOLIVE […]* […]* […]* […]* […]*

(535) Axe, Dove, Rexona, Sanex, Neutral, Nivea and Palmolive are the brands with the highest weighted distribution. The other brands appear to have less than full access to all retail channels.

Private label products

(536) Private label products only represented [0-5]*% of the Danish non-male deodorant market in 2009. This share has even decreased from [0-5]*% in 2008. Hence, private label products do not provide a possibility for retailers to counterbalance the increasing market power obtained by Unilever through the merger.

(537) In its reply to the Statement of Objections, Unilever contested the private label products market share in the non-male deodorants market ([0-5]*%), although it finally agrees that "the presence of private label products is limited in this market".

(538) The limited market presence of private label products is also confirmed by retailers. The leading retailer in the Danish market (accounting for at least [30-40]*% of the total deodorants sales) explained that their private label products represent less than [0-5]*% of total deodorant sales.

(539) In its reply to the Statement of Objections, Unilever maintains that private label products offer a significant competitive constraint on the Parties despite their limited market presence.

(540) The analysis about interaction between brands has not specifically considered private label products. Clear conclusions cannot be drawn based on the answers in the market investigation to the specific question about closeness of competition. When asked which brands retailers consider to be a close competitor to their private label products in terms of quality, product range and other characteristics, they indicated that Sanex, Neutral, Nivea and other retailers' private label products are close competitors to their private label products. However, it is not surprising that a retailer would mention the key mass-market brands as competitors to their private label products rather than the prestige brands. When asked which brands they consider as the closest substitutes for Sanex, only one retailer mentioned private label products.

(541) In general, the market investigation confirms the weak role of private label products in Denmark. Only one retailer expressed its intention to increase sales of private label products. However, when asked about plans to increase the share of their private label

products significantly in the next 1-2 years, it answers negatively, "because private label does not perform very well in personal care categories".

Conclusion on buyer power

(542) The high retailer concentration in Denmark could indicate that the largest retailers have a strong negotiation position in relation to suppliers of deodorants. However, Unilever and Sara Lee are in an equally strong position in their negotiations with retailers given their strong brands and broad presence in the retail sector. Merging the number one and number two suppliers in the market as well as four of the five largest brands would enhance this position substantially. [...]*

(543) Moreover, high retail concentration does not mean that all retailers in the market are in the same market position as regards the deodorant suppliers. Smaller retailers are in a very different position to the largest retailers when negotiating with suppliers compared.

(544) Finally, private label products have a very limited impact in Denmark. Hence, negotiation power arising from the possibilities retailers would have to substitute the Parties brands with private label brands is particularly limited in Denmark.

(545) Therefore, it can be concluded that countervailing buyer power post-merger cannot be considered as a mitigating factor which is likely to off-set any of the potential adverse effects of the proposed transaction.

Sufficient entry unlikely to occur

(546) Section IV.2.3.1.4 analysed, from a general point of view, that entry in deodorant markets is difficult because of high barriers resulting from the need to create brand awareness and access to shelf space. A potential entrant would also have to take into account the likely reaction of a powerful incumbent like Unilever, [...]*. In this section, particular features of the Danish market are discussed, complementing the general analysis.

(547) Unilever argues that the prospect of new entry is a particular competitive constraint in Denmark due to new entry of prestige brands. Unilever explains that "Unicare frequently refreshes its range in the market, for example launching David Beckham in September 2005, Hummel in October 2005 and Kappa in November 2008".

(548) First, the brand and price positioning of prestige brands is distant from the Parties brands. Since the existing prestige brands do not exercise a sufficient competitive constraint on the mainstream brands of the merged entity, this is equally likely to be the case for new prestige entry. Furthermore, these brands individually have very limited market shares.

(549) LECG has, on behalf of a competitor, submitted a detailed analysis of brands entering the Danish market between 2007 and 2010. The study shows that although 22 new brands were introduced in Denmark during that period, those brands obtained only a minimal market share. In June 2010 [5-10]*% of total sales was accounted for by these brands. Most of these examples are high end deodorants from strong fashion brands that are already well known. At the end of the period, only two brands had a market share above [0-5]*% (Kappa with [0-5]*% and Garnier with [0-5]*%). Kappa clearly belongs to the high priced segment (see Figure 15). Garnier Mineral is discussed separately in recital (553).

(550) In the reply to the Statement of Objection, Unilever argues that "even if prestige brands only obtained a small market share each", they account for a significant share of the market in aggregate. Unilever provided data in order to show that of the 30 brands gaining market share in Denmark in the past two years no less than 21 are prestige brands and half of these appeared to be new on the market. According to Unilever, this would clearly demonstrate that new entries by prestige brands are a competitive force in the market.

(551) The new data provided by Unilever confirms the limited market share achieved by the new prestige brands. Out of the 10 brands referred to by Unilever as new prestige brands, only two achieved a market share higher than 1%, which it further reinforces the view that these new prestige brands are very different from the mass market brands. The fact that there is a high turnover of brands within the prestige segment so that relatively young brands take up a significant share of total sales cannot bring into question that they (taken together) exercise a very limited constraint on the mainstream brands.

(552) In the reply to the Statement of Objection, Unilever also elaborates on the entry of Garnier Mineral, which was introduced in Denmark in February 2010. It is explained that Garnier Mineral's succeeded in achieving a weighted distribution of [80-90]*% and a relatively high proportion of shelf space ([0-10]*%) in the non-male market.

(553) It is difficult to predict how the market position of Garnier Mineral will develop although latest information indicates that Garnier Mineral has not been able to maintain its initial market share.Information on the latest developments of monthly market shares that Garnier Mineral achieved since its introduction show that Garnier Mineral has been loosing market share in recent months after having achieved a [5-10]*% value market share in April.

(554) Unilever reacted immediately to the entry by Garnier Mineral by launching Rexona Mineral in May 2010. [...]*

Figure 21: Garnier Mineral in Denmark since launch – Source: Nielsen

[...]*

(555) Despite a high number of new "prestige brands" in the Danish market, these entrants do not effectively constrain the Parties due to their different market position and very low market shares. The recent entry in the "mass-market" segment by Garner Mineral is difficult to assess at this stage. Market shares have in recent months been decreasing from the peak of [5-10]*% in April.

(556) It is concluded that sufficient new entry to counteract the adverse effects of the merger in the Danish non-male deodorant market cannot be expected to occur.

