Shorts International Ltd v Google LLC [2026] EWCA Civ 668

[2026] EWCA Civ 668

Case No CA-2025-00161 AND CA-2025-000164

IN THE COURT OF APPEAL (CIVIL DIVISION)

ON APPEAL FROM THE HIGH COURT OF JUSTICE

BUSINESS AND PROPERTY COURTS OF ENGLAND AND WALES

INTELLECTUAL PROPERTY LIST (ChD)

Michael Tappin KC sitting as a Deputy Judge of the High Court

IL-2023-000037

Venue Royal Courts of Justice, Strand, London, WC2A 2LL

Date 25 May 2026

Before

CHANCELLOR OF THE HIGH COURT

LORD JUSTICE STEPHEN PHILLIPS

LADY JUSTICE FALK

Between

SHORTS INTERNATIONAL LTDAppellant

GOOGLE LLCRespondent

Simon Malynicz KC, Daniel Selmi and Laura Adde (instructed by Sheridans) for Appellant

Lindsay Lane KC and Jessie Bowhill KC (instructed by Fieldfisher) for Respondent

Hearing Hearing dates: 10 - 11 March 2026

Approved Judgment

This judgment was handed down remotely at 10.00am on Monday 25 May 2026 by circulation to the parties or their representatives by e-mail and by release to the National Archives.

.............................

Chancellor of the High Court :

[1]This is a trade mark case. The respondent (“Google”) runs the well-known platform YouTube. The case is about a service available on YouTube called “Shorts”. The appellant, Shorts International Ltd (“SIL”), contended that Google's use of various signs in relation to that service amounted to infringement of their registered trade marks and passing off. Google denied this. Amongst other points, a major part of Google’s case was that the word "shorts" was descriptive of short-form content of the kind being provided on this part of the YouTube platform. Google also counterclaimed, contending that the registered trade marks were invalid essentially because they were descriptive of the goods and services for which they were registered. Google also contended that four of the marks should be revoked for non-use for certain goods and services.

[2]Following the trial in October 2024, Michael Tappin KC (sitting as a Deputy Judge of the High Court) gave judgment on 31 October 2024. For most of SIL’s trade marks (see below), the modest figurative elements were sufficient to confer distinctiveness on what would otherwise have been highly descriptive matter, but the other mark, for the word SHORTSTV alone, lacked distinctiveness and was invalid for all the relevant goods and services. The judge then found that Google’s use of “Shorts” did not did not infringe SIL’s marks because the word was descriptive and, for most uses of the word, the associated figurative material was different from the SIL marks. There was also a finding of revocation for non-use for some goods and services, and the claim in passing off failed. The consequential order was made on 18 December 2024.

[3]SIL was granted permission to appeal by the judge on some grounds and on further grounds by Lord Justice Arnold on 25 March 2025. Google was granted permission by the judge to cross-appeal on validity and non-use. Permission was not sought by SIL in respect of passing off and revocation.

[4]In Google’s oral submissions counsel clarified that Google’s cross-appeal was reactive to the appeal and was put on the basis that if Google was wrong in its opposition to SIL’s appeal then the court should find the marks were invalid.

Background

[5]SIL is a producer and distributor of short films which it broadcasts on its own “ShortsTV” television channel. It is the proprietor of five registered trade marks. The first four are as follows:

[6]These marks were applied for on 20 February 2018 and registered on 5 October 2018. Each is registered for goods and services in classes 9, 38, and 41. I will refer to all four of them as the 2018 Marks. I will refer to the first two above as the SHORTSTV 2018 Marks and the latter two as the SHORTS 2018 Marks.

[7]Later SIL obtained a further registered trade mark (no. 3428383) for the word SHORTSTV. This mark (the ‘383 Mark’) was applied for on 13 September 2019 and registered on 6 December 2019. The 383 Mark is registered for goods and services in classes 9, 38, 41, and 42.

[8]In 2021 Google launched its YouTube Shorts service in the UK, which is dedicated to showing user-generated videos less than 60 seconds long and in “vertical” format. That format corresponds to what is seen on a handheld mobile phone. Google uses the word “shorts” in connection with that service as a standalone word, in the name “YouTube Shorts” itself, and in various graphical forms:

[9]Each of these uses was alleged by SIL to have infringed its marks under s10(2) and s10(3) of the Trade Marks Act 1994.

[10]Google denied infringement under either s10(2) or 10(3) and also relied on a distinct defence under s11(2)(b) of the Trade Marks 1994 Act, essentially on the basis that its uses were of a sign which was not distinctive and were in accordance with honest practices in industrial or commercial matters.

[11]Google’s counterclaim for invalidity was advanced against all SIL’s marks, on the basis of allegations of a lack of inherent or acquired distinctive character under s3(1) of the 1994 Act. Google’s counterclaim for revocation for non-use has been mentioned above.

[12]SIL denied any lack of inherent distinctive character and alleged the marks had acquired distinctive character through use. SIL also advanced a fall back argument, submitting that to the extent the marks lack distinctive character for some goods or services, this could be remedied by limiting the specifications of goods and services to include the words “save for short films” in relation to classes 9, 41, and 42.

The judgment

[13]The judgment includes a convenient summary at [3] on which the following is based (keeping the judge’s numbering): i) At the relevant dates, the meaning of the word “shorts” extended beyond “short films” and included other short-form audiovisual content. ii) SIL has a number of registered trade marks incorporating the word “shorts”. The 2018 Marks are valid but the word mark, i.e. the 383 Mark, is invalid for most goods and services. However, all SIL’s trade marks have low inherent distinctive character, and SIL’s use of its trade marks in the UK has not been extensive enough to enhance that distinctive character. The allegation of non-use succeeds for some goods and services. iii) Google has used the word “shorts” in various ways in relation to the YouTube Shorts service. Some of those uses are purely descriptive of the type of material on the service. Others, in combination with the YouTube name or logo, or with a derivative logo, act as indications of origin. iv) None of Google’s uses of signs including the word “shorts” gives rise to a likelihood of confusion as to origin. While there are significant similarities between the signs used by Google which include the word “shorts” and SIL’s trade marks, the similarities concern the aspects which are descriptive rather than those which give SIL’s trade marks their (low) distinctive character. v) While the similarities between Google’s signs and SIL’s trade marks will give rise to a link in the minds of the limited group of UK consumers amongst whom SIL’s trade marks have a reputation, Google’s uses of signs including the word “shorts” will not cause damage to the distinctive character or repute of SIL’s trade marks. vi) If the judge had concluded that there was a likelihood of confusion or significant damage to the distinctive character or repute of SIL’s trade marks, then he would have rejected Google’s defence that it was using its signs in accordance with honest practices. However, as a result of the judge’s conclusions, SIL’s claim of trade mark infringement fails. vii) While SIL has protectable goodwill associated with its trade marks amongst a limited group of UK consumers, Google’s signs do not misrepresent its service as being provided by SIL or in some way authorised by SIL. Therefore there is no passing off.

[14]There is one further detail to add. Having found that the 383 Mark was invalid for most goods and services (referred to at summary paragraph (ii) above), the judge also rejected SIL’s fallback argument to amend the specifications to add “save for short films”.

Grounds of appeal

[15]SIL seeks to appeal most of the judge’s conclusions save for passing off and the partial revocation of the 2018 Marks. To that end, it advances 12 grounds of appeal: ground 1 relates to the meaning of “shorts”; grounds 2-3 challenge the invalidity of the 383 mark for some goods and services, grounds 4-10 challenge the findings of non-infringement under s10(2) and grounds 11 and 12 address infringement under s10(3).

[16]Google’s cross-appeal advances six grounds. Google grounds 1-4 contend the 2018 marks are invalid under s3(1)(b), (c) and (d), Google ground 5 is that the 383 mark should be held invalid for all goods and services, and Google ground 6 contends that the partial revocation for non-use of the 2018 Marks did not go far enough and they ought to have been revoked for all goods and services.

[17]SIL takes two points in a Respondent’s Notice. Respondent’s Notice point 1 challenges aspects of the reasoning which formed part of the conclusion that the SHORTSTV 2018 Marks were valid and had distinctive character, in order to bolster SIL’s appeal against the conclusion that the 383 Mark (the word SHORTSTV alone) was invalid. The submission is that in the SHORTSTV 2018 Marks the word alone had inherent or acquired distinctive character without consideration of the play symbol (this will be explained below). Respondent’s Notice point 2 is a counterpart of SIL’s ground 3 but applied to the 2018 Marks. It would operate as an answer to a point in Google’s cross-appeal.

[18]A number of the issues in this appeal are interrelated. It makes sense to deal with SIL’s appeals first, addressing first the issues of infringement (grounds 4-12) and meaning (ground 1). That means that validity on SIL’s appeal (grounds 2-3) will come at the end. After that I will address Google’s Grounds if need be.

