JUDGMENT OF THE GENERAL COURT (Sixth Chamber)

2 July 2025 ( * )

( EU trade mark – Opposition proceedings – Application for EU figurative mark JOLi – Earlier national word mark JOLIFE – Relative ground for refusal – No likelihood of confusion – Article 8(1)(b) of Regulation (EU) 2017/1001 )

In Case T‑473/24,

Villafrut Srl, established in Oppeano (Italy), represented by A. Rizzoli, lawyer,

applicant,

v

European Union Intellectual Property Office (EUIPO), represented by C. Bovar and D. Stoyanova-Valchanova, acting as Agents,

defendant,

the other party to the proceedings before the Board of Appeal of EUIPO being

Venture Fruit Global Ltd, established in Auckland (New Zealand),

THE GENERAL COURT (Sixth Chamber),

composed of M.J. Costeira, President, M. Kancheva and U. Öberg (Rapporteur), Judges,

Registrar: V. Di Bucci,

having regard to the written part of the procedure,

having regard to the fact that no request for a hearing was submitted by the parties within three weeks after service of notification of the close of the written part of the procedure, and having decided to rule on the action without an oral part of the procedure, pursuant to Article 106(3) of the Rules of Procedure of the General Court,

gives the following

Judgment

1 By its action under Article 263 TFEU, the applicant, Villafrut Srl, seeks annulment and alteration of the decision of the Fifth Board of Appeal of the European Union Intellectual Property Office (EUIPO) of 23 July 2024 (Case R 2468/2023-5) (‘the contested decision’).

Background to the dispute

2 On 22 July 2021, Venture Fruit Global Ltd filed an application with EUIPO to register the following figurative sign as an EU trade mark:

3 The trade mark applied for designated goods falling, after the limitation entered into on 14 December 2021, in the course of the proceedings before EUIPO, in Class 31 within the meaning of the Nice Agreement Concerning the International Classification of Goods and Services for the Purposes of the Registration of Marks of 15 June 1957, as revised and amended, and corresponding, in particular, to the following description: ‘Agricultural and horticultural products; fresh fruits and vegetables including pip-fruit’.

4 On 16 November 2021, the applicant filed an opposition to the registration of the mark applied for in respect of the goods referred to in paragraph 3 above.

5 The opposition was based, inter alia, on the earlier Italian word mark JOLIFE, registered on 22 April 2003 under No 889 342, designating goods in, inter alia, Class 31 and corresponding to the following description: ‘Fresh fruit and vegetables’.

6 The ground relied on in support of the opposition was the existence of a likelihood of confusion within the meaning of Article 8(1)(b) of Regulation (EU) 2017/1001 of the European Parliament and of the Council of 14 June 2017 on the European Union trade mark (OJ 2017 L 154, p. 1).

7 Following the request made by Venture Fruit Global, EUIPO invited the applicant to provide evidence of genuine use of the earlier mark cited in opposition. The applicant complied with the request within the prescribed period. The Opposition Division found that genuine use of the earlier mark had been proved only in respect of ‘fresh fruit and vegetables’ in Class 31.

8 On 6 November 2023, the Opposition Division partially upheld the opposition and consequently rejected the application for registration in respect of part of the goods designated by the mark applied for, namely the goods referred to in paragraph 3 above, on the ground that there was a likelihood of confusion with the goods in respect of which genuine use of the earlier mark had been proved, namely ‘fresh fruit and vegetables’ in Class 31.

9 On 14 December 2023, Venture Fruit Global filed a notice of appeal with EUIPO against the decision of the Opposition Division.

10 By the contested decision, the Board of Appeal upheld the appeal on the ground that there was no likelihood of confusion between the marks at issue in the mind of the relevant public, within the meaning of Article 8(1)(b) of Regulation 2017/1001, and partially annulled the decision of the Opposition Division.

Forms of order sought

11 The applicant claims that the Court should:

– annul the contested decision;

– uphold the opposition in respect of all the goods designated by the trade mark applied for and dismiss the trade mark application in its entirety;

– order EUIPO to pay the costs, including those relating to the earlier proceedings before EUIPO.

12 EUIPO contends that the Court should:

– dismiss the action;

– order the applicant to pay the costs, in the event a hearing is convened.

Law

13 In support of its action, the applicant relies on a single plea in law alleging infringement of Article 8(1)(b) of Regulation 2017/1001.