Overall conclusion

(557) It is therefore concluded that the notified concentration is likely to significantly impede effective competition on the market for non-male deodorants in Denmark.

IV.2.3.2.2.2. Male deodorant market

(558) According to the Parties, the Danish male deodorant market had a total value of EUR 4.4 million in 2009. It experienced an increase in value of [0-10]*% between 2008 and 2009. Unilever is active in the male deodorant market with its brands Axe and Rexona for Men. Sara Lee serves the Danish male deodorant market with its brand Sanex for Men and has sales with its Williams brand.

(559) The Parties would post-merger achieve a combined market share of [20-30]*% (Unilever [10-20]*%, Sara Lee [5-10]*%). Other competitors active in the Danish male deodorant market are Unicare ([20-30]*%) Saether ([10-20]*%), Beiersdorf ([10-20]*) and L'Oreal ([0-5]*%).

(560) Given the limited combined market share post-merger and the fact that retailers did not consider Sanex for Men or Williams to be close substitutes for Unilever's Axe or Rexona for Men brands, but rather Adidas, Hummel or Kappa, it is concluded that the proposed transaction would not result in a significant impediment of effective competition in the market for male deodorants in Denmark.

IV.2.3.2.3. Germany

(561) In Germany, the overall deodorant market had a value of EUR 634.5 million in 2009, (EUR 373.8 million non-male deodorant market; EUR 260.7 million male deodorant market). Independent of the relevant product market definition, the proposed transaction will only give rise to an insignificant change in the market structure since the combined market share as well as the increment is low and several sizeable competitors exist post-transaction.

Table 23: Germany – Deodorant market shares in value – 2009 – Source: Form CO.

Unilever Sara Lee Combined Competitors

Overall [20-30]*% [0-5]*% [30-40]*%

Beiersdorf [20-30]*%

Henkel [5-10]*%

Coty [5-10]*%

Male [30-40]*% [0-5]*% [30-40]*%

Beiersdorf [20-30]*%

Coty [10-20]*%

Henkel [5-10]*%

Non-male [20-30]*% [0-5]*% [30-40]*%

Beiersdorf [20-30]*%

Private Label [10-20]*%

Henkel [5-10]*%

(562) The combined market share on the male as well as the non-male deodorant market would be below [30-40]*% and the increment brought about the transaction would be at most [0-5]*%. In both markets, several competitors, in particular Beiersdorf with almost [20-30]*% market share, but also Henkel, Coty and Private Label are present. In addition, all retailers considered Sara Lee's Duschdas brand to be a close competitor to Henkel's FA brand and Beiersdorf 8x4 brand, all positioned at the lower end of the market, where Unilever is not active.Respondents to the market investigation did not anticipate any anti-competitive impact on competition as sufficient alternatives would be available post-transaction.

(563) Based on these elements, it is concluded that the concentration is not likely to lead to a significant impediment of effective competition as regards the markets for male and non-male deodorants in Germany.

IV.2.3.2.4. Ireland

IV.2.3.2.4.1. Non-male deodorant market

(564) A number of factors indicative of significant non-coordinated effects are present in the non-male deodorant market in Ireland.

Similarities with the market structure in the United Kingdom

(565) As mentioned by the Parties in the Form CO "In important respects, the market for deodorants in Ireland is similar in character to the market for deodorants in the UK, and it follows that the key aspects of the competitive analysis are the same. The structure of supply of the deodorants market in Ireland is very similar to that of the UK, as can be seen below. Procurement by [...]* across its UK and Irish businesses as a whole is leading to increasing alignment of terms of trade to retailers generally, for the reasons mentioned below. Many channels of advertising and media support are common to both markets. As such, Unilever views conditions of competition within Ireland as derivative of those prevailing in the UK market. "

(566) The Parties accept that, in important respects, the market for deodorants in Ireland is similar in character to the market for deodorants in the United Kingdom. For that reason, the competitive assessment of the Irish market will, in some aspects, closely follow the assessment of the deodorant market in the United Kingdom.

Merging firms have high market shares

(567) According to the Parties, the non-male deodorant market in Ireland had a total value of EUR 12.4 million in 2009, [90-100]*% of which is covered by supplier brands and the remaining [0-5]*% by private labels. It experienced a decrease in value of [20-30]*% between 2007 and 2009. Unilever is mainly active through its brands Dove, Sure (which is the best selling brand), Impulse and Vaseline while Sara Lee serves market with its brands Sanex and Radox.

Table 24: Parties market shares in value by brands – Ireland, non-male deodorants. Source: Form CO

2009

000 € %

UNILEVER [...]* [60-70]*

- DOVE [...]* [10-20]*

- IMPULSE [...]* [5-10]*

- SURE [...]* [30-40]*

- VASELINE [...]* [0-5]*

SARA LEE [...]* [5-10]*

- RADOX [...]* [0-5]*

- SANEX [...]* [5-10]*

COMBINED [...]* [60-70]*

ACCANTIA [...]* [0-5]*

BEIERSDORF [...]* [10-20]*

COLGATE [...]* [10-20]*

COTY [...]* [0-5]*

GILLETTE [...]* [0-5]*

CUSSONS [...]* [0-5]*

HENKEL [...]* [0-5]*

REVLON [...]* [0-5]*

PRIVATE LABELS [...]* [0-5]*

OTHERS [...]* [0-5]*

TOTAL 12 449 100.0

(568) The Parties achieve a combined market share of [60-70]*%. It will exceed by five times the share of its nearest competitors Colgate-Palmolive with its brands Palmolive and Soft&Gentle ([10-20]*%), Beiersdorf with its brand Nivea ([10-20]*%) and Revlon with its brand Mitchum ([0-5]*%). Other branded players as well as private labels ([0-5]*%) hold insignificant market positions (below [0-5]*%) in the non-male deodorant market.

(569) The market has decreased in value terms by [20-30]*% between 2007 and 2009, probably due to the dire economic situation in Ireland. Although all brands have seen their sales drop, the impact on firms' market position varies significantly. Unilever has been the most successful supplier since its market share has increased from [50-60]*% in 2007 to [60-70]*% between 2007 and 2009. Though to a smaller extent, Beiersdorf has also increased its market share (from [5-10]*% in 2007 to [10-20]*% in 2009) as well as Revlon (from [0-5]*% to [0-5]*%). Conversely, Colgate-Palmolive market share eroded significantly during the same period (from [10-20]*% in 2007 to [10-20]*% in 2009). Sara Lee's market share also slightly decreased (from [5-10]*% to [5-10]*%) but this decrease is mainly due to Radox (whose share went from [0-5]*% to [0-5]*%). [...]*[...]*. Sanex's share remained stable (from [5-10]*% in 2007 to [5-10]*% in 2009).

