MONTRES BREGUET S.A. & Ors v SAMSUNG ELECTRONICS CO. LTD & Anor

This is my decision following a liability trial of a claim for trade mark infringement brought by the claimant watch makers against the Samsung Electronics group. The claim relates to digital watch faces (watch face applications, or “apps”) that could be downloaded to Samsung’s smartwatches from the Samsung Galaxy App store (the “SGA store”).

The Claimants are all members of the Swatch group of watch makers. I shall refer to the Claimants collectively as “Swatch”, and the group of which they are members as the “Swatch group”. All of them are incorporated and based in Switzerland, with the exception of the Tenth Claimant which is incorporated in Germany. Each has a distinct brand and associated trade marks, ranging from Swatch in the “basic range” (with prices below CHF 200), Tissot, Mido and Hamilton in the “middle range” (between CHF 300 to 1,000), Longines in the “high range” (CHF 1,000 to 5,000), Omega in the “luxury range” (CHF 3,000 to 10,000) and Breguet, Blancpain, Glashütte Original and Jaquet Droz in the most exclusive “prestige range”, with some prices of those watches running into hundreds of thousands of Swiss francs, or even more. Members of the Swatch group also own other brands, namely Harry Winston, Union, Rado, Certina, Balmain and Flik Flak.

The First Defendant, Samsung Electronics Co. Ltd (“Samsung”), is incorporated in South Korea and is the parent entity of the well-known technology group. The claim against the Second Defendant, a UK subsidiary, was stayed, so Samsung is the only active defendant.

Swatch claim that between October 2015 and February 2019 a total of 23 of their trade marks were infringed by 30 watch face apps being made available in the SGA store, and that those apps were admitted by Samsung to have been downloaded around 160,000 times in the UK and EU. Although each of the apps was developed by a third party (the “app developers”), Swatch maintain that Samsung was intimately involved in, and controlled, the whole process by which the apps were made available. Their primary case is that Samsung is liable as a primary tortfeasor, alternatively that it is a joint tortfeasor. An alternative claim for passing off, covering a further six apps as well, was not pursued at the trial, and claims for infringement of two additional trade marks with international registrations were also not pursued.

Samsung denies liability. Its defences include that it has not used the marks, that the signs on the apps were not used in a way that indicated that there was any connection with Swatch, and that the conditions for joint liability were not established. Further, it relied on Article 14 of Directive 2000/31/EC of the European Parliament and of the Council of 8 June 2000 (the “e-Commerce Directive”).

It will be necessary to refer at various points in this judgment to the appearance of trade marks or signs on watch faces. There was no dispute that the conventional location of a trade mark on a watch, whether in word or figurative form or both, is generally the centre of the upper portion of the dial. For convenience, I will use the term “Dial Branding” to refer to the appearance of the mark or sign on the watch face.

I would like to thank Counsel and other members of the respective legal teams for their assistance and co-operation, including in providing further written submissions in response to questions from the court arising out of the pleadings in respect of individual apps, and in agreeing the contents of Appendices 1 and 2 to this judgment. I do need to add, however, that because of the way in which the case was presented the court was required to do a significant amount of work to produce those Appendices, despite their content being uncontroversial. I would recommend that, in future, greater consideration is given in advance to what would best assist the court to determine the issues in the case, which in this case included disputes about each app.

Apart from some evidence of the alleged infringements, documentary evidence provided by Swatch principally comprised documents provided under a Civil Evidence Act notice evidencing the background and history of the individual Claimants, material on the luxury watch market and the Swatch group’s place in it, Samsung smartwatch promotional material, and video extracts from Samsung developer conferences in 2018 and 2020.

Samsung’s documentary evidence included material relating to the smartwatch, Samsung’s review process, terms and conditions for app developers, download and sales data and, under a Civil Evidence Act notice, evidence in respect of Apple’s app review process, Apple and Google’s infringement report processes, Samsung’s infringement report process, potentially infringing apps shown on the Apple app store and Google Play, and the full version of the watch designer tool presentation at the 2018 Samsung developer conference.

I heard oral evidence from three witnesses for each party. There was no expert evidence. Swatch’s witnesses were Carole Aubert, Peter Steiger and Sylvain Dolla. Samsung’s witnesses were Loi Le, Younggu Kim and Taegu Lee. Mr Le gave evidence through a Vietnamese interpreter and Mr Kim and Mr Lee through a Korean interpreter. All witnesses gave evidence in person, other than Mr Le who gave evidence by videolink from Hanoi. There were no issues with the credibility of any of the witnesses.