14 The applicant claims that the Board of Appeal wrongly concluded that there was no likelihood of confusion. First, the applicant submits that the finding of the Board of Appeal that the level of attention of the public varied from low to average for the goods at issue is erroneous. Secondly, the applicant contests the visual and phonetic comparison of the signs at issue, in that the Board of Appeal considered that those signs were short and had a low degree of visual similarity and ‘at most’ an average degree of phonetic similarity. Thirdly, the applicant submits that the overall assessment of the likelihood of confusion is wrong, since the differences between the signs at issue cannot outweigh their similarities.

15 EUIPO disputes the applicant’s arguments.

16 Under Article 8(1)(b) of Regulation 2017/1001, upon opposition by the proprietor of an earlier trade mark, the trade mark applied for is to be refused registration where, because of its identity with or similarity to an earlier trade mark and because of the identity or similarity of the goods or services which the two trade marks designate, there is a likelihood of confusion on the part of the public in the territory in which the earlier trade mark is protected. The likelihood of confusion includes the likelihood of association with the earlier mark.

17 According to the case-law, a likelihood of confusion is the likelihood that the public might believe that the goods or services at issue come from the same undertaking or, where appropriate, from economically linked undertakings. The likelihood of confusion must be assessed globally, according to the perception that the relevant public has of the signs and the goods or services at issue, and taking into account all factors relevant to the circumstances of the case, in particular the interdependence of the similarity between the signs and the similarity between the goods or services covered (judgment of 4 July 2019, FTI Touristik v EUIPO , C‑99/18 P, EU:C:2019:565, paragraph 20).

18 For the purposes of applying Article 8(1)(b) of Regulation 2017/1001, a likelihood of confusion presupposes both that the marks at issue are identical or similar and that the goods or services which they cover are identical or similar. Those conditions are cumulative (see judgment of 22 January 2009, Commercy v OHIM – easyGroup IP Licensing (easyHotel) , T‑316/07, EU:T:2009:14, paragraph 42 and the case-law cited).

The relevant public

19 The Board of Appeal took the view that the goods concerned were aimed at both the general public and professionals and that it was appropriate to take into account the degree of attention of consumers belonging to the general public. In its view, the degree of attention of the latter varied from low, in the case of low-priced products intended for mass consumption, to average, in the case of higher-priced foodstuffs and beverages.

20 The Board of Appeal then considered that, in view of the earlier national trade mark on which the opposition was based, the relevant territory was that of Italy.

21 The applicant claims, in essence, that, when assessing the level of attention of the relevant public, the Board of Appeal erroneously included products which were not at issue in the present case and, consequently, wrongly overestimated that public’s level of attention as being average. It claims that the reference in the contested decision to foodstuffs is very broad and imprecise because, in particular, the price and frequency of purchase of foodstuffs can vary the level of attention of the relevant public, which is not the case with ‘fruit and vegetables’ in Class 31. In addition, the reference to beverages is erroneous, as they are not covered by the marks at issue. Consequently, as regards the goods at issue, namely ‘fresh fruit and vegetables’, the applicant considers that the level of attention of the relevant public is low.

22 EUIPO disputes the applicant’s arguments.

23 According to settled case-law, for the purposes of the overall assessment of the likelihood of confusion, the average consumer of the category of goods concerned is deemed to be reasonably well informed and reasonably observant and circumspect. In that regard, it should also be borne in mind that the average consumer’s level of attention is likely to vary according to the category of goods or services in question (see judgment of 22 June 1999, Lloyd Schuhfabrik Meyer , C‑342/97, EU:C:1999:323, paragraph 26 and the case-law cited).

24 In the present case, it should be noted, first, that the reference to foodstuffs is relevant, given that fruit and vegetables form part of that generic category and, secondly, that beverages were mentioned only by way of example. As EUIPO points out, that reference is part of a general reasoning and, in any event, only ‘fresh fruit and vegetables’, expressly referred to in paragraph 68 of the contested decision, were taken into account in the overall assessment of the likelihood of confusion.

25 It should also be noted, as EUIPO points out, that, for ‘fresh fruit and vegetables’, the level of attention of the relevant public varies, in the case-law, from low (see, to that effect, judgment of 15 October 2018, Apple and Pear Australia and Star Fruits Diffusion v EUIPO – Pink Lady America (WILD PINK) , T‑164/17, not published, EU:T:2018:678, paragraph 51) to average (see, to that effect, judgment of 13 November 2024, Chiquita Brands v EUIPO – Compagnie financière de participation (Representation of a blue and yellow oval) , T‑426/23, not published, EU:T:2024:807, paragraph 24).