(571) The combined high market share suggests that the merged entity is likely to enjoy significant market power in the absence of any mitigating factors

(572) Furthermore the overlap between the Parties is significant with Unilever holding a [60-70]*% market share pre-merger and Sara Lee a [5-10]*% share. This is indicative of strong non-coordinated effects which would likely lead to higher prices and/or reduced quality and choice. Following the acquisition of Sara Lee, Unilever will control four out of the six major brands in the non-male deodorant market in Ireland (Sure, Dove, Impulse and Sanex). No other supplier holds more than one significant brand in the market.

(573) In its reply to the Statement of Objections, Unilever contested Sara Lee's position by referring to a retailer which argued that "Sanex is insignificant on the Irish market" . However, Sara Lee only represents [0-5]*% of the retailer's turnover in deodorants . This share is not representative of the structure of the Irish market in which Sanex reaches a market share of [0-5]*% on the overall deodorant market and even [5-10]*% in the non-male deodorant market. In addition, this retailer is not even mentioned as one of top five retailers in terms of deodorant sales in Ireland (neither for Unilever nor for Sara Lee). Thus, this statement cannot be considered to be representative of Sanex's position on the Irish market.

Merging firms are close competitors

(574) Besides market shares, another important aspect for assessing the unilateral effects arising from the proposed merger is the closeness of the Parties' non-male deodorants. The higher the degree of substitutability between their products, the more likely it is that the Parties will significantly raise prices to retailers. In addition to pointing to high market shares, retailers explained that the Parties are close competitors. In particular, Dove and Sanex share skin-caring attributes which explains why two out of three retailers considered that the closest competitor of Sanex would be Dove. One retailer also indicated that Sanex would be the closest competitor to Vaseline.

(575) Unilever submitted that a number of significant suppliers will continue to impose a competitive constraint on Unilever post-transaction. It notably argues that the responses in the market investigation pointed to Nivea (Beiersdorf), Soft&Gentle (Colgate-Palmolive) and Natrel (Henkel) as close to Sanex . However, closer inspection of these brands and their characteristics leads to the conclusion they do not exercise a particularly strong constraint on the main brands.

(576) Soft&Gentle (the main brand of Colgate-Palmolive) is described by the Parties as follows: "Soft & Gentle has a focus on fragrances, although the product range includes variants targeted at efficacy and skin care. All the products in the Soft & Gentle range contain an element of fragrances or exotic oils, alongside the antiperspirant effect and functionality of the product. Advertising is directed at the female market with an emphasis on mood enhancing, confidence, feel good-factor and escapism."The characteristics of Soft&Gentle do not include it in the range of "skin-care" deodorants where Dove, Vaseline, Sanex and Nivea are positioned. In addition Colgate-Palmolive with its Soft&Gentle brand was the one with the most declining market share on the market (losing [...]* percentage points from 2007 to 2009). Only one out of three Irish retailers considered Soft&Gentle to be a close competitor to Dove, and another mentioned the brand as among the closest competitors to Sure. None of them mentioned the brand as among the closest competitors to Sanex.Moreover, Soft&Gentle is not considered as interacting significantly Sanex neither in the interaction index provided by the Parties nor in the gross switching analysis

(577) Regarding Natrel (Henkel's second brand after Right Guard), one Irish retailer mentioned this brand as the closest competitor to Sanex whereas the two other considered Dove as the closest Sanex's competitor. However, Natrel has an extremely

388 Form CO, paragraph 6.256.

389 Questionnaire to customers on deodorants sent on 23 April 2010, questions 27 and 30.

390 Form CO, paragraph 6.64.

391 Questionnaire to customers on deodorants sent on 23 April 2010, question 30.

117

limited position knowing that Henkel achieve a market share of [0-5]*% with its two brands (Right & Guard and Natrel). In addition, Natrel was not discussed in the Form CO of the Parties. As such, Natrel is unlikely to constrain the combined entity's position to any appreciable extent.

(578) Finally, Nivea was not mentioned by any of the retailers as the closest brand to Sanex but has been referred as the closest brand to Dove by two retailers. Indeed, Nivea is in a cluster of close competitors together with other brands with a skin caring proposition - Dove, Sanex and Vaseline. It may not be appropriate in this particular case to label any of these particular brands as "the closest" for the purpose of drawing conclusions about the competitive constraints. While Nivea may score closer to Dove in some brand positioning characteristic, other elements suggest that it may be Sanex which is more closely competing with Dove in some format.

Assessment on the basis of format

(579) Looking at market shares by formats reveals additional elements regarding the closeness of substitution between Dove and Sanex. In addition to being skin-friendly brands, Dove and Sanex have a particular strength in the roll-on format ([20-30]*% of the non-male Irish market) since their market share for roll-on formats is significantly above their overall non-male share ([20-30]*% in roll-on formats and [10-20]*% overall for Dove, [10-20]*% in roll-ons formats and [5-10]*% overall for Sanex).

(580) Shares of sales between roll-ons and aerosolsformats are more equally spread for Nivea, Colgate-Palmolive and Vaseline brands whereas Sure is rather an aerosol brand and Impulse is exclusively available in spray format. Other brands (Right Guard, Secret, Adidas) are only available in spray formats except for Mitchum.

Table 25: Ireland: Non-male deodorants, 2009: market shares – Source: Form CO.

ROLL-ON AEROSOL

000 € % 000 € %

UNILEVER [...]* [40-50]* [...]* [60-70]*

- DOVE [...]* [20-30]* [...]* [10-20]*

- IMPULSE [...]* [0-5]* [...]* [10-20]*

- SURE [...]* [20-30]* [...]* [30-40]*

- VASELINE [...]* [0-5]* [...]* [0-5]*

SARA LEE [...]* [10-20]* [...]* [0-5]*

- RADOX [...]* [0-5]* [...]* [0-5]*

- SANEX [...]* [10-20]* [...]* [0-5]*

COMBINED [...]* [60-70]* [...]* [70-80]*

BEIERSDORF IRELAND [...]* [10-20]* [...]* [10-20]*

COLGATE PALMOLIVE [...]* [10-20]* [...]* [10-20]*

COTY [...]* [0-5]* [...]* [0-5]*

GILLETTE [...]* [0-5]* [...]* [0-5]*

392 Aerosols and roll-on formats represent [90-100]*% of the formats sold in the non male deodorant market in Ireland.

118

HENKEL [...]* [0-5]* [...]* [0-5]*

REVLON [...]* [5-10]* [...]* [0-5]*

OTHERS [...]* [0-5]* [...]* [0-5]*

TOTAL 2 506 100.0 9 478 100.0

(581) Table 25 therefore confirms the market investigation that Sanex is a close competitor to Dove (and to a lesser extent to Nivea).