Ms Aubert is an in-house counsel at the Swatch group, where she is largely concerned with brand protection and has particular responsibility for online infringement. Before joining the group in July 2018 she worked for the Federation of the Swiss Watch Industry for 13 years as head of its internet unit, overseeing online enforcement on behalf of members. She also worked during part of that period as a partner in a law firm.

Ms Aubert’s written evidence assisted in providing some context and background to the Swatch group’s approach to brand protection and this claim. I found her oral evidence in cross-examination to be somewhat general and of less assistance to the issues in the case. To some extent this was a reflection of her role in the group, although she appeared also to be less familiar with the details of the litigation than might have been expected.

One theme of Ms Aubert’s cross-examination was the fact that Swatch had neither chosen to sue the app developers, nor had it taken legal action against other watch face app providers, including a relatively well known one called Svartling. A point was also taken about an apparent delay between Samsung providing evidence identifying an apparently infringing app in the Apple store and Swatch taking steps to get it taken down.

These points do not affect the question whether Samsung is liable. Fundamentally, it is up to the Swatch group to determine what if any legal action it should take and against whom, as opposed to relying solely on “notice and take down” procedures (see [70] below). It is understandable that the group takes a selective approach. As Ms Aubert said in her written evidence, its resources are not infinite. It is clear from Ms Aubert’s evidence that Swatch’s decision to bring proceedings against Samsung was at least partly prompted by a perception that Samsung was “not operating at industry standard” in taking proactive steps to prevent infringement.

As regards app developers, I accept Ms Aubert’s evidence that Swatch’s experience has been that the level of detail that had been provided by Samsung about them was unlikely to make attempting to track down and pursue them practicable. I also note that Samsung has taken no action against the app developers either, despite the fact that it could be expected to have more detail, including banking information, and has also had a direct legal relationship with them under the terms and conditions it entered into with them (discussed further below).

Further, as I understood Ms Aubert’s evidence, whilst the Swatch group does make a significant number of requests for allegedly infringing matter to be removed from the internet, it is quite cautious in its approach, such that its requests are generally accepted. I also understood her to draw distinctions between the level of concern Swatch would have about apparently infringing material appearing on (say) a phone, as opposed to their concern about such material appearing on a smartwatch. I would further point out that, as discussed below, Samsung markets its smartwatch for its watch-like qualities, which include a round dial. There are some differences in that respect from (for example) the Apple smartwatch.

Mr Steiger has had a number of roles in the Swatch group since joining it 32 years ago, in particular in the areas of finance and logistics. He is currently the group’s Chief Controlling Officer and is a member of its Executive Group Management Board. He also has specific control of some of the group’s country organisations, including the UK.

Mr Steiger’s evidence reflected his long experience at the group and the knowledge he has gained. He emphasised the importance of the branding and the “Swissness” of the group’s watches. He also explained his understanding of the perspective of customers acquiring watches in the luxury or prestige ranges, who would either be watch enthusiasts or purchasing as a gift, and were looking for a brand that they could trust and with a reputation that would reflect well on them.

Mr Dolla has worked with the Swatch group for 17 years, and was appointed CEO of Tissot in June 2020, having previously been CEO of Hamilton. He is also a member of the Extended Group Management Board of the group’s holding company. As CEO of Tissot he is responsible for brand strategy for the Tissot brand.

Mr Dolla’s evidence was confident and impressive. He gave a clear explanation of brand values including, in particular, the fact that consumers were looking for something timeless, with which they could develop an emotional connection. A deliberate decision had been made not to enter the smartwatch market, in the sense generally understood of a limited life product. Tissot had introduced its own “connected” watch with some functions (the Tissot T-Touch Connect), which was very important to the brand, but had chosen to do so by developing its own operating system, intended to operate on a long term basis on a “beautiful watch”, and to run for months without conventional recharging (by using solar power). The group had declined approaches that major technology companies had made seeking collaboration.

Mr Dolla’s concern about the impact of the watch faces complained about appearing on Samsung smartwatches was very evident.