[19]Since at least some of the grounds are essentially challenges to findings of fact, it is worth referring to the cases summarising the correct approach on appeal. The relevant general principles were summarised by Phillips LJ in Kynaston-Mainwaring v GVE London Ltd

[2022] EWCA Civ 1339

; [2023] RTR 17 at [16]-[19] and further guidance, specifically in the context of trade mark infringement, was given by the Supreme Court in Iconix Luxembourg Holdings SARL v Dream Pairs Europe Inc

[2025] UKSC 25

;

[2025] RPC 15

at [93]-[95]. There is no need to set out either of these passages in this judgment and attempting to summarise them even further would risk appearing to add an unnecessary gloss. It bears noting that in Iconix the Supreme Court overturned the decision of the Court of Appeal to overturn the trial judge’s evaluative conclusion on infringement.

SIL’S APPEAL – INFRINGEMENT

[20]Section 10(2) of the 1994 Act provides that:

“A person infringes a registered trade mark if he uses in the course of trade a sign where because— (a) the sign is identical with the trade mark and is used in relation to goods or services similar to those for which the trade mark is registered, or (b) the sign is similar to the trade mark and is used in relation to goods or services identical with or similar to those for which the trade mark is registered, there exists a likelihood of confusion on the part of the public, which includes the likelihood of association with the trade mark.”

[21]In other words, to establish infringement under s10(2), it must be the case that:(i) a sign was used by a third party within the relevant territory;

(ii) the use was in the course of trade;

(iii) it was without the mark proprietor’s consent;

(iv) it was of a sign which is at least similar to the mark;

(v) it was in relation to goods or services which are at least similar to those for which the mark is registered; and

(vi) it gave rise to a likelihood of confusion (Match Group LLC v Muzmatch Ltd

[2023] EWCA Civ 454

at [26]).

[22]There is no dispute that (i)-(iv) are satisfied. The issues are whether Google’s use was in relation to goods or services, and gave rise to a likelihood of confusion (i.e. (v) and (vi)). For a sign to be used in relation to goods or services, it must be used for the purpose of distinguishing the goods or services in question, i.e., as a trade mark as such (Montres Breguet v Samsung

[2023] EWCA Civ 1478

at [83]).

[23]For the use to give rise to a likelihood of confusion, it must cause the average consumer to mistake the sign for the mark (so-called “direct confusion”) or believe that the goods and services denoted by the sign come from the same undertaking, or an economically linked undertaking, as that responsible for the goods and services denoted by the mark (“indirect confusion”) or vice versa (“wrong way round confusion”) (Liverpool Gin Distillery Ltd v Sazerac Brands LLC

[2021] EWCA Civ 1207

at [10]-[14]).

[24]At [174], the judgment sets out the standard summary of the applicable principles on likelihood of confusion. The purpose of the standard summary has been explained a number of times and is addressed by the Supreme Court in Iconix at [38]. The summary was (and remains) as follows:

“(a) the likelihood of confusion must be appreciated globally, taking account of all relevant factors; (b) the matter must be judged through the eyes of the average consumer of the goods or services in question, who is deemed to be reasonably well informed and reasonably circumspect and observant, but who rarely has the chance to make direct comparisons between marks and must instead rely upon the imperfect picture of them he has kept in his mind, and whose attention varies according to the category of goods or services in question; (c) the average consumer normally perceives a mark as a whole and does not proceed to analyse its various details; (d) the visual, aural and conceptual similarities of the marks must normally be assessed by reference to the overall impressions created by the marks bearing in mind their distinctive and dominant components, but it is only when all other components of a complex mark are negligible that it is permissible to make the comparison solely on the basis of the dominant elements; (e) nevertheless, the overall impression conveyed to the public by a composite trade mark may, in certain circumstances, be dominated by one or more of its components; (f) and beyond the usual case, where the overall impression created by a mark depends heavily on the dominant features of the mark, it is quite possible that in a particular case an element corresponding to an earlier trade mark may retain an independent distinctive role in a composite mark, without necessarily constituting a dominant element of that mark; (g) a lesser degree of similarity between the goods or services may be offset by a greater degree of similarity between the marks, and vice versa; (h) there is a greater likelihood of confusion where the earlier mark has a highly distinctive character, either per se or because of the use that has been made of it; (i) mere association, in the strict sense that the later mark brings the earlier mark to mind, is not sufficient; (j) the reputation of a mark does not give grounds for presuming a likelihood of confusion simply because of a likelihood of association in the strict sense; and (k) if the association between the marks creates a risk that the public might believe that the respective goods or services come from the same or economically-linked undertakings, there is a likelihood of confusion.”

[25]As is also familiar, although the summary is drafted by reference to registration, the same principles apply to infringement and when using it for that purpose one must consider the actual use of the sign complained of in the context in which the sign has been used (e.g. Iconix at [39]). So in the jargon of trade mark law in the context of infringement, one often refers to the comparison “between mark and sign”, referring to the registered mark and the alleged infringing sign.

[26]Of particular importance to this appeal is the fact that the determination of likelihood of confusion requires a global assessment. The summary above is nothing more than it purports to be, a convenient summary of applicable principles. It is not an algorithm or a route map. It is a mixture of topics to consider along with reminders of some common points of principle which often arise in relation to those topics. In the end what matters most is that the task is a global assessment.

[27]A point not mentioned in the summary is the flipside of sub-paragraph (h). As Arnold LJ noted in Lifestyle Equities CV v Royal County of Berkshire Polo Club Ltd

[2024] EWCA Civ 814

at [36]-[39], trade marks with a less distinctive character enjoy narrower protection than those with highly distinctive character and one way in which that principle manifests itself is that when the common element between a mark and a sign has low distinctiveness, that points against there being a likelihood of confusion (see also PlanetArt LLC v Photobox Ltd

[2020] EWHC 713 (Ch)

[28]My own decision as a High Court Judge in Nicoventures Holdings Ltd v London Vape Co Ltd

[2017] EWHC 3393 (Ch)

provides a simple illustration of the point. When the marks consist of descriptive words with some unremarkable styling, and the common element between them (VAPE) is descriptive of the relevant goods (electronic cigarettes, i.e. vapes), the conclusion there was that there was no likelihood of confusion.

s10(2) – grounds of appeal

[29]Ground 1 is about the meaning of “shorts”, grounds 4-6 focus on the comparison between the mark and the sign and grounds 7-10 are about the global assessment.

Ground 1

[30]At [46] the judge summarised SIL’s case about what “shorts” means. SIL’s position was (and is) that while the word had other meanings, e.g. short trousers, its principal meaning, in context and at the relevant time, was “short films”; and by “short films” what SIL also contends is that short films “are just like feature films on theatrical release, but shorter”. They are said to share with feature films the characteristics of having narrative structure, a plot, scenes and actors, and having (or aspiring to have) professional-looking production values.

[31]After considering the evidence and the issues in detail, the judge held at [58] that at the relevant time the meaning “extended beyond ‘short films’ and included other short-form audiovisual content”. Counsel for SIL contended that SIL did not challenge this factual conclusion but did challenge what was said to be the legal analysis. This requires a bit of explanation. Counsel’s objective is to attack an important aspect of the reasoning relevant to the overall assessment of likelihood of confusion under s10(2), namely the reaction of the average consumer when they encounter the Google signs.

[32]To step back, as the judge identified at [239], SIL’s best case was based on the SHORTS 2018 marks, which do not include “TV”, as registered for goods and services identical to those which Google has used its signs, such as “video recordings” and “presentation of videos”. It was also based on Google’s use of the composite sign below, which did not include any reference to YouTube. The sign was:

[33]I will refer to the logo appearing there above the word as the lozenge logo.

[34]At [240] the judge held this use did not give rise to a likelihood of confusion. Critical to this conclusion were two related statements in that paragraph. The first was that the word “shorts” was descriptive of the material in question. The second was that the average consumer would appreciate that the similarity between the SHORTS 2018 marks and the sign arose from the fact that the marks are registered for, and the sign is being used in relation to, goods and services for which the word “shorts” indicates characteristics. Also relevant was the play symbol, but at this stage dealing with ground 1 of the appeal I am focussing on “shorts”.

[35]To attack this conclusion on appeal SIL advances a case based on the following steps: i) First, it is said to be settled law that for the purposes of the exercise under s10(2) the average consumer is taken to be someone familiar with the registered trade mark. ii) Next it is said that to these people, the people familiar with the registered trade mark, “shorts” means a specific thing, which they know. It means short films. iii) Next it is said that when they see Google’s use and they see what it is used for (user generated videos lasting no more than 60s), they see that it is not a short film service. And even if this causes these people to think that “shorts” must be being used there with an extended meaning, to include the short audiovisual content provided by Google’s service, that does not involve the word being descriptive in the same way that “shorts”, as registered, is descriptive of short films. iv) Therefore, it is said, either way (i.e. whether the average consumer’s thinking stops at the first sentence of the previous sub-paragraph or includes the second sentence), one cannot discount the common element between mark and sign on the basis that it is descriptive, as the judge did, because at most the two are not descriptive in the same sense. If the goods to which the registered mark relates are not the same as the goods the sign is used for, then why, SIL asks rhetorically, would the average consumer discount the common element? Therefore there is a likelihood of confusion.