Switching data analysis

(582) Unilever also provided interaction indices for Sanex for Women for the year 2009. This data submitted on 26 July 2010build on detailed purchase records of individuals and, as previously explained in recital (374) when addressing the Belgian market, should be interpreted cautiously.

(583) Although contaminated by these factors, the switching data may however also contain information about actual substitutability and hence closeness of competition. Each brand may in a given period lose customers to (all) other brands and simultaneously win customers from all other brands. The sum of lost and won customers between two brands, the total gross switching is thus an indicator how closely the two brands are competing. The interaction indices presented in Unilever's submission of 26 July 2010 are shown in Figure 22.

Figure 22: Interaction index for non-male Sanex in Ireland

[...]*

(584) The Parties explain that, although the data show that Sanex for Women interacts more closely with Vaseline ([180-200]), Dove for Women ([110-130]) and Sure for Women ([110-130]), or even [180-200] with Sure for Girl), Sanex for Women has also an equal or higher level of interaction with several third parties' brands, notably Nivea ([110-130]) Coty/Adidas ([200-220]), Revlon/Mitchum ([200-220]) and Henkel/Right Guard for Men and Women ([150-170]/[110-130]). A higher degree of interaction is also evident with Sanex for Men ([240-260]) and Radox for Men ([280-300]).

(585) The interaction indices presented by Unilever are a comparison of the actual gross consumer switching of the brand (here, Sanex for Woman) with the "expected" switching based on the proportions of the competing brands' market shares in the total deodorant category (combining male and non-male). A high interaction index alone

393 "M.5658 Unilever/Sara Lee Body Care: Supplementary submission regarding interaction indices" submitted on 26 July 2010.

394 With regards to Right Guard, Mitchum for Women or Adidas of Women, none of the three Irish retailers mentioned these brands as the closest competing brands to Dove, Rexona, Vaseline or Sanex. In addition, these brands have low shares of respectively [0-5]*% for Mitchum, [0-5]*% for Right Guard and [0-5]*% for Adidas.

395 The market shares are calculated on the basis of the brands' sales within the panel. This may vary from the market shares based on Nielsen data to some extent, but in general provides a rather consistent picture with those.

119

therefore does not account for the absolute size of a brand in the market. When the market share is taken into account, the actual consumer switching as observed in the panels is significantly lower than the interaction index alone would suggests.

(586) Therefore, while interaction indices provide some insights about the closeness of brands, the gross switching figure from and to a brand provides a more illustrative picture about the actual switching patterns of consumers. The Parties have submitted the underlying data used by GFK/Europanel containing the actual switching figures. Based on the same data as used for the interaction index, Figure 23 shows in descending order for each brand their share of total gross switching to and from Sanex for Women in 2009. Consistent with the method used by GFK/Europanel, the total gross switching is compared with the proportions of the competing brands' market shares in the total deodorant category ("value of competitors").

Figure 23: Gross switching for non-male Sanex in Ireland

[...]*

(587) On this basis of this data, the proportion of buyers switching from Sanex to other brands in the market was calculated. In that respect, of all the Sanex for Women customers that switched with another non-male brand, [50-60]*% interacted with a Unilever brand. Of all the Sanex for Women customers that switched with a non-Sanex brand (either male or non-male) [50-60]*% interacted with a Unilever brand.

(588) The level of switching from the panel data is in line with Unilever's market share. In fact, this only confirms that the Parties' very high market shares are a good proxy for analysing the effects of the transaction.

(589) The indication from the consumer panel data that around [40-50]*% of all Sanex switching within the non-male market in Ireland occurs in relation to Unilever's brands means that Unilever does indeed represent a very significant constraint on Sanex. Even if the other half of switching is with other brands, the merger instantly removes half of the total competitive constraint previously exercised on Sanex. This means that the incentives for price increase of Sanex are significant post-merger.

Conclusion on closeness of competition.

(590) The market investigation demonstrated that the majority of retailers considered Dove to be a close competitor to Sanex. In addition, the format analysis indicates that Dove and Sanex are the two brands which are particularly strong in roll-on. Finally, switching analysis data shows Sanex is switching a lot with Sure and Dove.

(591) It is concluded that the merger would eliminate an important competitive force on the market which significantly contributes to constrain Unilever on the non-male deodorant market in Ireland.

Lack of countervailing buyer power

(592) The reasons why a potential price increase after the merger is unlikely to be defeated by countervailing buyer power is analysed in section IV.2.3.1.3. These reasons equally apply to the Irish non-male deodorant market.

396(593) The penetration of private labels in this market is extremely limited ([0-5]*%) although Irish retailers have in general submitted during the market investigation that they achieve significantly higher margins in private labels .

(594) In its reply to the Statement of Objections, Unilever argues that one retailer planned to introduce private label without clear timeline.

(595) Nevertheless, Unilever does not mention that the two retailers do not predict an increase in the share of private labels in the next 1-2 years. One argues that the demand is too low and that "The deodorants category has been so deflated in the past 12-18 months that the price gap between branded and private label deodorants is negligible. Also the strength of many of the deodorant brands is such, and the retail price so low, that own brand is not very attractive to consumers in this category " and the second mentions that the private labels were mainly "launched in response of the economic situation".

(596) As a consequence, customers would not be in a position to switch to private labels after the merger to counterbalance the increasing market power obtained by the new entity.

(597) In its reply to the Statement of Objections, Unilever argues that Irish market is highly concentrated and reiterates that given that the combined share of top three retailers is over [70-80]*% (and the top five over [80-90]*%), the customer structure is particularly concentrated, inferring buyer power. It also puts forward that even the smallest of the top five accounts for a significant percentage of Parties sales ([10-20]*% for Unilever and [5-10]*% for Sara Lee). Unilever further argues that losing sales even for this relatively smaller distributor would mean not being able to make-up significant business elsewhere. Unilever also submits that there is no evidence that small retailers get a worse deal compared to larger retailers. Because smaller retailers have sufficient alternatives to Unilever's deodorants at their disposal they are able to achieve good terms with Unilever.