26 It follows that the Board of Appeal did not make an error of assessment in considering that the level of attention of the relevant public varied from low to average.

The c omparison of the goods at issue

27 The Board of Appeal found that the goods designated by the mark applied for, included in Class 31, were identical to the goods in the same class covered by the earlier mark, since the general category of ‘fresh fruit and vegetables’ covered by the earlier mark either included the goods of the mark applied for or was included in them.

28 Those assessments of the Board of Appeal are not contested by the applicant and there is nothing in the case file to call them into question.

The comp arison of the signs at issue

29 As regards the mark applied for, the Board of Appeal found, in paragraph 38 of the contested decision, that it consisted of the word element ‘joli’, written in stylised black bold type, and a purely decorative curve, placed below that word element. The Board of Appeal then found, in paragraph 39 of the contested decision, that the word element ‘joli’ was devoid of meaning for a significant proportion of the Italian-speaking public, with a below-average level of attention, who did not recognise the French word ‘joli’.

30 With regard to the earlier mark, the Board of Appeal first held, in paragraph 40 of the contested decision, that that mark consisted of a single six-letter word element, the element ‘jolife’. The Board of Appeal then noted, in paragraphs 41 to 45 of the contested decision, that the earlier mark would not be broken down by the relevant public. In its view, the letters ‘jo’ had no meaning for the targeted consumer and, since the letter ‘j’ was not part of the Italian alphabet, the sign as a whole would be perceived as a foreign word by a significant part of the relevant public in Italy. The Board of Appeal concluded that, irrespective of whether the Italian-speaking public recognised the English word ‘life’, it would perceive the earlier mark as a whole. Finally, in paragraph 46 of the contested decision, the Board of Appeal added that there was no visual indication enabling the relevant consumer to break down the earlier mark into two terms, ‘jo’ and ‘life’ or ‘joli’ and ‘fe’.

31 Those assessments of the Board of Appeal are not contested by the applicant and there is nothing in the case file to call them into question. By contrast, the applicant contests the findings of the Board of Appeal relating to the visual and phonetic comparison of the marks at issue.

32 The overall assessment of the likelihood of confusion must, as regards the visual, phonetic or conceptual similarity of the signs at issue, be based on the overall impression produced by them, taking into account, in particular, their distinctive and dominant elements. The perception of the marks by the average consumer of the goods or services in question plays a decisive role in the global assessment of that likelihood of confusion. In that regard, the average consumer normally perceives a mark as a whole and does not engage in an analysis of its various details (see judgments of 12 June 2007, OHIM v Shaker , C‑334/05 P, EU:C:2007:333, paragraph 35 and the case-law cited, and of 4 March 2020, EUIPO v Equivalenza Manufactory , C‑328/18 P, EU:C:2020:156, paragraph 58 and the case-law cited).

The visual comparison

33 The Board of Appeal found that the marks at issue coincided as regards the sequence of the letters ‘j’, ‘o’, ‘l’ and ‘i’ and differed as regards their ending and the stylisation and figurative element of the mark applied for. According to the Board of Appeal, as the signs at issue were relatively short, the difference linked to the additional syllable ‘fe’, placed at the end of the earlier mark, would be clearly perceptible to the relevant public. The Board of Appeal considered that, for that type of sign, the ending was as important as the beginning. It therefore concluded that the marks at issue had a low degree of visual similarity.

34 The applicant contests that analysis, arguing that the signs at issue cannot be regarded as short signs and that, even if they were, the difference between their final parts would have only a limited visual impact. According to the applicant, consumers pay more attention to the beginning of a trade mark than to its end. That is why the four identical letters in the same position at the beginning of the signs attract consumers’ attention more than the last syllable ‘fe’ of the earlier mark. Furthermore, the applicant submits that the Board of Appeal erred in considering that the last two letters of the earlier mark, ‘fe’, constituted 50% of the word element of the mark applied for. Finally, the applicant points out, citing the Court’s case-law, that the figurative element of the sign concerned plays a secondary role in relation to its word element. Accordingly, in the applicant’s view, the Board of Appeal should have found that there was at least an average degree of visual similarity.