[36]The proposition at step (i) was common ground and we did not hear argument about it. It is based on a statement made by Arnold LJ sitting at first instance in Sky plc v SkyKick UK Ltd

[2018] EWHC 155 (Ch)

at [275] and I will take this opportunity to say I agree with it. As the judge, Mr Tappin KC, himself said at [168], as applied to s.10(2), this statement appears to be nothing more than a different way of expressing the point that, when considering whether there is a likelihood of confusion, the court must have regard to the impact of the alleged infringing sign on the proportion of consumers to which the trade mark is particularly distinctive (see Interflora Inc v Marks & Spencer plc

[2014] EWCA Civ 1403

at [123]). I will add only this, the fact that the average consumer is taken to be familiar with the registered mark does not ignore imperfect recollection. In the comparison, they will rely on whatever imperfect picture of the trade mark they have kept in mind - see paragraph [20] of Arnold LJ’s judgment in Tesco v Lidl

[2024] EWCA Civ 262

, cited with approval in the Supreme Court by Lord Briggs in Iconix at [30].

[37]The purpose of ground 1 of the appeal is to justify the next step, i.e. step (ii). The judge did not approach it this way and the submission is that by not doing so the judge was adopting what counsel for SIL called a “single meaning rule”. This is a reference to [17] of Arnold LJ’s judgment in Tesco v Lidl also cited by Lord Briggs in Iconix at [30]. In that paragraph Arnold LJ explained that the assessment from the perspective of the average consumer does not involve the imposition of a single meaning rule akin to that applied in defamation law. The point he was making was that this means that a conclusion of infringement is not precluded by a finding that many consumers of whom the average consumer is representative would not be confused. As Arnold LJ said, if the finding is that a significant proportion of the relevant public is likely to be confused, then a finding of infringement may properly be made.

[38]SIL’s case is that although (it says) it is not challenging the judge’s finding of fact that the meaning of “shorts” extended beyond its principal meaning of short films and included other short-form audiovisual content, what the judge did was impose this single meaning on all average consumers whereas, to the average consumer familiar with the registered mark, the predominant meaning was short films (as in step (ii)) and so the rest of the analysis follows.

[39]Ingenious though it is, in my judgment this submission is wrong. The relevant finding of fact about the meaning of shorts was about the meaning of that term to the relevant average consumer. The judge’s conclusion that it extended beyond short films and included other short form audiovisual content was plainly open to him on the evidence. There was no finding that there were relevant average consumers for whom “shorts” only meant short films and for whom the meaning did not extend to other short form audiovisual content.

[40]The fact that in law the test for likelihood of confusion posits an average consumer familiar with the registered mark does not turn that average consumer into someone for whom, on the facts of this case, “shorts” means only short films and does not have that extended element.

[41]Nor does SIL gain support from the judge’s conclusions on passing off, in which he did find the marks had protectable goodwill amongst a limited group of people in the UK (referred to as film makers, others in the film industry and film buffs). That was not and did not amount to a finding that there was a class of people for whom “shorts” only meant short films.

[42]The key finding at [240], that the average consumer would appreciate that the similarity between the SHORTS 2018 marks and the sign arose from use in relation to material for which “shorts” was descriptive was open to the judge and involved no error of law. I would dismiss ground 1.

[43]One does not get into the possibly novel question of whether it is enough to amount to a relevant likelihood of confusion as to trade origin (which is the test), when the average consumer sees a sign used in one descriptive sense which has a common element with the registered mark with which they are deemed to be familiar, when that common element, as it appears in the mark, is understood as descriptive albeit in a different sense. I will say only that I have doubts that it would satisfy s10(2).

[44]As part of the argument (at step (iii) above) Counsel for SIL submitted that the judgment in Technopol v OHIM

Case C-51/10

P 10 March 2011 at [50] laid down the degree of conviction on the part of the average consumer which is required before they can be taken to identify a mark or sign as descriptive in a legally relevant manner. There is nothing in this point on this judgment. The judge came to a clear conclusion (at [240]) expressed in definitive terms about the thinking of the average consumer. He was entitled to do so.

[45]Before leaving ground 1 it is worth identifying a different point on meaning which the judge dealt with in holding that shorts was not limited to short films but also included other short audio-visual content. The issue was the meaning of “short films” and whether it could function as a category of goods. At [30] above I have set out what SIL contended that term meant, but in the judgment at [60] to [62] the judge considered and rejected this definition. For one thing there is no consensus about their maximum length (at [60]), although the differences may not be large (the Oscars use up to 40 minutes while one of SIL’s witnesses said it was up to 45 minutes). Just as if not more important was a debate about what made a piece of audiovisual content a “film” (judgment [61]). SIL said it had to have narrative structure, a plot, scenes and actors and having (or aspiring to) professional-looking production values. But the judge identified serious problems with all these elements (for example animations were accepted but have no actors, and documentaries (also accepted) have neither actors nor plot). All of the points relied on by SIL were characteristics of the content itself and often subjective. I can find no fault with these aspects of the judge’s reasoning. For these reasons, the judge concluded at [62] that SIL’s definition was not a proper basis to identify a category of goods as opposed characteristics of goods. That point will come up again below.

Grounds 4-5

[46]By ground 5 the judge “wrongly regarded some of the allegedly infringing signs as purely descriptive and thus non-infringing”. The specific point is a challenge to a finding at [224] that two instances of Google’s use of the word “shorts” as a standalone word were purely descriptive and thus not “in relation to goods or services”, i.e., trade mark usage, for the purposes of s10(2).

[47]The first one is in a webpage called the creator page as follows:

[48]In this image one can see that just above the various images and below the ball there is a sort of menu of tabs running horizontally with six options:

“Home Videos Shorts Live Playlists Community”

. The image shows what you get if the Shorts tab is selected. In the context of this dispute, the significance of this argument is that there is no logo beside “Shorts” (cf. the top left menu in which the same word appears beside the lozenge logo). At [224] the judge held that the average consumer would understand the word “Shorts” in that menu as being used purely descriptively to refer to the type of material rather than as a trade mark.

[49]SIL’s allegation was that this use would be seen as trade mark use because “Shorts” was the name of Google’s new short-form user-generated content video service. However the judge clearly had that in mind overall, and despite it, the finding was that this very particular usage of “Shorts” was not an example of the word being used as a badge of origin. The fact “Shorts” was the name of Google’s new short-form user-generated content video service does not automatically render every use of the word trade mark usage. The conclusion was open to the judge and I reject that part of SIL’s case. Given that the word appears beside other descriptive labels like Home, Videos and so on, I would reach the same conclusion as the judge, but that is not the test on appeal.

[50]SIL also took a point based on the same single meaning argument advanced in ground 1, arguing that this would be seen as trade mark use by that proportion of the relevant public for whom “shorts” meant “short films”. For the same reasons given in relation to ground 1, that issue does not arise on the judge’s findings of fact.

[51]The other use of “Shorts” to which [224] applies was on a similar horizonal menu on a similar page of the mobile app. This time the options are “All Shorts Video Unwatched Watched”. This stands or falls with the first image and does not add anything new.

[52]Ground 4 is that the judge omitted to compare the registered mark with a sign comprising the word Shorts alone (with no other matter such as the word YouTube or the lozenge logo). Succeeding on ground 5 was a necessary prerequisite for this ground because those two instances were the ones SIL relied on of use of Shorts without other material. Therefore ground 4 does not arise.

Ground 6

[53]Ground 6 is a submission that in the context of s10(2) the judge “wrongly allowed the criterion of distinctiveness to pollute his comparison of the Marks”. To understand SIL’s point there are two preliminary points to make. The first is to appreciate that one way in which a mark may lack distinctiveness is because it is descriptive. There are other dimensions to the distinctiveness of a mark but in this case the descriptive nature of the word “shorts” and the play symbol play an important role. The second is that SIL accepts that distinctiveness plays a role in the overall assessment of the likelihood of confusion. The issue is about its involvement at the comparison stage.

[54]The judge set out his assessment of infringement under s10(2) in a well structured passage from [223] to [243], concluding at [243] that the case on s10(2) fails. That passage is in five parts as follows: i) [223] – [229] “Identification of the signs and context”, ii) [230] – [231] “The distinctive character of SIL’s marks”, iii) [232] – [234] “Comparison of the marks and sign”, iv) [235] – [237] “Comparison of the goods and services”, v) [238] – [243] “Overall assessment of the likelihood of confusion”. (The numbering of these parts is mine.)