(598) Even though the three biggest retailers account for [70-80]*% of the national deodorant sales, important variations occur amongst them: the main retailer accounts for more than [30-40]*% of the national deodorants sales whereas the second and the third retailers account respectively for [20-30]*% and [10-20]*% of sales. Fourth and fifth retailer are significantly smaller (both hold [5-10]*% each).

(599) First, the concentration of customers on which Unilever bases its argument cannot itself be indicative of countervailing buyer power. While the market share of the largest retailer is [30-40]*%, it remains small compared to the Parties' combined market share of [70-80]*% of all deodorant sales (male and non-male). Also, as a

396 In addition, the private label share has remained stable in the last three years ([0-5]*% in 2007).

397 Questionnaire to customers on deodorants sent on 25 June 2010, question 1.

398 Questionnaire to customers on deodorants sent on 25 June 2010, question 23.

399 Questionnaire to customers on deodorants sent on 25 June 2010, question 23 a).

121

retailer indicates, end-consumers expect to find Unilever's strong and popular brands on the shelf, and retailers cannot ignore this .

(600) In such a market structure, it would seem unrealistic for the retailer to construct a deodorant shelf without one or several important Unilever brands which makes Unilever an unavoidable trading partner in relation to prices setting, promotions, new listings and placement on shelves. In a market where all other suppliers would hold only one non-male brand, it is clear that a shelf with only Nivea, Soft & Gentle and Palmolive would not be complete in terms of set of brands that the customer expects to find in the stores and product attributes.

(601) The fact that top five retailers have a high proportion of the total market does not therefore give these retailers buyer power, as could potentially be the case with perfectly substitutable homogeneous goods where customers could switch all their demand to other readily available alternatives. This is simply not the case here since consumers consider the merger would bring together close substitutes (Sanex is perceived as being close to Vasenol and Dove) for which the number of alternatives are limited. The finding is equally relevant for the retailers, who must assemble a portfolio of brands on the shelf that adequately cover the needs of all customers.

(602) Second, as explained in Section IV.2.3.1.3, even if a large retailer were to derive some degree of bargaining power from its large size or a sophisticated purchasing strategy, there is no generally convincing reason why other retailers should also be positively affected. Thus, even if countervailing buyer power were to exist for the larger retailer, it is unlikely to sufficiently off-set potential adverse effects of a merger since it would only ensure that a particular segment of customers, with particular bargaining strength, is shielded from significantly higher prices or deteriorated conditions after the merger.

(603) Third, when comparing smaller and larger retailers, it is not argued that large buyers have bargaining power. The analysis rather answers the question whether smaller retailers would benefit from a hypothetical situation in which large buyer had some bargaining power. In this context, even if large retailers would have buyer power, Unilever's arguments that the same degree of buyer power lies with small retailers cannot be sustained.

(604) The very strong position of Unilever compared to its competitors and the fact that Unilever is an unavoidable trading partner, are also reflected by the weighted distribution in retail stores of the different suppliers in Ireland. Weighted distribution represents the share of sales accounted for by all the stores in which the brand is stocked. It is therefore a measure of the retail exposure given to a brand. Table 26 shows that Unilever managed to secure widespread distribution compared to its competitors .

Table 26: Deodorants weighted distribution, Ireland

400 Questionnaire to customers on deodorants sent on 25 June 2010, question 16

401 See paragraph 67 of the Horizontal Merger Guidelines.

402 [...]*

122

Ireland

Same Period YAGO Previous Period

LYNX [...]* [...]*

DOVE [...]* [...]*

SURE [...]* [...]*

IMPULSE [...]* [...]*

[...]* [...]*

NIVEA [...]* [...]*

[...]* [...]*

SOFT & GENTLE [...]* [...]*

[...]* [...]*

SANEX [...]* [...]*

RADOX [...]* [...]*

[...]* [...]*

GARNIER [...]* [...]*

MEN EXPERT [...]* [...]*

[...]* [...]*

MITCHUM [...]* [...]*

[...]* [...]*

RIGHT GUARD [...]* [...]*

[...]* [...]*

(605) Lynx, Dove, Sure, Impulse, Nivea and Soft&Gentle are the brands with the highest weighted distribution in Ireland (above [...]*). Post-transaction, Unilever will control five out of the seven brands with the highest weighted distribution rate, and reinforce its position as unavoidable trading partner for the retailers.

(606) Unilever also states that the retailers have alternative sources of supply than the Parties brands and that consumers would rather switch brand than switch store. To substantiate this position, Unilever points to figures which show that loyalty to major

123

Irish retailers is generally significantly higher ([...]*) than to Unilever and Sara Lee brands ([...]*) .

(607) This statement was not confirmed in the market investigation. Indeed, all Irish retailers stated that if their store were not to offer a particular deodorant brand of Unilever/Sara Lee, the customers would switch to other retailers (for the whole shopping basket or for the deodorants only) .

(608) Furthermore, the loyalty figures do not – as claimed by Unilever – confirm that customers would rather continue buying from their retailer than switch brands when their preferred deodorant brand is not available. First, these figures try to answer two isolated questions – loyalty to a deodorant brand, and loyalty to a shop – but do not allow a conclusion as to how many consumers would switch retailers if their deodorant brand were not available. Second, even if customers were, on average, more loyal to their retailer than to the brand, there is a significant share of customers for which the loyalty for a brand is important. As switching customers would probably re-direct their entire shopping basket to another retailer, the losses for the retailer could be significant even if a relatively smaller percentage of consumers would switch shops. Finally, if customers were to stay with the same retailer if their deodorant brand were be available, one would expect that retailers were to exploit such behaviour and sell only those items on which they can earn the highest margins. The market reality is different.

(609) To illustrate the retailer buyer power, Unilever quotes the reply of only one competitor on the Irish market which stated “Retailers can constrain SKU listings and breadth of distribution. Retailers work on their own timescales in Ireland, often making late confirmations of promos/listings/reversing prior commitments and have the power not to run promotion plans which could be disruptive to manufacturers" Delisting is used as a threat by retailers".