Mr Le has worked at Samsung since 2012, at the Samsung Vietnam Mobile R&D Center (“SVMC”) in Hanoi. He is now a senior software engineer, and his main responsibility is to manage the Samsung Galaxy App validation process. His team, currently comprising 14 people, is responsible for the content review of the apps uploaded to the SGA store by app developers. He reports to the App Review Team in Korea, managed by Younggu Kim.

Mr Le’s responses were straightforward, clear and informative about his team and their working practices. He also candidly accepted that it was a mistake to have allowed the apps in dispute through the review process, and explained that at the time his team had little knowledge of luxury brands.

Mr Kim is a senior professional at Samsung in South Korea, whose role includes managing the app review process for the SGA store and overseeing training in respect of it. His background is in electronic engineering. As already indicated he is responsible for the App Review Team at the SVMC. He has been involved in app review work since 2011, well before the SVMC team was established in 2016.

Mr Kim’s evidence about Samsung’s approach to content review, and in particular the intellectual property aspects and how its processes have developed, was, like Mr Le’s, helpful and informative. His evidence also made it clear that the apps in dispute would have been rejected if the SVMC team had known about the Swatch brands.

Mr Lee is a principal professional at Samsung, based in South Korea. His role is specific to smartwatch apps, overseeing how they are uploaded to the SGA store and having responsibility for Samsung’s polices relating to smartwatch apps in that store, including benchmarking against competitors such as Apple and Xiaomi. He has worked in content management since 2010.

Mr Lee’s evidence in cross-examination was in some respects a little less straightforward than that of Mr Le and Mr Kim. He did candidly accept that a smartwatch is a watch, but maintained his written evidence that the difference between an analogue watch face and a watch face app on a smartwatch was easily perceived, despite agreeing with Samsung’s 2020 marketing material for its Galaxy watch referred to at [45] below. He also suggested that it was rare for the “always on” feature (see below) to show a brand name, despite a substantial proportion of the apps in dispute doing so (a fact that he accepted). As discussed below, I also had some difficulty with his evidence – which I consider in any event to be an expression of opinion – that most consumers would identify that apps were sold by third parties before deciding to download them.

Article 9 of the EU Trade Mark Regulation ((EU) 2017/1001) (the “EUTM Regulation”) provides as follows:

For the purposes of these proceedings it was not suggested that there was any material difference between the wording of Article 9(2), which applies to EU marks, and its equivalent in domestic legislation, s 10(1)-(3) of the Trade Marks Act 1994 (“TMA 1994”) (Footnote: ¹).

It is uncontroversial that, to establish “double identity” infringement under Article 9(2)(a) (equivalent to s 10(1) TMA 1994), six conditions must be satisfied: (i) there must be use of a sign by a third party within the relevant territory; (ii) the use must be in the course of trade; (iii) the use must be without the consent of the proprietor of the trade mark; (iv) the use must be of a sign which is identical to the trade mark; (v) the use must be in relation to goods or services which are identical to those for which the trade mark is registered; and (vi) the use must be such as to affect or be liable to affect the functions of the trade mark.

To establish infringement under Article 9(2)(b) (equivalent to s 10(2) TMA 1994), six conditions must also be satisfied. The first three are the same as under Article 9(2)(a) ((i) to (iii) above). The remaining three are: (iv) the use must be of a sign which is at least similar to the trade mark; (v) the use must be in relation to goods or services which are at least similar to those for which the trade mark is registered; and (vi) the use must give rise to a likelihood of confusion on the part of the public.

Infringement under Article 9(2)(c) (equivalent to s 10(3) TMA 1994) requires nine conditions to be satisfied: (i) the trade mark must have a reputation in the relevant territory (the EU for EU registered marks, and the UK for UK registered marks); (ii) there must be use of a sign by a third party within the relevant territory; (iii) the use must be in the course of trade; (iv) it must be without the consent of the proprietor of the trade mark; (v) it must be of a sign which is at least similar to the trade mark; (vi) it must be in relation to goods or services; (vii) it must give rise to a “link” between the sign and the trade mark in the mind of the average consumer; (viii) it must give rise to one of three types of injury, that is to say, (a) detriment to the distinctive character of the trade mark, (b) detriment to the repute of the trade mark, or (c) unfair advantage being taken of the distinctive character or repute of the trade mark; and (ix) it must be without due cause.

It was not disputed that the relevant dates for assessment of the infringement claims were between October 2015 and February 2019, when Swatch contends that the uses complained of commenced (Levi Strauss & Co v Casucci SpA (C-145/05) [2006] ETMR 71 at [17]-[20]).