[55]Paragraph [240] has been considered already. It is part of the overall assessment section. The paragraph includes references to the descriptiveness of the word “shorts” and the play symbol. Of course SIL contends this reaches the wrong conclusion overall, but ground 6 of the appeal is not focussed on the presence there of the reference to descriptiveness. SIL’s point is that in the section on comparison of the marks and sign descriptiveness is also referred to, at [232] when considering the sign as compared to the 2018 Marks, and again at [233] when considering the 383 Mark.

[56]Taking paragraph [232], this is the place in which the judge focusses on the similarities (or not) between the marks and the signs. The paragraph starts with visual similarities, noting that there is visual similarity in that both mark and sign contain the word “shorts” but also noting visual differences, e.g. that the play symbol (the triangle) appears in the O of “shorts” in the marks but not in the signs. Next it deals with aural similarities, noting that they are obvious. Then paragraph [232] turns to conceptual similarities, noting that they too are obvious but also noting that the word “shorts” is descriptive of the type of material (and the play symbol indicates the material can be played). Finally paragraph [232] concludes observing that the similarities between mark and sign are at a level which differs from that which has given the SHORTS 2018 marks their distinctive character and so avoided the invalidity attack (this will be addressed in the validity section below). The point the judge is making is that it was particular aspects of the figurative elements in the SHORTS 2018 Marks which gave them their distinctive character, however the similarities which do exist are not with those elements which gave the marks their distinctive character.

[57]Pausing here, it is clear that the judge’s findings in [232] were recognising a high degree of visual, aural and conceptual similarity but also noting what the nature of that similarity was. This is reflected in the summary paragraph in the judgment at [3](iv) which notes the significant similarities albeit that they concern aspects which are descriptive rather than the elements which give the marks their distinctive character.

[58]SIL submitted that taking descriptiveness into account here in [232] was wrong and contrary to authority. One aspect of SIL’s case, which I quite accept, is that this factor should not be given inappropriate weight in the overall assessment. As I say, I agree but only in the sense that no factor should be given inappropriate weight. The argument is that the mention of it in this part of the judgment indicates, when taken with the other parts of this whole section on 10(2) from [223] to [243], that it has been given undue or inappropriate weight, as a form of double counting. I cannot accept that. The judge was right and entitled to take that factor into account in the global assessment and, reading the relevant passages in the judgment as a whole, he was not giving it undue or inappropriate weight.

[59]SIL sought to rely on Adidas v Marca Mode (C-102/07) submitting that there the CJEU warned against “double counting” of distinctiveness. I agree one should not double count that or any other factor but that is not what the CJEU in that case was concerned with. The point in Adidas was that the CJEU was rejecting a submission that the public interest of leaving descriptive signs free for other traders to use should be taken into account as part of the test under (the equivalent of) s10(2). The CJEU held that that public interest consideration was a factor relevant to validity and did not play a role in the likelihood of confusion. That is a different issue and is not what the judge did here. The CJEU also held that the defence under (the equivalent of) s11(2) was also not part of the s10(2) test, which again is not relevant to what the judge did.

[60]SIL’s other point was that to mention it in the part concerned with comparing the marks to assess their similarity was itself wrong in principle. In other words it is the location of this reasoning which is an error of law. The best case for SIL on this issue is BSH Bosch and Siemens v. EUIPO(known as KOMPRESSOR)

Case C-43/15

8 November 2016, particularly paragraphs [58]-[60]. However properly understood, I do not believe that is what the ruling means but to explain why will require some detail about the case.

[61]BSH sought to register a mark consisting of a drawing of a fan and the words “compressor technology” for goods in Nice classes 7, 9 and 11. LG Electronics opposed, based on earlier marks for the word KOMPRESSOR registered amongst other things for goods in class 7 such as washing machines. Based on the issue of likelihood of confusion the EUIPO Opposition Division upheld the opposition for some goods (such as washing machines). BSH appealed to the First Board of Appeal and then to the General Court, on the ground that the sign KOMPRESSOR was descriptive of goods containing a compressor (such as a washing machine) and so the earlier marks had minimal distinctive character and thus even slight differences should be sufficient to preclude a likelihood of confusion. These appeals failed and BSH appealed to the CJEU. The CJEU dismissed the appeal.

[62]Thus BSH were refused registration of a mark consisting of descriptive words and a figurative drawing not because the mark was inherently unregisterable, but because there was a likelihood of confusion with an earlier mark consisting solely of (more or less) the same descriptive word. On the face of it that conclusion might appear at odds with Lifestyle Equities, PlanetArt and Nicoventures (above). However comparing outcomes on the facts is not productive, what matters are the legal principles.

[63]The key passages in the CJEU’s judgment are at [47]-[48] and [56]-[64]. In these passages BSH’s argument, described at [48], is characterised as being that the descriptiveness of what the marks had in common had to be taken into account so that a conclusion of likelihood of confusion was not permitted in that case. The CJEU’s concern at [58] was that accepting this would have the effect of disregarding the similarity of the marks as a factor (see also L'Oréal v OHIM (C-235/05 P)

[2006] ECR I-57

for a similar point). It would deprive the degree of similarity, which did exist, of any significance.

[64]It is this passage which can be read as a prohibition on taking descriptiveness into account at the comparison stage, however that is not what the court is saying. The argument the court was faced with treated the assessment of similarity as binary (similar or not) and potentially preclusive. After all if the marks were found simply to be “not similar” then there could be no likelihood of confusion. In a case like the present one, a conclusion that the mark and sign are “not similar” (thereby precluding a likelihood of confusion) would have been plainly wrong. Notably the judge did not do this. On the contrary he formed a nuanced view of similarity which recognised the obvious similarities while also recognising what they consisted of. The CJEU’s decision was concerned with avoiding preclusion. It was not concerned with prohibiting such a nuanced finding.

[65]That the CJEU’s concern is about preclusion can be seen at the end of [62] and in [63] where the point being made is that there must be at least the possibility of finding a likelihood of confusion, even when the mark has a weak distinctive character. Then at [64] the CJEU adopts more or less the same reasoning as it had identified at [58] had been adopted by the General Court below, i.e. that BSH’s argument would have the effect of disregarding the similarity as a factor and therefore was wrong.

[66]Thus overall, while BSH may be a surprising outcome on the facts, the legal principle it stands for does not reveal any error in the judge’s approach in this case. To the contrary his consideration of the similarity of mark and sign, taking into account as it did the descriptive nature of what they had in common, did not reach a bald preclusive conclusion. It contains no error of law.

[67]The fact that a common element between the mark and sign in this case is descriptive, is relevant. Absent authority, the fact this point is mentioned in a passage when the judge is comparing the mark and the sign and assessing the significance of what they have in common and how they differ, including at a conceptual level, is entirely natural. To find that a natural approach runs into a point of law does not make the task of judging any easier.

[68]More broadly, it should be noted that the assessment of likelihood of confusion is a global exercise which judges are entitled to conduct in different ways. It is clearly necessary to take account of relevant factors and disregard irrelevant ones. Creating legal sub-tests with rigid methodological criteria is a legal tendency I would prefer to avoid, in any situation. Trade mark law is already complex and difficult to state concisely. A further accretion of sub-tests serves no practical purpose other than creating the possibility of finding foot faults by trial judges in order to support appeals against what are, in truth, evaluative conclusions.

[69]There is one further dimension to this point which is worth identifying. One of SIL’s submissions was in effect that the judge erred in not reaching an interim conclusion on similarity. SIL argues that the judge ought to have held that the mark were “highly similar based on their visual, aural and conceptual similarities”. Just as it would have been wrong to summarise the result of the comparison as “not similar”, so it would be just as wrong on the facts of this case, particularly being in mind conceptual similarity, to summarise the result as SIL contend for, without adding more.

Ground 7

[70]Ground 7 relates to the judge’s overall assessment at [238]-[243] and contends that it was “rationally insupportable to find that the placement of the play symbol (in one of the uses complained of) was sufficient to dispel confusion where the goods/services were identical, the signs contained no other words, and both featured a red and white play symbol, albeit in a different position”.

[71]To understand this ground one needs to start with the judge’s rejection of Google’s case that the 2018 Marks were invalid for lack of distinctive character. The issue is dealt with in more detail below in the section on SIL’s validity appeal. At this stage the points to note are that “shorts” and the play symbol were descriptive, and the font and the contrast between the lettering was insignificant. What gave the marks distinctive character so that they were valid was the overall impression of the mark as a whole comprising the combination of the word and the play symbol rendered in red and located inside the O of shorts. Then when it came to infringement it was this combination which the judge noted was absent from the Google signs at [240]. Therefore the average consumer would not mistake the sign for the mark nor would they believe the goods or services came from an undertaking economically linked to the mark. The characterisation of the judge’s conclusion in ground 7 as being that placement of the play symbol was “sufficient to dispel confusion”, is not accurate.