(610) The position towards retailers of a supplier which holds market shares of [0-5]*% in the overall deodorant segment and [0-5]*% in the non-male segment cannot be compared with Unilever which will hold, post merger, market shares of [70-80]*% on the overall deodorant market and [60-70]*% in the non male segment. With regards to delisting, [...]*

(611) Finally, it is not sufficient that buyer power exists prior to the merger; it must also exist and remain effective following the merger. This is because a merger of two suppliers may reduce buyer power if it thereby removes a credible alternative. In the present case, the merger will remove one of the two main suppliers considered to be the closest competitors in the market, since the market investigation did not reveal that there is a set of credible alternative suppliers which could replace Sanex in terms of product characteristics. It will also bring together four out of the six main brands in the deodorant market, which has a direct impact on the ability and incentives of retailers to take advantage of their alleged countervailing buyer power.

403 Reply to the Statement of Objections, paragraph 13.18 and Figure 13.1.

404 Questionnaire to customers on deodorants sent on 25 June 2010, question 17.

124

(612) Therefore, it can be concluded that countervailing buyer power post-merger would not be sufficient to off-set any of the potential adverse effects of the merger.

Sufficient entry unlikely to occur

(613) The market investigation indicated that entering the deodorant market – either from a neighbouring personal care market or as a new entrant - would be difficult. Even if the technology required and the investment to manufacture deodorants could be considered as "accessible", high barriers to entry exist in the form of significant advertisement and promotion (A&P) expenditure to create brand awareness and get access to shelf space from retailers.

(614) The general perception of the market is that these barriers to entry exist and are significant. As stated by a retailer "There are perceived high barriers to entry as a consequence of the dominant position of brands such as Sure and Lynx" although "Physically there is no barrier that should stop a new entry eg intro of Garnier range" .

(615) [...]* (see recital (332))

(616) In addition to the barrier to entry coming from the creation of brand awareness and access to shelf space, a potential entrant would have to take into account the likely reaction of a powerful incumbent like Unilever. [...]*, to the contrary Unilever’s increased market share, the addition of another brand like Sanex rather strengthens its ability and incentive to fight entry.

(617) Unilever submitted that the prospect of new entry is a real competitive constraint in Ireland. In order to illustrate it, Unilever mentioned a potential launch of the Playboy brand by Coty in 2010 and more vigorous competition from Gillette in the light of the wider European launch of deodorants under this brand in July 2010. These potential entries and expansions concern the male market and as such they are of limited relevance to assess potential constraints exercised by new entries in the non-male market.

(618) With respect to the recent entry of Garnier Mineral, as it only took place in January 2010, it is probably too early to conclude on the outcome of this entry. However a preliminary assessment follows.

(619) First, it is worth noting however that Unilever reacted immediately to the entry of Garnier Mineral, by launching Sure Mineral in Ireland in January 2010.

(620) [...]*

(621) While Unilever correctly points out that these measure are a sign of competitive interaction and are in no way sign of illegal conduct, they do contribute to making entry more difficult to succeed and hence less likely to be undertaken.

405 Questionnaire to customers on deodorants sent on 23 April 2010, question 61.

125

(622) In its reply to the Statement of Objections, Unilever emphasises that Garnier achieved a weighed distribution of [...]* and a market share of [0-5]*% in non-male-specific deodorants, arguing that this demonstrates that new entrants can penetrate the market.

(623) Given the costly investments and the time needed for the launch of Garnier Mineral, a share of [0-5]*% for a company as L'Oreal shows that the brand seems to have difficulties in establishing itself on the shelves. In addition, the low impact of Garnier Mineral entry has been confirmed by all the responding retailers. Indeed, one retailer argued the entry of Garnier Mineral had a "low impact in the deodorant market", the second one stated that Garnier Mineral is present in the shelves but that "it would not be part of their top seller list"and the last one notably recently refused to list 6 out of 12 proposed Garnier SKU's .

(624) Despite the significant efforts of L'Oreal which is one of the very few companies potentially able to make such a relatively large-scale entry into the market, the developments so far strongly suggest that Garnier Mineral will not be a success in the Irish market. With regards to the entry of Garnier in Ireland, L'Oreal itself explained "the market dynamics which gave Garnier reasons to believe in a successful entry and a reasonable prospect of profit have vanished…"

(625) Therefore, it is unlikely that new entry will occur in the Irish non-male-deodorant market so as to counteract adverse effects of the merger.

Overall conclusion

(626) It is therefore concluded that the notified concentration is likely to significantly impede effective competition on the market for non-male deodorants in Ireland.

IV.2.3.2.4.2. Male deodorant market

(627) The Irish male deodorant market had a total value of EUR 12.6 million in 2009. Due to the impact of the financial and economic crisis, it experienced a decline in value of [10-20]*% between 2008 and 2009. Unilever is active in the male deodorant market with its main brands Lynx and Sure for Men. In addition, Brut and Vaseline are sold on the market. Sara Lee serves the Irish male deodorant market with its brand Sanex for Men.

(628) The Parties would post-merger achieve a combined market share of [70-80]*% (Unilever [70-80]*%, Sara Lee [0-5]*%). Other competitors active in the Irish male market are Beiersdorf ([5-10]*%), Procter & Gamble ([5-10]*%), Coty ([0-5]*%), Henkel ([0-5]*%) and Revlon ([0-5]*%). In spite of this high market share, the proposed transaction will not result in a significant impediment of effective competition for several reasons.

406 Questionnaire to customers on deodorants sent on 23 April 2010, question 59.

(629) First, the increment resulting from the transaction is limited as Sara Lee had a market share of [0-5]*% in the Irish male deodorant market. Thus, the competitive landscape would not change significantly due to the merger.

(630) Second, the wide majority of retailers does not consider Sanex for Men to be a close 410 competitor to either Lynx or Sure for Men. In their response to the market investigation, they considered Coty's Adidas or Beiersdorf's Nivea for Men as close competitors to Unilever's male brands, which are more oriented towards efficacy and 411 fragrance compared to Sanex for Men's health proposition.In addition, [50-60]*% of Sanex for Men deodorants are sold as roll-ons, while roughly [90-100]*% of Lynx and 412[80-90]*% of Sure for Men deodorants are sold as aerosols.

(631) Finally, several competitors of similar or larger size compared to Sara Lee remain in the market with male brands like Gillette, Right Guard, Nivea for Men, Adidas or Mitchum for Men. Indeed, retailers active in Ireland either have not listed Sara Lee or indicated that sufficient alternatives would be available to replace it in the male deodorant 413market.