Details of the 23 trade marks in issue are set out in Appendix 1 to this judgment, including the goods that they are most relevantly registered for and on which Swatch relied.

Details of the 30 apps in issue are set out in Appendix 2. By way of explanation:

As already mentioned, the apps were made available in the SGA store at various times between October 2015 and February 2019. A number were available for free and others required a modest payment. Samsung admitted that during that period the 30 apps now in dispute were downloaded 157,715 times in total by end users in the EU. With the additional six apps the subject of the original claim, Samsung’s figure for total downloads is 163,769. Swatch accepts Samsung’s figures for the purposes of determining liability only. Not surprisingly, apps available for free were generally downloaded in greater numbers.

The Swatch group began monitoring the SGA store in December 2018, when it became aware that potentially infringing apps were available in it. Swatch first made contact with Samsung’s Swiss subsidiary in that connection by a letter dated 21 December 2018, which demanded that the apps, together with any other infringing apps, be taken down and sought disclosure of app developers’ details and details of downloads. The letter referred to trade mark infringement but also made additional allegations, including in respect of copyright in design features. By 9 January 2019 all the apps identified by Swatch had been taken down.

Swatch issued proceedings in February 2019. The claim identified some additional apps not referred to in the correspondence. These were then also promptly taken down by Samsung.

Mr Lee described a smartwatch as a wearable electronic device with a digital interface that provides operational controls through software apps. Smartwatches tell users the time but also come with a number of additional features not provided by traditional watches, such as the ability to make calls, listen to music, read messages or emails, make payments and track health and activity levels. Mr Lee said that Samsung smartwatches function “more like a wearable computer or mobile phone than like a traditional watch”, and said that they were predominantly purchased for their ability to do things other than display the time. The fact that smartwatches are purchased not simply to tell the time was supported by documentary evidence and is in any event obvious.

The face of a smartwatch consists of a digital display which, unless an “always on” feature is enabled, is switched on when prompted, whether by motion or tapping the screen. When switched on the default screen is a watch face which tells the time, although features such as incoming calls may be displayed instead. If an “always on” feature is available and enabled then a watch face will be displayed continually, although to preserve battery life and avoid screen damage the version of the watch face shown will generally be a simpler version as compared to the one displayed while the smartwatch is in active use, using around 15% of the number of pixels. If that feature is not enabled the watch face will change to a blank screen after a set number of seconds, the default being 15 seconds.

Samsung smartwatches allow significant flexibility for the consumer to configure the software on their smartwatch as they choose, including by accessing the SGA store to download apps.

At least during the relevant period, a Samsung smartwatch did not connect directly to the internet. Instead, it relied on a wireless connection to a smartphone with which it was linked (although not necessarily a Samsung model of a smartphone). In order to create the connection the end user would have to download the Samsung Wear app to the phone.

Samsung first launched a smartwatch in 2013. From 2018 Samsung’s smartwatch models have been called the “Galaxy Watch”. Models launched before then were called the “Samsung Gear” or “Galaxy Gear”.

Since 2015 smartwatches produced by Samsung have been round in shape, like a conventional watch, and with a bezel (Footnote: ²) that also has the appearance of that for a watch. They are held on to the wrist by a strap, again like a conventional watch.

Samsung’s smartwatches are, and were at relevant times, specifically marketed by it for their watch-like qualities. For example, I saw marketing material from 2020 describing the Samsung Galaxy watch as “truly watch-like”, with “ticking sounds and hourly schemes to provide a realistic watch-like feel”, and with a “rich watch face design created with depth and lighting effect”. The advertisement adds that, in addition to the preloaded styles, “about 60,000 styles can be downloaded from the Galaxy Apps” (Footnote: ³). Earlier marketing material included: an undated video promoting the Samsung Gear S3 design, describing it as a “true watch” and depicting a watchmaker, work bench and tools; a webpage from 2016 advertising the launch of that product as redefining the smartwatch experience by injecting “real watch” value with designs “firmly grounded in timeless watch aesthetics”, emphasising Samsung’s “drive to add new value to the smartwatch experience by creating pieces that are watches, first and foremost” and referring to the innovation of an “Always on Display” which “greatly adds to the device’s ‘real watch’ aesthetic”; and a 2019 webpage advertising the Samsung Galaxy watch as having “the feel of a real watch” and describing it as having a “round face, watch hand shadowing and ticking sound [to] give it the feel of a real watch”.