[72]SIL also submitted that the judgment subjected the mark to a level of scrutiny which the average consumer would not undertake. The argument is that when face with a scrolling screen, where a very small version of Google’s logo would appear, the average consumer would be unable to discern whether there was a small play symbol within the letter “O”, especially when accounting for imperfect recollection of the earlier mark. No doubt they would but that does not reveal anything rationally unsupportable about the judge’s conclusions.

[73]In my judgment the judge was right that word “shorts” and the play symbol themselves are not distinctive. This is the decisive factor. Amongst other things it means that such distinctive character as there is in the SHORTS 2018 Marks arises from the combination of shorts with the play symbol placed within the letter “O” (and the colouring). An argument based on the inability of the average consumer to spot, as present in the sign, one of the few distinctive elements of the earlier mark is not a promising approach to finding a likelihood of confusion. SIL’s argument about imperfect recollection in effect seeks to strip the mark of its distinctive character.

[74]In my judgment this ground does not reveal reasoning by the judge which was rationally unsupportable. The judge’s conclusions were open to him.

Grounds 8-10

[75]None of grounds 8 to 10 were pursued separately from the points I have addressed above in relation to grounds 1 to 7. There is no need to consider them further. Conclusion

[76]Accordingly, I would dismiss SIL’s appeal on s10(2).

s10(3) – the law

[77]SIL also contends that its marks were infringed under s10(3). That section provides that:

“A person infringes a registered trade mark if he uses in the course of trade, in relation to goods or services, a sign which— (a) is identical with or similar to the trade mark, … where the trade mark has a reputation in the United Kingdom and the use of the sign, being without due cause, takes unfair advantage of, or is detrimental to, the distinctive character or the repute of the trade mark.”

[78]To establish infringement under s10(3), it must be the case that:(i) the mark has a reputation in the UK;

(ii) a sign was used by a third party within the UK;

(iii) the use was in the course of trade;

(iv) it was without the mark proprietor’s consent;

(v) it was of a sign which is at least similar to the mark;

(vi) it was in relation to goods or services;

(vii) it gave rise to a “link” between the sign and the mark in the mind of the average consumer;

(viii) it gave rise to one of three types of injury, namely (a) detriment to the mark’s distinctive character, (b) detriment to the mark’s repute, or (c) unfair advantage being taken of the mark’s distinctive character or repute; and

(ix) it was without due cause (Match Group (above) at [55]).

[79]The judge summarised the law on s10(3) at [244]-[254] in a passage which SIL does not criticise, and he then decided the issues at [255]-[260]. The judge held that the marks did not have a reputation in the UK and so the claim under s10(3) failed on that basis. He went to consider the other infringement issues on the hypothetical basis that the marks at least had a reputation amongst consumers of “short films”, holding that Google’s use would be likely to give rise a link, but would not be detrimental to the distinctive character or repute of the marks. So the hypothetical 10(3) claim would have failed too.

[80]Ground 11 of SIL’s appeal challenges the conclusion on reputation and ground 12 originally addressed both detriment to distinctive character and detriment to repute, but the point on repute was not pursued.

Ground 11

[81]The legal principles on reputation were set out by the CJEU in General Motors Corp v Yplon SA (C-375/97)

[1999] ECR I-5421

at [24]-[27], as follows:

“24. The public amongst which the earlier trade mark must have acquired a reputation is that concerned by that trade mark, that is to say, depending on the product or service marketed, either the public at large or a more specialised public, for example traders in a specific sector. 25. It cannot be inferred from either the letter or the spirit of Article 5(2) of the Directive that the trade mark must be known by a given percentage of the public so defined. 26. The degree of knowledge required must be considered to be reached when the earlier mark is known by a significant part of the public concerned by the products or services covered by that trade mark. 27. In examining whether this condition is fulfilled, the national court must take into consideration all the relevant facts of the case, in particular the market share held by the trade mark, the intensity, geographical extent and duration of its use, and the size of the investment made by the undertaking in promoting it.”

[82]Thus the reputation must exist among the relevant public, which in turn depends on the product or service marketed. While there is no minimum requirement for knowledge by a given percentage of that group, the mark must have a reputation among “a significant part of the public concerned by the products or services covered by that trade mark.” (see [26]).

[83]Arnold J (as he then was) in Sky v Skykick

[2018] EWHC 155 (Ch)

at [307] made the point that this is not a particularly onerous requirement. I agree. Nevertheless it is a requirement which must be satisfied, and without it the enhanced protection for a mark which s10(3) provides is not available. Marks without a reputation with a significant part of the public concerned are always entitled to the protection provided by s10(1) and s10(2).

[84]The judge’s decision that no reputation existed for the purposes of s10(3) was made at [255]. The reasoning was as follows. First, for this purpose the relevant date was June 2021, the date Google started the acts complained of. Second, the judge referred back to his earlier findings about SIL’s use of the marks in the UK. The essential facts were found at [63]-[81] and then summarised at [128] when considering (and rejecting) the submission that SIL’s marks had acquired distinctiveness through use as at 8 February 2022. Next the judge found that SIL’s marks were not known by a significant part of the public concerned with any of the categories of goods or services for which the marks were registered. Then, even if one were to consider “cinematographic films”, only a very small part of the UK public concerned with such goods knew of SIL’s marks in that context. Finally the judge did not accept that “short films” could be a proper sub-category of goods and services for the reasons explained above.

[85]Pausing here, it is worth noting that although the judge went on to consider the further issues under s10(3) on the hypothetical basis of the existence of a reputation among consumers of short films, that was not based on a finding of fact that that such a reputation did exist even if, which he did not accept, such a sub-category was a proper one. I will come back to this after reviewing the evidence.

[86]SIL’s appeal on ground 11 is put on the basis that the judge erred in his assessment of the evidence. SIL submits that the judge ought to have held that the evidence was sufficient to show a reputation at least in respect of consumers of short films. To succeed SIL needs to establish two things. First that the facts support a finding of reputation - at least for consumers of short films (whatever that category amounts to) and second, if so, that the judge erred in holding that that category was not a proper one.

[87]In my judgment this is another aspect of SIL’s appeal which simply fails on the facts. I will set out below the points made by SIL in its skeleton in full. I do this in fairness to SIL and to avoid summarising the points they wish to make. SIL’s submissions are as follows: i) SIL was founded in 2000 to create a foundation for short films, and a short film industry, under the name Britshorts Ltd. It adopted its current name nearly 20 years ago – in 2005. A rebranding exercise in 2017/2018 shifted branding from the use of “Shorts” to “ShortsTV”; ii) SIL has distributed Oscar-nominated short film compilations since 2006 (being the only company in the world that does so), initially via iTunes and later as theatrical releases in North America (since 2009). UK releases began by at least 2016 and, in 2024, around 30 cinemas showed the compilations, selling approximately 900 tickets (other figures were not available). The Judge accepted that at least some use had been made of the [SHORTSTV 2018] Marks at the start and end of the compilations since the re-branding exercise in 2017/2018 (at judgment [67]). See also judgment [68], where he noted by reference to SIL’s evidence that branding ([SHORTSTV 2018] Marks) appears on posters and trailers to promote SIL’s Oscar nominated short film compilations, which carry the [SHORTSTV 2018] Marks. Furthermore, Battersea Cinema, Riverside Studios and Phoenix Cinemas used SIL’s material on their websites, showing the 615/649 Marks, and Chiswick cinema showed such material on a screen in its foyer; iii) From 2014-2020, SIL’s main production activities involved packaging collections of short films for distribution on SIL’s own channels (including hosted shows), producing content around the making of short films, including filmmaker interviews and “behind the scenes” programmes, and running short film competitions, including one called ‘The Pitch’. SIL’s evidence also referred to a short film produced by SIL called Freeze-Frame, commissioned by a Sino-British partnership in about 2014. SIL’s evidence provided screenshots showing the use of a variant of the [SHORTS 2018] Marks at the end of Freeze-Frame and one of the films arising from The Pitch, and of the use of the [SHORTSTV 2018] Marks at the end of compilations of short films distributed through Amazon (notwithstanding that there was no evidence as to the revenue generated through distribution in the UK); iv) SIL had a linear TV channel on Sky called “ShortsTV” for a little over a year from 2008 and on Amazon Prime Video from November 2020 until late 2022. Whilst there was no evidence of viewing figures for the Sky channel (nor the Amazon Prime Video channel in the UK), SIL’s evidence showed that, worldwide, this channel received around 2% of the viewership of the Amazon Prime Video - Video on Demand (VoD) service; v) SIL launched its Amazon Prime Video VoD channel in the UK between October 2020 and January 2021 (Amazon Prime being one of the largest subscription services in the world). The Amazon Prime Video channel had the [SHORTSTV 2018] Marks visible on the screen in various ways on and between SIL’s content, as well as “ShortsTV” being used orally in voiceovers between content. The Judge accepted “that it is more likely than not that, since launch, the [SHORTSTV 2018] Marks have been used as the name of the channel, and appeared within the ShortsTV branded area within Amazon Prime Video” (at judgment [73]); vi) See also judgment [74]-[81] for other examples in relation to SIL’s FAST channels, EST, SIL’s app, SIL’s YouTube channel, other social media and SIL’s publicity at film festivals etc., which is all one piece with the evidential picture above showing use of SIL’s marks.