35 EUIPO disputes the applicant’s arguments.

36 In that regard, it should be noted that the marks at issue both contain the sequence of the letters ‘j’, ‘o’, ‘l’ and ‘i’, and that ‘joli’ constitutes the sole word element of the mark applied for, as well as the first four letters of the earlier mark JOLIFE, which implies a certain degree of similarity between those marks.

37 Thereby, it should also be noted that the signs at issue have a number of visual differences.

38 In the first place, although the sequence of letters ‘j’, ‘o’, ‘l’ and ‘i’ is certainly common to the signs at issue, it must be said that its specific graphic representation in the sign applied for attracts the public’s attention in some way, having regard to its particular typeface. In addition, the marks at issue are visually distinguished by the curve below the word element of the mark applied for and by the unusual use of upper-case letters for the ‘JOL’ element and a lower-case letter for the ‘i’.

39 It is true that, where a trade mark is composed of word and figurative elements, the former are, in principle, more distinctive than the latter, because the average consumer will more easily refer to the goods in question by quoting the name of that trade mark than by describing its figurative element (judgment of 14 July 2005, Wassen International v OHIM – Stroschein Gesundkost (SELENIUM-ACE) , T‑312/03, EU:T:2005:289, paragraph 37). Nevertheless, a figurative mark, by definition, contains a particular figurative element which, unless it is wholly insignificant, must, even if it is dominated by a word element, as in the present case, be taken into account in the overall assessment of the similarity of the signs at issue (see judgment of 11 December 2024, López-Ibor Aliño v EUIPO – Dimensión Estratégica Quality Research (LOPEZ-IBOR ABOGADOS) , T‑672/22, not published, EU:T:2024:892, paragraph 60 and the case-law cited). Thus, it is apparent from an analysis of the mark applied for that, although they are not co-dominant, the figurative elements of that mark, consisting essentially of letters with a slight stylisation, the existence of a curve under the word element thereof and the unconventional use of lower-case and upper-case letters, are not, as the Board of Appeal recognised in the present case, negligible.

40 In the second place, the earlier mark contains two final letters in addition to those of the mark applied for, namely the letters ‘fe’.

41 In that regard, it should be noted, as does EUIPO, that the Board of Appeal did not describe the signs at issue as short signs, but as ‘quite short’ or ‘relatively short’ signs. Nor did it apply the case-law on short signs, according to which the relevant public is likely to perceive more clearly the differences between two signs of that kind (see, to that effect, judgment of 3 December 2014, Max Mara Fashion Group v OHIM – Mackays Stores (M & Co.) , T‑272/13, not published, EU:T:2014:1020, paragraph 47 and the case-law cited), but considered, in the present case, that the elements at the end of the signs at issue were just as important as their elements at the beginning and their central elements.

42 The Court has recognised that, although the relevant public generally pays greater attention to the beginning of a trade mark than to the end (judgment of 25 March 2009, L’Oréal v OHIM – Spa Monopole (SPA THERAPY) , T‑109/07, EU:T:2009:81, paragraph 30), in respect of word marks which are relatively short, such as those in the present case, the elements at the beginning and end of the sign were as important as the central elements (see, to that effect, judgments of 20 April 2005, Krüger v OHIM – Calpis (CALPICO) , T‑273/02, EU:T:2005:134, paragraph 39; of 15 March 2012, Cadila Healthcare v OHIM – Novartis (ZYDUS) , T‑288/08, not published, EU:T:2012:124, paragraph 52; and of 28 April 2021, FCA Italy v EUIPO – Bettag (Pandem) , T‑191/20, not published, EU:T:2021:226, paragraph 39).

43 In the present case, it is true that the first four letters of the marks at issue are identical and make up the entire word element of the mark applied for. However, in accordance with the case-law, it cannot be ruled out that signs may be considered to be visually similar to a low degree, even though they contain a sequence of identical letters (see, to that effect, judgments of 11 December 2013, Eckes-Granini v OHIM – Panini (PANINI) , T‑487/12, not published, EU:T:2013:637, paragraph 42; of 27 February 2019, Aytekin v EUIPO – Dienne Salotti (Dienne) , T‑107/18, not published, EU:T:2019:114, paragraph 48; and of 11 October 2023, Pascoe pharmazeutische Präparate v EUIPO – Novartis Pharma (PASCELMO) , T‑435/22, not published, EU:T:2023:610, paragraphs 50 to 54).