(632) It is therefore concluded that the proposed transaction would not result in a significant impediment of effective competition in the market for male deodorants in Ireland.

IV.2.3.2.5. Greece

(633) The Greek deodorant market had a total value of EUR 39 million in 2009 (EUR [...]* for the male deodorant market; EUR [...]* for the non-male deodorant market). Unilever is mainly active in Greece with its brands Axe, Dove and Rexona, while Sara Lee achieved its turnover in 2009 with the Sanex brand.

Table 27: Greece – Deodorant market shares in value – 2009 – Source: Form CO.

Unilever Sara Lee Combined Competitors

Overall [30-40]*% [5-10]*% [30- 40]*%

Beiersdorf [10-20]*%

Procter & Gamble [20-30]*%

Sarantis [10-20]*%

Male [40-50]*% [0-5]*% [40- 50]*%

Beiersdorf [10-20]*%

Procter & Gamble [5-10]*%

Sarantis [20-30]*%

Non-male [20-30]*% [5-10]*% [30- Beiersdorf [10-20]*%

410 See replies to Questionnaire to customers on deodorants sent on 23 April 2010, question 30.

411 See replies to Questionnaire to customers on deodorants sent on 23 April 2010, question 27.

412 See Annex 7.1 to Form CO.

413 See replies to Questionnaire to customers on deodorants sent on 23 April 2010 question 54.

127

Unilever claims that the survey panel is carried out at individual level in Spain and not at household level. This does not alleviate the risk of contamination since individual customers may purchase deodorants for the whole family.

135

contamination effect. Among non-male brands, the interaction indices of Sanex are the highest with Lactovit followed by Nivea, Dove, Rexona for Women and Byly.

(669) The interaction indices presented by Unilever are a comparison of the actual gross consumer switching of the brand (here, Sanex for Women) with the "expected" switching based on the proportions of the competing brands' market shares in the total deodorant category (combining male and non-male).A high interaction index alone therefore does not necessarily suggest substantial actual switching from and to small brands. When the market share is taken into account, the actual consumer switching as observed in the panels is significantly lower than the interaction index alone would suggest.

(670) Therefore, while interaction indices provide some insights about the closeness of brands, the gross switching figure from and to a brand provides a more illustrative picture about the actual switching patterns of consumers. Upon request, the Parties have submitted the underlying data used by GFK/Europanel containing the actual switching figures. Based on the same data as used for interaction index, Figure 25 shows in descending order for each brand their share of total gross switching to and from Sanex for Women in 2009. Consistent with the method used by GFK/Europanel, the total gross switching is compared with the proportions of the competing brands' market shares in the total deodorant category ("value of competitors").

Figure 25: Gross switching for Sanex for Women in Spain

[...]

(671) The main interactions of Sanex for Women can be found with all other brands and the private label category. However, the category "other brands" includes several small brands aggregated together (such as Palmolive, Fa, Coty, Kinesia, Heno de Pravia) and it is doubtful that this high interaction is representative of significant closeness between Sanex and one of these individual brands. The main interaction with individual brands is with Rexona and Dove (and Axe, although it is less than Axe's market share suggests). However the interaction with Axe indicates that the contamination effect is likely to be an important factor in the data, in particular given the fact that according to TNS Worldpanel data only [0-5]*% of Spanish females use male deodorant brands.

(672) In general, and with the exception of Axe (for which gross switching is smaller than the proportionate market share) and Sanex for Men (for which it is the opposite), gross switching of Sanex for Women is relatively proportionate to the market shares of the competitors. This shows that market shares can be considered as a reasonable proxy to estimate the relative share of customers switching from/to Sanex for Women to/from other brands.

441 The market shares are calculated on the basis of the brands' sales within the panel. This may to some extent vary from the market shares based on Nielsen data, but in general provides a rather consistent picture with those.

442 Form CO, Table 6.8 at page 54, and more recent and extended figures from additional submission of Unilever of 27 July 2010.

136

(673) Of all the Sanex for Women switchers that switched to or from a non-Sanex brand (either male or non-male) [20-30]*% switched to or from a Unilever brand. This figure confirm that the risk of losing sales to Unilever appear to be an important constraint on Sanex prior to the merger. These switches would not constrain Sanex after the transaction since the switches would no longer be considered lost to the merged entity.

(674) In order to assess whether Sanex represents a competitive constraint on Unilever's brands, Unilever has been asked to provide gross switching data for Dove and Rexona for Women respectively. As show in Figure 26 Sanex for Women is the main interacting brand with Dove and Rexona for Women, with the exception of aggregated small brands which cannot be considered as individually representative.

Figure 26: Gross switching for Dove in Spain

[...]

Figure 27: Gross switching for Rexona for Women in Spain

[...]

Assessment on the basis of format and product attributes

(675) A detailed analysis of different attributes associated by customers of non-male deodorants in Spain shows [reference to parties’ internal documents].

(676) In addition, it has been tried to identify in more detail in which sub-segments Sanex is positioned and where a significant overlap with Unilever's brands exists. Moreover, the combination of price points and product characteristics could complement the qualitative analysis of closeness between Unilever's brands and Sanex.The sub-segmentation is based on a classification provided by Unilever.

(677) The analysis shows that Sanex has a sizeable position in the skin care, AP, efficacy and the no white mark sub-segments. It has significant interaction with Unilever in all these sub-segments. Sanex competes in all the sub-segments where Rexona is present (skin care, fragrance, efficacy, and no white marks). The same is true with Dove (with the exception of the efficacy sub-segment) which in addition is present in the hair minimising sub-segment. Nivea is also present in these four sub-segments. Lactovit is also only present in two (skin care and AP). Fa only competes in the fragrance sub-segment and Byly is not present in the no white marks sub-segment. Conversely, the Parties' brands are hardly active in the fragrance and the hair minimising sub-segments.

443 [...]

137

See replies from [retailer] and [retailer].

See replies from [retailer].

145

(706) The key issue is whether the growth of private label products constraints the merging parties' brands. A majority of retailers considers that the increase of private labels primarily impacted alternative brands. One retailer explains that the rise of private label products has no significant effect on the main brands in their assortment, which are Sanex, Rexona, Axe and Nivea, but rather on second-tier brands such as Lactovit, Heno de Pravia, Tulipa Negro and Byly. A second retailer does not think that its future private label products could easily replace Sanex or Unilever’s brands. A third distributor is of the opinion that the growth has no major effect on branded products and seems to consider that Byly’s or Puig’s brands do not play in the same league as Unilever’s or Sara Lee’s. That retailer does not carry the Lactovit brand and argues that Byly plays a limited role in several segments of the non-male market, such as sprays or anti-perspirant deodorants. Only one of the retailers holds that all brands (including Sanex) have lost shares as a result of private label products’ growth (its share of private labels – [5-10]*% - is much smaller than [retailer's] or [retailer's].) .