[88]These submissions put SIL’s case as attractively as it can be, but they do not face up to the major deficiencies in the evidence.

[89]Starting with sub-paragraph (i), it is not in dispute but alone it does not prove the point.

[90]The Oscar nominated short film compilations are referred to at sub-paragraph (ii). The judgment at [128] recognised that there had been use of the SHORTSTV 2018 Marks at the start and end of these compilations and in publicity materials, but noted that there was no evidence about the number of people who saw these marks when viewing the compilations at cinemas or on publicity materials. The fact 900 people bought tickets in 2024 is irrelevant. If anything the reference to it highlights the absence of evidence focussed on the relevant date more than two years earlier.

[91]In relation to packaging collections of short films (sub-paragraph (iii)), as that paragraph notes there was no evidence of revenue generated through distribution in the UK (judgment [69]). Also at [69] the judge noted that distributions through the BBC did not include SIL’s marks.

[92]In relation to sub-paragraph (iv) and the linear TV channel ShortsTV, this is addressed by the judge at sub-paragraph (ii) of [128]. He held that there were no relevant viewing figures. The 2% figures quoted in sub-paragraph (iv) are worldwide figures and do not help.

[93]In relation to sub-paragraph (v) and SIL’s Amazon Prime Video VoD (video on demand) channel, the judge noted at [73] there was surprisingly little evidence about its viewership. It was launched in the UK between October 2020 and January 2021. There was evidence of 1000 monthly UK subscribers in 2024 but that is again too late and no basis for a finding about June 2021. One of SIL’s witnesses also accepted that the numbers were likely to be lower in previous years. The only hard evidence identified in the judgment (and SIL do not suggest there was any more) was that 71 people viewed the most streamed title in July 2021. SIL’s founder Mr Pilcher agreed that at that time viewership was very small.

[94]In relation to compendious points in sub-paragraph (vi), they do not help either. The judge had them well in mind. To take a single example, the passage refers to SIL’s FAST channels (FAST stands for free, advertising supported streaming TV) but as the judge held at [74] these FAST channels were launched in the UK “between June/July 2021 and October 2021” and so the judge was entitled to place no weight on them when the relevant date is June 2021.

[95]Standing back, it is obvious that there will have been people in the UK who had heard of SIL and seen SIL’s marks, but that is not the test. It is apparent that SIL’s evidence to show a reputation for the marks in the UK in June 2021 was paper thin. The only numbers in hard evidence are tiny. The judge’s conclusions that the marks were not known by a significant part of the public concerned with any of the categories of goods or services or even just cinematographic films, was plainly right. It is also manifest that this weak evidence does not establish that a significant part of the public defined as consumers of short films would have known the mark either. That is plain whatever the boundaries of the sub-category might be. The evidence establishes no such thing.

[96]In oral argument there was reference to the judge’s decision in the context of passing off, in which he was prepared to accept that by June 2021 SIL had protectable goodwill amongst “filmmakers, some others in the film industry, and film buffs” (at [286]). That does not assist SIL in relation to s10(3) and in fact if anything it makes the point the other way. It is yet another, different, and even narrower group of consumers.

[97]Therefore, as the judge held at the end of [255], SIL did not establish a reputation in the UK for the purpose of s10(3). There is no need to consider the question of sub-categories, which I prefer to leave to a case in which it matters. Nor do the issues under ground 12 arise.

Conclusion

[98]Accordingly, the appeal in respect of grounds 11-12 is dismissed.

SIL’S APPEAL – VALIDITY

[99]I will now deal with SIL’s appeal concerning validity of the 383 Mark (that is the registration for the compound word SHORTSTV). In the end it amounts to some fairly brief points but they will require a substantial run up both on the law and on the facts.

[100]To recap, the judge found that for the majority of the goods and services registered, the 383 Mark was invalid for a lack of inherent distinctiveness on each of grounds s3(1)(b)-(d) and that it had not acquired distinctive character through use. The only goods and services for which it was not invalid were not things relied on by SIL as part of its infringement case. As a fall back SIL sought to amend to add the words “save for short films” in various places in the specifications. The judge rejected this on a number of bases.

[101]Ground 2 of the appeal challenges the assessment of the distinctive character of the 383 Mark, contending that it is inherently distinctive or ought to have been held to have acquired distinctive character through use. Ground 2 also takes an unspecific intermediate point, submitting that for some of the goods or services on which SIL lost, the mark ought to have been found distinctive. The only specific example of this to which my attention has been drawn was computer software. It is not the court’s task, either at first instance or on appeal, to identify any other examples for itself.

[102]Ground 3 challenges the rejection of the proposed amendment.

[103]There is no need to consider s3(1)(d) (customary in the trade) any further because although the judge did hold that the 383 Mark was invalid on this ground (see [118]) it is clear that on the facts this conclusion is not logically distinct from the conclusions under s3(1)(b) (devoid of any distinctive character) and s3(1)(c) (consists exclusively of descriptive indications).

The law

(c) The law

[104]The legislation relating to this is s3(1) of the 1994 Act, which provides as follows: “The following shall not be registered—(a) signs which do not satisfy the requirements of section 1(1),

(b) trade marks which are devoid of any distinctive character,

(c) trade marks which consist exclusively of signs or indications which may serve, in trade, to designate the kind, quality, quantity, intended purpose, value, geographical origin, the time of production of goods or of rendering of services, or other characteristics of goods or services,

(d) trade marks which consist exclusively of signs or indications which have become customary in the current language or in the bona fide and established practices of the trade: Provided that, a trade mark shall not be refused registration by virtue of paragraph (b), (c) or (d) above if, before the date of application for registration, it has in fact acquired a distinctive character as a result of the use made of it.” (a) signs which do not satisfy the requirements of section 1(1), (b) trade marks which are devoid of any distinctive character, (c) trade marks which consist exclusively of signs or indications which may serve, in trade, to designate the kind, quality, quantity, intended purpose, value, geographical origin, the time of production of goods or of rendering of services, or other characteristics of goods or services, (d) trade marks which consist exclusively of signs or indications which have become customary in the current language or in the bona fide and established practices of the trade:

[105]Despite the order in which the sub-sections appear in s3(1), it is convenient to start with s3(1)(c) (consists exclusively of descriptive indications) and then address s3(1)(b) (devoid of any distinctive character). The reason is because, as Arnold J observed in Starbucks (HK) Ltd v British Sky Broadcasting Group plc

[2012] EWHC 3074 (Ch)

at [91](v) there is an overlap between s3(1)(b) and s3(1)(c) such that anything within s3(1)(c) will also be within s3(1)(b), (see Agencja Wydawnicza Technopol sp. z o.o. v OHIM (C-51/10 P) [2011] ETMR 23 (Technopol) at [46]-48]).

s3(1)(c) (consists exclusively of descriptive indications)

(c) s3(1)(c) (consists exclusively of descriptive indications)

[106]The text of s3(1)(c) is set out above. The provision applies to trade marks which consist exclusively of signs or indications which may serve to designate a characteristic of the goods or services. A comprehensive survey of the relevant legal principles for assessing this was conducted by the CJEU in Technopol at [33]-[50]. Since no error of law is alleged in respect of them, I will not rehearse those principles here.

[107]Many marks consist of a word which would have been unregistrable for the relevant goods or services and some additional figurative styling (sometimes also referred to as non-verbal matter). Sometimes that figurative styling is quite modest, nevertheless the mark is registered thanks to that figurative matter. However once it has acquired the registration the proprietor commences proceedings against people who are using the word in the same context either without any figurative matter at all or with different figurative elements. Of course each case has to be decided on its facts but notably, in Starbucks at [117] Arnold J identified and deprecated this practice, coining the memorable expression, figurative figleaf, to describe the matter in the registered mark. One example was the mark in Starbucks itself (the word “now” with six lines radiating from the “o”), which was used to bring a claim against use of the word “now” alone.