44 Consequently, the Board of Appeal was right to consider that the group of letters ‘fe’ had a significant impact on the perception of the signs at issue, since that final element of dissimilarity is as important as the elements at the beginning, so that it contributes to mitigating, in the context of an overall impression, the similarity produced by the initial letters ‘joli’.

45 Therefore, contrary to the applicant’s assertion, in the present case the beginning of the signs at issue will not necessarily have a greater effect on consumer perception than the ending of the earlier mark, and the latter will be perceptible by the relevant public.

46 As regards the applicant’s argument that the last two letters of the earlier mark do not constitute 50% of the word element of the mark applied for, it must be noted that that mark consists of four letters and that two letters represent half of four. The assessment of the Board of Appeal that the last two letters of the earlier mark constitute 50% of the four letters of the mark applied for must therefore be confirmed.

47 Consequently, the visual similarity of the signs at issue is limited solely to the presence of the common letter sequence ‘j’, ‘o’, ‘l’ and ‘i’.

48 The assessment of the similarity between two trade marks cannot, in principle, be limited to taking into consideration only one component of a complex trade mark consisting of a sequence of common letters and comparing it with another trade mark. It is only in circumstances where the overall impression conveyed to the relevant public by a complex trade mark is dominated by one or more of its components and all the other components of that mark are negligible that the assessment of similarity may, by way of exception, be made solely on the basis of the dominant element (see, to that effect, judgment of 12 June 2007, OHIM v Shaker , C‑334/05 P, EU:C:2007:333, paragraphs 35 and 41 and the case-law cited).

49 When comparing the marks at issue, account must also be taken of the figurative elements of the trade mark applied for, which, as is apparent from paragraph 39 above, are not insignificant (see, to that effect, judgment of 11 December 2024, LOPEZ-IBOR ABOGADOS , T‑672/22, not published, EU:T:2024:892, paragraphs 87 and 88).

50 Accordingly, the differences between the signs at issue, due to the figurative elements of the mark applied for and the presence of additional letters in the earlier mark, namely those making up the syllable ‘fe’, contribute, to a not inconsiderable extent, to visually distinguishing the signs at issue.

51 In the light of the foregoing considerations, the Board of Appeal’s conclusion that the signs at issue, taken as a whole, show little visual similarity must be upheld.

The phonetic comparison

52 The Board of Appeal found that the marks at issue coincided as regards the sound of the sequence of letters ‘j’, ‘o’, ‘l’ and ‘i’ and differed in the number of syllables and the rhythm of pronunciation. According to the Board of Appeal, as the signs at issue are relatively short, the difference in ending linked to the additional syllable ‘fe’ in the earlier mark would be clearly perceptible. Accordingly, the Board of Appeal concluded that the marks at issue were phonetically similar to at most an average degree.

53 The applicant disputes that analysis and submits that the difference in the final part ‘fe’ of the earlier mark is not such as to compensate for the identical sound of the first four letters ‘j’, ‘o’, ‘l’ and ‘i’ of the signs at issue. Furthermore, the marks are devoid of conceptual meaning and are pronounced with the same rhythm and intonation, according to the Italian rules of pronunciation, which produces a similar overall phonetic impression. According to the applicant, the Board of Appeal should therefore have found that there was at least an average degree of phonetic similarity.

54 EUIPO disputes the applicant’s arguments.

55 In that regard, it should first be noted that the earlier mark will be pronounced in three syllables, namely ‘jo’, ‘li’ and ‘fe’, whereas the mark applied for will be pronounced in two syllables, namely ‘jo’ and ‘li’, so that they differ in length.

56 While it is true that there is a partial phonetic identity between the signs at issue resulting from the presence of the sequence of common letters ‘j’, ‘o’, ‘l’ and ‘i’, the Board of Appeal was right to conclude that there is at most an average degree of phonetic similarity between the signs at issue, taking into account the additional syllable in the earlier mark.

57 That conclusion is not called into question by the applicant’s argument that the Board of Appeal should have recognised a degree of phonetic similarity which was at least average, and not at most average. In that regard, it must be noted, as does EUIPO, that the applicant does not substantiate its argument. Moreover, as EUIPO also points out, the applicant acknowledges, as did the Board of Appeal, that the degree of phonetic similarity between the signs at issue to be taken into account in assessing the likelihood of confusion is average.

The conceptual comparison

58 The Board of Appeal found that none of the signs at issue had any meaning for the relevant public. It therefore concluded that the conceptual comparison was neutral. By contrast, for the section of the public that understood the French word ‘joli’, it considered that the signs were conceptually different.