(707) That specific trend in the Spanish market is particularly evident in Figure 33, which shows the relative share of private label and suppliers’ brands in the Spanish non-male market since early 2007. It is clear from the graph in Figure 33 that Unilever and Sara Lee maintained their relative position despite the growth of private label products, whilst alternative brands have significantly lost market share. If these trends continue, the Spanish market would be evolving towards the constitution of two major blocks, namely Unilever /Sara Lee on the one hand and private label products on the other hand.

Figure 33: Relative share of private label products and suppliers’ brands in the Spanish non-male market since 2007

[...]

(708) On 15 September 2010, Unilever provided upon request the gross switching figures related to private label products as well as Nivea, Byly and Lactovit, for the period

See non-confidential minutes from a conference call with [retailer] on 1 October 2010. See also reply from [retailer] questionnaire to customers on deodorants sent on 14 September 210, question 4.

See reply from [retailer] to questionnaire to customers deodorants sent on 14 September 2010, question 6 “We would not be able to replace any of them. Customers who like a brand will not change easily, despite small differences on price”

See reply from [retailer] to questionnaire to customers on deodorants sent on 14 September 2010, question 4. "Ninguna marca ha perdido venta como consecuencia da un increment de la marca privada porque las marcas de fabricante han crecido algo mas que las marcas privadas ».

See reply from [retailer] to questionnaire to customers on deodorants sent on 14 September 2010, question 8.

See reply from [retailer] to questionnaire to customers on deodorants sent on 14 September 2010, question 9.

See reply from [retailer] to questionnaire to customers deodorants sent on 14 September 2010, question 8.

Competitors interaction Spain Q2.

146

between March 2009 and 2010. Consistent with the method used by GFK/Europanel, the total gross switching is compared with the proportions of the competing brands' market shares in the total deodorant category ("value of competitors").

(709) Those switching data show that the level of switching to and from private label products for Sanex is smaller than the value of its market share. This indicates that customers of private label products switch less to and from Sanex than they should do taking into account's Sanex's market position. The level of switching to and from Axe is also smaller than its market share would suggest. Those data show that the main brand in the male market (Axe) and the main brands in the non-male market (Sanex) are less constrained by private label products than rival brands. Indeed Byly and Lactovit account for a higher degree of gross switching to and from private label products than their market shares would suggest.

Figure 34: Gross switching for Private labels in Spain [...]

(710) This trend has not been reversed in 2010 according to Nielsen figures. Table 33 shows the evolution of several non-male brands in Spain during the period between August 2009 and August 2010. It confirms that Sanex and Unilever brands are maintaining their level of sales despite a (more moderate) growth of private label products. Nivea, Byly, Briseis, Puig and Fa are sharply declining.

Table 33: evolution of several non-male brands in Spain between August 2009 and August 2010. August-Sanex Dove Rexona PL Nivea Byly Briseis Puig Fa 2009- August 2010

+[0-5]*% +[0-5]*% -[0-5]*% +[10-20]*% -[5-10]*% -[0-5]*% -[5-10]*% -[10-20]*% -[20-30]*%Evolution of sales (in value in %)

[One specific retailer] represents the large majority of private label products ‘penetration and growth.

Unilever explains in detailthat growth of private label products in Spain is not confined to [the specific retailer] and that even if it were, the evidence of [the specific retailer]’s ability to increase its private label range indicate that other retailers also have the same possibility to follow the same strategy, if expedient.

(712) With respect to the first point, Unilever indicated that other discounters such as [retailer] and [retailer] have private label products sales of deodorants comparable to [the specific retailer]. Whereas [40-50]*% of [the specific retailer]’s value sales of deodorants are private label products, the same figure is [20-30]*% for [retailer] and [20-30]*% for [retailer]. Those three retailers represent one third of Unilever’s sales of deodorants. [retailer] increased its private label products sales in value from [10-20]*% to [10-20]*%. According to Unilever, retailers continue to seek enlarging their private label

Reply to the Statement of Objections, paragraphs 10-18 to 10-25

147

range, as evidenced by replies made by [retailer] and [retailer] in the market investigation.

(713) Even if private label products' growth were considered to be a phenomenon specific to [the specific retailer], Unilever submitted this would still have market-wide relevance. Unilever described Spanish retailers as benchmarking their prices against [the specific retailer], selling products available at [the specific retailer] at lower prices than those not listed by [the specific retailer] and comparing their shopping carts with the same basket bought at [the specific retailer].

(714) As to the first issue, it is correct that “discounters” such as [retailer] and [retailer] have a significant level of private label deodorants available on their shelves and that their objective is to increase the penetration of private label products. At this stage however, the wide gap between [the specific retailer] and its retail competitors is not on the verge of being filled. As indicated by Unilever, [10-20]*% of deodorants sales in [retailer] were achieved in 2009 with private label products, [5-10]*% in [retailer], [0-5]*% in [retailer] (these three retailers accounting for [30-40]*% of Unilever’s sales of deodorants). This hardly compares with the [40-50]*% achieved in [the specific retailer] in 2009 and [50-60]*% for the first six months of 2010 .

(715) The development of private label products is a priority for [the specific retailer]. [the specific retailer] accounts for the wide majority of sales of non male private label deodorants in Spain ([70-80]*%). It sales of all deodorants (including branded and private label products) is much smaller ([10-20]*%). In 2009, [the specific retailer] accounted for [60-70]*%of the growth of sales of private label products despite its already high starting base. Even if private label products' growth is not strictly confined to [the specific retailer], it has not taken off in other retail chains. Thus, the growth observed in the sales of private label products is mainly the result of [the specific retailer]'s expansion in the deodorant category.

(716) What is more, competing retail chains did not confirm in the market investigation that they intend to follow the same path as [the specific retailer]. One retailer explains that “We are neither able to open 100 shops per year nor our claim for our customers is PL.

Reply to the Statement of Objections, figure 10-1.

Form CO, Table 6-35.

See reply from [retailer] to questionnaire to customers on deodorants sent on 14 September 2010, question 1.