[108]In Starbucks Arnold J held that the mark in that case was unregistrable under s3(1)(c) or alternatively under s3(1)(b) as being devoid of distinctive character (at [116]). The word “now” was clearly descriptive in that case but one might have said that the lines could have taken the mark out of the ambit of s3(1)(c) given that the sub-section is concerned with signs consisting exclusively of descriptive matter. Looked at that way a descriptive word coupled with some minor figurative element could never be invalid (or unregistrable) under s3(1)(c) and would always have to be considered under s3(1)(b). Arnold J’s conclusion on the facts relating to s3(1)(c) was that the average consumer would regard “now” as descriptive of a characteristic of the service in question, namely its instant and immediate nature and, he held, the figurative elements of the mark did not affect that conclusion. On appeal in Starbucks it was not necessary to examine this issue. The Court of Appeal (Kitchin LJ giving the leading judgment) upheld Arnold LJ’s conclusion that the mark was devoid of distinctive character (i.e. invalid under s3(1)(b)).

[109]At [92] the judge in the present case held that s3(1)(c) can apply in circumstances in which there is some figurative matter involved, provided it is done with care and bearing in mind an observation of Geoffrey Hobbs QC, sitting as the Appointed Person, in FLYING SCOTSMAN TM

[2012] RPC 7

at [28]. There Mr Hobbs QC characterised the circumstances in which s3(1)(c) could be applied in such a case as being when what he called the non-verbal content was “insignificant in the context of the sign as a whole”.

[110]As the judge in the present case also noted, this approach to s3(1)(c) is also supported by the judgment of the General Court in Caffè Nero Group Ltd v EUIPO

Case T-37/16

. There the court upheld an application of the equivalent of s.3(1)(c) to reject a mark consisting of the words “Caffè Nero” written in black in a particular font, with “Caffè” above “Nero” and in smaller letters, against a blue background. The court held that the figurative elements did not “convey an immediate and lasting impression which members of the public may retain”.

[111]I agree with the judge. Provided it is done carefully, s3(1)(c) can be applied to cases in which the mark consists of descriptive indication (or indications) along with some other figurative (or non-verbal) matter. If, in the context of the mark as a whole, that figurative matter does not convey an immediate and lasting impression which the average consumer would retain, then it is insignificant. Without more, such a mark would fall foul of s3(1)(c).

[112]Two final points are worth noting at this stage. First, as the judge also noted (at [91]) a characteristic of goods or services within s3(1)(c) can include the potential content of those goods or services (citing two Appointed Person decisions: Linkin Park [2016] 74 at [44]-[47] and FLYING SCOTSMAN at [15]-[17]). Second, as the judge noted at [93], POSTKANTOOR (Koninklijke KPN Nederland NV v Benelux-Merkenbureau (C-363/99)

[2004] ECR I-1619

) at [98]-[100] has a useful summary of the approach to take in assessing a compound word mark. As the CJEU held, a mark consisting of a word composed of elements, each of which is descriptive, is itself descriptive, unless there is a perceptible difference between the word and the mere sum of its parts. Assuming the compound word has not acquired its own meaning, for the compound as a whole to be more than the sum of its parts, it would need to create “an impression which is sufficiently far removed from that produced by the mere combination of meanings lent by the elements of which it is composed” (POSTKANTOOR at [100]).

s3(1)(b) (devoid of any distinctive character)

(b) (devoid of any distinctive character)

[113]As the judge noted at [120], for a mark to possess distinctive character for the purposes of s3(1)(b), it must serve to identify the goods or services in respect of which registration is applied for as originating from a particular undertaking, and thus to distinguish the goods or services from those of other undertakings (OHIM v BORCO-Marken-Import Matthiesen GmbH & Co KG (C-265/09 P)

[2010] ECR I-0000

at [31]).

[114]The question is always whether the mark as a whole serves to distinguish the goods or services in respect of which registration is applied for. Therefore a mark combining two or more elements, both of which independently lack distinctiveness, cannot necessarily be held to be devoid of distinctive character (see Eurohypo AG v OHIM (C-304/06 P)

[2008] ECR I-3297

at [41]) nor can it necessarily be held to be distinctive. In the case of a compound mark, like any other, the assessment of its distinctive character must be based on the overall impression the mark as a whole creates in the mind of the average consumer.

Acquired distinctiveness (s3(1) proviso)

[115]A mark which is not inherently distinctive can acquire distinctive character through use. Whether this is so is a question of fact. The judge at [126] identified a set of helpful general principles relevant to assessment as summarised by Arnold J at [156]-[163] of W3 Ltd v easyGroup Ltd

[2018] EWHC 7 (Ch)

. Neither party took a point on this and there is no need to set them out. As the judge also noted (at [127]), a pertinent observation was made by Jacob J in British Sugar plc v James Robertson & Sons Ltd

[1996] RPC 281

at [306] about how hard it is to show that a descriptive (or laudatory) word has acquired distinctiveness through use. Compelling evidence is needed.

Amendments to exclude goods or services

[116]When s3(1) is found to apply in relation to goods or services as registered, it may be possible to amend the specification down to a class of goods or services which avoids the problem. The overriding priority when assessing a proposed amendment must be clarity on the register (see for instance s32(2)(d): “The application shall contain—…(d) a representation of the trade mark, which is capable of being represented in the register in a manner which enables the registrar and other competent authorities and the public to determine the clear and precise subject matter of the protection afforded to the proprietor”). Therefore, to the extent that an amendment undermines this clarity, or in other words offends legal certainty, the court (or office) is entitled to refuse it.

[117]Negative exclusions raise particular difficulties and the law is not easy to apply. It derives from POSTKANTOOR at [113]-[115]. To understand what was said you need to know that POSTKANTOOR is the Dutch for POST OFFICE and that the applicant was seeking to register the word POSTKANTOOR for, amongst other things, services such as the issue of postage stamps. The office held the mark applied for lacked distinctive character on the basis that it was descriptive of goods and services provided in relation to a post office. As one way of attempting to deal with that objection, the applicant sought to amend the specification of services to add the words “provided they are not connected with a post office”. This was rejected. The eighth question referred to the CJEU was about this aspect.

[118]The CJEU held (at [113]) that an applicant is entitled to limit the goods and services within a class to some of those goods and services, for example because the mark would lack distinctiveness in relation to other goods and services within that class. However (at [114]) the CJEU went on to hold that a negative limitation to limit the goods or services concerned to those which “do not possess a particular characteristic” was not permissible. It was contrary to legal certainty as to the extent of protection of the mark ([115]).

[119]In Omega Engineering v Omega SA

[2012] EWHC 3440 (Ch)

at [43]-[57] Arnold J addressed this, noting the difficulties its application had caused and referring to Geoffrey Hobbs QC’s decision as the Appointed Person, in Croom’s Trade Mark Application

[2005] RPC 2

at [30]. There Mr Hobbs QC held that what POSTKANTOOR was excluding was limitations which referred to “characteristics that may be present or absent without changing the nature, function or purpose of the goods”. Arnold J agreed with that approach and noted other decisions of Appointed Persons in which that approach had been taken.

[120]I can see that on the facts of POSTKANTOOR a postage stamp is the same thing whether its supply was “connected with a post office” or not, and I can also see that this way of applying POSTKANTOOR seems to work in the cases in which it has been applied by those sitting as the Appointed Person. I have no reason to think that those cases are flawed in any way.

[121]However I confess I do not find that that explanation of POSTKANTOOR helps at least in a case like the present one. The vice in POSTKANTOOR was two fold. The amendment was drafted in negative terms and the exclusion used as its defining language (more or less) the very characteristic of the services of which the mark was descriptive. That is confusing for the public since the problem in the first place is that the word as it appears in the mark is descriptive of that characteristic and that problem has not changed even when the limitation is applied. Most importantly this approach puts all the risk on other businesses and users of signs. It is like someone trying to register LARGE for goods, being told LARGE lacks distinctive character because it refers to the size of goods and then suggesting that the specification be amended to say in effect “goods save for large ones”. If it was accepted one would end up with the word LARGE registered for “goods save for large ones”. The fact that in theory the relevant goods now do not include any large ones, and so in a sense a pedant could say the mark cannot in fact be descriptive of them, is unreal and, it creates serious uncertainty for the public.

The judgment on validity

[122]To understand the judge’s assessment of s3(1)(c) for the 383 Mark I need to start, as the judge did, with his assessment of the SHORTS 2018 Marks, because they were linked. It is also convenient as I go through to highlight the points made on appeal, because then they can conveniently be dealt with in the context in which they arose.