59 Those assessments of the Board of Appeal are not contested by the applicant and there is nothing in the case file to call them into question.

The intrinsic distinctiveness of the earlier mark

60 In paragraph 58 of the contested decision, the Board of Appeal concluded that the earlier mark had normal distinctive character.

61 That assessment by the Board of Appeal is not contested by the applicant and there is nothing in the case file to call it into question.

The overall assessment of the likelihood of confusion

62 The Board of Appeal considered, after taking into account the degree of attention of the relevant public, the identity of the goods concerned, the absence of any possibility of decomposing the earlier mark, the low degree of visual similarity of the signs at issue, their at most average degree of phonetic similarity, their neutral conceptual comparison or their lack of conceptual similarity, that the overall impression of the signs at issue was clearly different. Accordingly, there was no likelihood of confusion.

63 The applicant contests that assessment, in that it is based on erroneous premises, according to which the signs at issue are short, visually similar to a low degree and phonetically similar to at most an average degree.

64 In that regard, the applicant reiterates its arguments concerning the visual and phonetic comparison of the signs at issue, arguing that the differences in the ending of the earlier mark and in the figurative element of the sign applied for cannot outweigh the similarities between the signs at issue. In addition, it stated that, even if the relevant public displayed an average level of attention, a likelihood of confusion could not be ruled out, owing to the principle of imperfect memory retained by consumers. Finally, the Board of Appeal erred in considering that the earlier mark was ‘not recognisable’ within the mark applied for, because this would mean that there would never be a likelihood of confusion where an earlier mark was longer than the mark applied for or where the marks did not consist of identical letters.

65 EUIPO disputes the applicant’s arguments.

66 It should be borne in mind that, according to the Court of Justice’s settled case-law, a likelihood of confusion on the part of the public must be assessed globally, taking into account all factors relevant to the circumstances of the case (see judgment of 4 March 2020, EUIPO v Equivalenza Manufactory , C‑328/18 P, EU:C:2020:156, paragraph 57 and the case-law cited).

67 The overall assessment of the likelihood of confusion involves a degree of interdependence between the factors taken into account and, in particular, between the similarity of the marks at issue and that of the goods or services designated by them. Thus, a low degree of similarity between the goods or services designated by the trade marks may be offset by a high degree of similarity between those trade marks, and vice versa (judgments of 29 September 1998, Canon , C‑39/97, EU:C:1998:442, paragraph 17, and of 14 December 2006, Mast-Jägermeister v OHIM – Licorera Zacapaneca (VENADO with frame and others) , T‑81/03, T‑82/03 and T‑103/03, EU:T:2006:397, paragraph 74).

68 In the present case, it follows from the foregoing considerations that the relevant public consists of the general public and professional consumers, with a low to average level of attention, and that the signs at issue are visually similar to a low degree and phonetically similar to at most an average degree, while the conceptual comparison remains neutral for the predominant part of the relevant public and results in a difference for the part of that public which understands the French word ‘joli’. The similarities between the signs are not particularly significant in the present case. Despite the presence of the sequence of identical letters ‘j’, ‘o’, ‘l’ and ‘i’ at the beginning of the signs at issue, the overall impression produced by the mark applied for is distinct from that produced by the earlier mark, owing to the presence of the figurative elements of the mark applied for and the two additional letters of the earlier mark.

69 The circumstances of the present case are therefore such that the relevant public will clearly distinguish the signs at issue, even if they have a certain degree of similarity, due to the presence of the sequence of common letters ‘j’, ‘o’, ‘l’ and ‘i’, and even if the goods at issue are identical and the earlier mark has an average degree of distinctiveness.

70 In the present case, therefore, the Board of Appeal was right to hold that there was no likelihood of confusion.

71 That conclusion cannot be called into question by the applicant’s reliance on the principle of imperfect recollection in order to establish that a likelihood of confusion cannot be excluded. It is true that the average consumer only rarely has the chance to make a direct comparison between the different marks but must place his or her trust in his or her imperfect recollection of them (judgment of 22 June 1999, Lloyd Schuhfabrik Meyer , C‑342/97, EU:C:1999:323, paragraph 26) and that that principle is applicable even to a public displaying an average level of attention (see, to that effect, judgment of 14 February 2019, Torro Entertainment v EUIPO – Grupo Osborne (TORRO Grande MEAT IN STYLE) , T‑63/18, not published, EU:T:2019:89, paragraph 66).