[123]The section started at [94] to [96] and focussed on goods such as video recordings and cinematographic films (see [94] in particular). As mentioned already in the context of the appeal about placement of the play symbol (ground 7), the judge noted that the SHORTS 2018 Marks consisted of the word “shorts” in an ordinary sans serif capitalised font, either black on a white background or white on a black background, with a red triangle in the “O”. He also noted the finding that one meaning of the word “shorts” was short-form audiovisual content, including but not limited to “short films”, and held that the word “shorts” would be recognised by the average consumer as a description of a characteristic of the goods, at least in the case of video recordings and cinematographic films (and so on). Next the judge made the point (at [95]) that he needed to consider whether the presentation of the word “shorts” in the SHORTS 2018 Marks meant they did not consist exclusively of signs which may serve to designate characteristics of those goods. On this he held (at [95]) that the font and the contrast between the lettering and the background were insignificant in the context of the mark as whole. That left the red triangle in the letter “O”, which he held would be identified by the average consumer as a play symbol and therefore as something designating a characteristic of the goods, namely that they can be played. However he then went on to address the combination of the word “shorts” and the play symbol rendered red and located within the “O”. The way these elements were combined were held not to consist exclusively of signs which may serve to designate characteristics of those goods. The overall impression created by the combination was one which went beyond the mere juxtaposition of two descriptive indications. Therefore the objection to the SHORTS 2018 Marks under s3(1)(c) (consists exclusively of descriptive indications) failed.

[124]In the next passage ([97] to [108]) the judge went on to consider the validity of the same marks if he was wrong in his prior conclusion on validity, but now focussing on the other goods and services they covered. The reason for doing this was because that would lay the groundwork for consideration of the 383 Mark. It was a sensible and convenient way of essentially examining whether “shorts” was descriptive of these other goods and services. Two examples are sufficient for present purposes. The first is at [98]. Here the judge addressed electronic media such as DVDs. He noted that these are carriers of audiovisual content (see Linkin Park above) and held in effect that “shorts” and the play symbol are apt to describe a DVD which has a collection of short form audiovisual content on it. Therefore if he had been wrong about the status of the SHORTS 2018 marks under s3(1)(c) and in fact those marks were invalid because they consisted exclusively of descriptive indications relating to films, then that conclusion on invalidity also followed for goods such as DVDs because those things could carry such films.

[125]Then at [99] the judge addressed computer software. On this he reasoned as follows. “Computer software” could include software for accessing audiovisual entertainment. However as in the previous paragraph, the descriptive aspects of the SHORTS 2018 Marks were also apt to describe the content which computer software for accessing audiovisual entertainment would access. Therefore again, if he had been wrong and in fact the SHORTS 2018 Marks were invalid under s3(1)(c), then that same conclusion would also have followed for computer software.

[126]Pausing here, on appeal SIL submit that “shorts” is not descriptive of computer software. However this does not face up to the point the judge made, which I have just described. He dealt with it in the context of the 2018 Marks but was doing this specifically because he understood it would be relevant to the 383 Mark. The judge’s reasoning here makes sense and contains no error. I reject that aspect of the appeal.

[127]Turning back to the judgment, from [100] to [108] the judge worked through other categories of goods and services but nothing independent now turns on those.

[128]At [109] the judge turned to the SHORTSTV 2018 Marks. He noted SIL’s case that including “TV” in red meant that these marks could not be worse than the SHORTS 2018 Marks (which I think must be right) but observed that as he was about to explain in addressing the 383 Mark, their position was not better either.

[129]At [110] the judgment turns to address the 383 Mark itself. Again to recap, this is the compound word SHORTSTV registered for a wide range of goods and service, including video recordings and cinematographic films. He held that “shorts” would be understood in the manner he had already described. Next he held that TV would be understood by the average consumer to mean television and taken on its own, would designate a characteristic of goods and services, that they are or can be provided by televisual means. The question therefore (at [111]) was whether the combination of “SHORTS” and “TV” created “an impression which is sufficiently far removed from that produced by the simple combination of those elements”. These latter words in quotation marks were taken from POSTKANTOOR at [100]. The judge held that the impression given by the combination of “SHORTS” and “TV” was simply that the goods are shorts provided by televisual means and that the services involve the provision of shorts by televisual means, nothing more. He noted counsel’s submission that SHORTSTV had unusual syntax but did not see why. He also noted the SIL’s counsel had submitted that consumers were familiar with TV channels being named “XTV” where X was a reference to the type of content. The judge said that this rather supported the view that there was nothing unusual about the syntax. He also noted that counsel’s point had been that consumers would take “XTV” as a reference to a channel (i.e. as having a particular origin), even though X is a type of content, but the judge explained that while with use consumers may come to understand “XTV” to refer to a particular channel, that does not mean that without any use of that mark they would assume that it did. He also held that the evidence did not establish that they would.

[130]Pausing here, on appeal SIL submits again that XTV is a well-established formula and that “TV” in such a compound word is understood by consumers to indicate that the word functions as a proper name, i.e. an indication of origin. The second point is the submission that SHORTSTV has unusual syntax, citing SAT.1. However as I have just explained above, the judge dealt with these points head on and rejected them. His reasons were cogent and I can see no ground on which this court could or should interfere.

[131]Turning back to the judgment, at [113] the judge noted that Google were contending that SIL had effectively conceded that the addition of “TV” did not confer distinctiveness. This was because in the context of non-use SIL had argued that use of the SHORTSTV 2018 Marks counted as use of the SHORTS 2018 Marks and therefore necessarily SIL was contending that the addition of “TV” to the SHORTS 2018 Marks did not alter their distinctive character. The judge held it was a matter for him to determine whether the addition of “TV” confers distinctiveness, rather than to rely on an implied concession but he also observed that SIL did not have any effective answer to this point, and it was consistent with the view he would have arrived at in any event.

[132]Overall therefore the word SHORTSTV was held to be invalid under s3(1)(c) for goods such as video recordings and cinematographic films, as well as a wide range of other goods and services. In his annex 3 the judge set out the goods and services which this finding of invalidity applied to and which it did not.

[133]I have already covered the points made by SIL on appeal on this ground in context. Standing back I can see no justification for overturning this conclusion.

[134]The judge then addressed s3(1)(b) (devoid of any distinctive character) at [119] to [123] and held that this grounds stood or fell with his decisions on s3(1)(c) (consists exclusively of descriptive indications).

[135]Having got this far, SIL’s first Respondent’s Notice point cannot assist SIL. It was addressed to the distinctive character of the element SHORTSTV in the SHORTSTV 2018 Marks and sought to ensure that SIL could not be taken to have conceded the lack of distinctive character of that word alone in the context of the SHORTSTV 2018 Mark. That problem did not arise for SIL and the point does not now make any difference.

[136]At [124] the judge considered whether the 383 Mark had acquired a distinctive character. The relevant date for this mark was 8 February 2022. After accurately summarising the law at [124] to [127] the judge reviewed the factual evidence in detail at [128] to [131]. His conclusions at [132] were that: “[ …] In my judgment, having regard to the evidence to which I have referred, it is clear that exposure to SIL and its marks was largely limited to filmmakers and some other people in the film industry and to film buffs. Even then the evidence is such that it is not possible to assess how many of that group of people had been exposed to any of SIL’s Marks prior to 8 February 2022. In my judgment a significant proportion of the relevant group of persons (i.e. the general public in the UK) had not come to identify goods and services as originating from a particular undertaking because of any of SIL’s Marks.” [as written [132] contained a typographical error in the date]

[as written [132] contained a typographical error in the date]

[137]Although this conclusion is challenged on appeal, as I have shown above in the slightly different context of reputation for the purposes of s10(3), it is quite obvious that the judge reviewed the evidence in detail, made no error doing so and came to conclusions which were open to him on that evidence.

[138]Finally at [134] to [141] the judge dealt with the proposed limitation to exclude “short films”. Again he accurately summarised the relevant cases. On the facts the judge rejected the proposal for three reasons. First because there is no clear and consistent meaning of “short films” (see above), second on the basis of POSTKANTOOR and OMEGA (at [139]) and third (at [140]) that even if “short films” was sufficiently clear (and avoided POSTKANTOOR) the limitation did does not go far enough because on his finding of fact (now unsuccessfully challenged at ground 1) “shorts” is wider than “short films” and so the class for which the mark is descriptive is wider than what would be excluded by the limitation on any view.

[139]On appeal, in my judgment this point can be disposed of without grappling with POSTKANTOOR because the judge was plainly right to refuse the limitation both on the basis that short films is not a sufficiently clear limitation and also because ground 1 having failed, the judge was right that the amendment did not go far enough. SIL’s submission in its skeleton is that “short films are short films” and consumers know them when they see them. That simply does not grapple with the difficulties.

[140]I would therefore dismiss this appeal on grounds 2 and 3.

GOOGLE’S CROSS-APPEAL

[141]Given Google’s clarification in oral argument that its cross-appeal was put on the basis that the court should find the marks were invalid if Google was wrong in its opposition to SIL’s appeal, there is no need to address the cross-appeal.

Conclusion

[142]For the above reasons, SIL’s appeals in respect of both infringement and validity are dismissed.

Phillips LJ:

[143]I agree.

Falk LJ:

[144]I agree.