72 Nevertheless, in the present case, the visual and phonetic differences between the signs at issue, such as those based on the ending of the earlier mark and the pronunciation of its last syllable, will not go unnoticed. Consequently, the relevant public, with a low to average level of attention, will perceive those differences, even if it only retains an imperfect image of those signs, which have a different overall impression.

73 Finally, as regards the applicant’s argument that the Board of Appeal considered that there was no likelihood of confusion because the earlier mark was ‘not recognisable’ within the mark applied for, it should be noted that the Board of Appeal took into account all the relevant factors and their interdependence in analysing the likelihood of confusion, and not only the fact that the earlier mark was not contained within the mark applied for. It follows that, contrary to the applicant’s claim, the Board of Appeal did not consider that there was no likelihood of confusion where the earlier mark was longer than the mark applied for or where those marks were not composed of the same letters. The Board of Appeal simply noted that, as the earlier mark could not be broken down into two words, ‘joli’ and ‘fe’, and was regarded as an indivisible whole, the relevant public would have a different overall impression of the marks at issue.

74 Furthermore, the examples drawn from the General Court’s case-law and EUIPO’s decision-making practice, relied on by the applicant, relating to the recognition of the existence of a likelihood of confusion in cases where the earlier mark is longer than the mark applied for or is not composed of the same letters, placed in the same order, as the latter, are not relevant to the present case.

75 First, as EUIPO submits, with regard to the judgments of 29 March 2023, Machková v EUIPO – Aceites Almenara (ALMARA SOAP) (T‑436/22, not published, EU:T:2023:167), of 12 July 2023, Nurel v EUIPO – FKuR Property (Terylene) (T‑325/22, not published, EU:T:2023:397), and of 28 February 2024, Drinks Prod v EUIPO – Wolff and Illg (IGISAN) (T‑164/23, not published, EU:T:2024:132), the circumstances are different from the present case, in that the signs were not considered to be relatively short and share the same beginning and ending.

76 Secondly, as regards the judgment of 29 November 2023, Claro v EUIPO – Claranet Europe (Claro) (T‑661/22, not published, EU:T:2023:762), the earlier mark was broken down into two terms, ‘clara’ and ‘net’, and the Court recognised that the term ‘clara’ was the more distinctive. Therefore, unlike in the present case, that term and the term constituting the contested sign in the case giving rise to that judgment were of the same length and differed only in their last letter. The same is true of the decision of the Second Board of Appeal of 23 October 2023, DEVO v devonfw (Case R 336/2023-2), where the earlier mark had been broken down into ‘devon’ and ‘fw’, which is not the case here. Moreover, it should be borne in mind that the legality of EUIPO decisions must be assessed solely on the basis of Regulation 2017/1001, as interpreted by the EU judicature, and not on the basis of decision-making practice prior to those decisions (see, to that effect, judgment of 26 April 2007, Alcon v OHIM , C‑412/05 P, EU:C:2007:252, paragraph 65).

77 Thirdly, with regard to the judgment of 30 March 2022, Copal Tree Brands v EUIPO – Sumol + Compal Marcas (COPALLI) (T‑445/21, not published, EU:T:2022:198), the case concerned Article 8(5) of Regulation 2017/1001, namely the question of whether there was a likelihood of the mark applied for taking unfair advantage of the repute of the earlier mark ‘COMPAL’, and not the assessment of the likelihood of confusion within the meaning of Article 8(1)(b) of Regulation 2017/1001.

78 In the light of all the foregoing considerations, the action must be dismissed in its entirety.

Costs

79 Under Article 134(1) of the Rules of Procedure of the General Court, the unsuccessful party is to be ordered to pay the costs if they have been applied for in the successful party’s pleadings.

80 Even though the applicant has been unsuccessful, EUIPO contended that it should be ordered to pay the costs only in the event that a hearing is convened. Since no hearing was held, each party must be ordered to bear its own costs.

On those grounds,

THE GENERAL COURT (Sixth Chamber)

hereby:

1. Dismisses the action;

2. Orders Villafrut Srl and the European Union Intellectual Property Office (EUIPO) each to bear their own costs.

Costeira

Kancheva

Öberg

Delivered in open court in Luxembourg on 2 July 2025.

V. Di Bucci

M. van der Woude

Registrar

President

* Language of the case: English.