Samsung’s business model involves producing and marketing the hardware (in this case the smartwatches) pre-installed with a comparatively small number of apps designed by Samsung so that the hardware functions. This will include around 20 watch faces that the consumer can choose from. Samsung also offers some further apps that it has developed, including watch face apps, in the SGA store. However, a substantial majority of apps available in the SGA store are developed and provided by third-party app developers. Mr Lee described this as typical of businesses operating in the electronics industry. He also gave evidence that, based on the available figures, only 26% of UK Samsung smartwatch users had made active use of the SGA store, with those who did downloading an average of 4.6 watch face apps each. Further, about 15% of watch face apps are made available for free, frequently becoming paid for apps if they gain popularity.

Samsung smartwatches are generally in use for a relatively short period. The average replacement cycle for a Samsung smartwatch in the UK is about 27 months.

The SGA store was established and is managed by Samsung, specifically for Samsung devices. It is an online platform through which both Samsung and third-party app developers can make available apps to consumers who have Samsung products, including but not limited to Samsung smartwatches.

The SGA store comprises both “back-end” and “front-end” servers. The back-end comprises a developer portal where app developers can upload apps and an administration site where Samsung reviews apps before they are made available. The front-end is the part that is available to consumers.

Also available to developers via the back-end portal is the “Galaxy Watch Studio tool”. That tool provides user-friendly instructions on how to create Samsung smartwatch watch face apps that work on Samsung’s Linux-based “Tizen” operating system, for uploading to the SGA store.

Once an app has been designed and uploaded to the SGA store, it is automatically submitted to Samsung’s app review process, which must be passed before it is made available to consumers as a “for sale” version on the front-end server. The review process is discussed further below.

In order to access the SGA store consumers must set up an SGA store account with Samsung, which involves providing a name, email address, country of residence and phone number (plus optional payment details, although those can be provided when a purchase is made). Watch face apps available in the SGA store can only be viewed or downloaded by consumers with a Samsung smartwatch, which must also be connected to a smartphone (see above). Once connected, watch face apps may be viewed and downloaded to the smartwatch by using either the phone or the smartwatch. The process for accessing the SGA store and obtaining apps is discussed further below.

The SGA store has a search function, although Ms Aubert’s evidence was that it was restricted to the app name and that Swatch did not find it helpful in identifying potentially infringing apps. Ms Aubert also criticised the fact that the contents of the store could not be viewed from a desktop browser and that there was no system of alerts when new apps were added containing keywords corresponding to previous searches. Rather, Swatch had to buy Samsung’s product and then try to investigate using a phone and smartwatch, conducting manual searches without the benefit of any alerts and also with no means to search using picture recognition. Further, the apps available could differ between countries, and it was necessary to specify a country when setting up the required Samsung account, potentially requiring multiple accounts to allow comprehensive searching.

I accept this evidence. I also note and accept Mr Lee’s evidence that consumers may search for other features apart from keywords, such as styles and colours, and by using a drop-down menu (none of which would be particularly useful for Swatch), and his explanation that search results will not necessarily be comprehensive because unpopular apps may be excluded. I further note that Swatch commissioned a report from a third party, the Mintz Group, which confirmed the existence of some difficulties in searching the SGA store, including an inability to search all watch faces at once rather than within specified categories, and common misspellings.

During the period in issue, any third party who wished to upload an app to the SGA store was first required to obtain an account by registering with Samsung’s SGA Seller Portal by providing their name, email address, country of residence and bank account, and agreeing to Samsung’s terms and conditions. The terms and conditions included, among other things:

These terms and conditions were applicable in respect of all apps uploaded to the SGA store, including but not limited to watch face apps.

As already mentioned, Samsung has organised developer conferences, which are available online. Their aim clearly includes the provision of guidance about developing apps for the SGA store, with a view to ensuring that a good range of high quality apps is available.

Each third-party app uploaded to the SGA Store is, and was at relevant times, subject to both a technical and a content review prior to being made available to consumers. The technical review is limited to checks for viruses and to ensure the app’s functionality and compatibility with Samsung’s operating system. The content review was historically conducted by a team based in Poland, but in 2016 a new team led by Mr Le was set up as part of the SVMC, and that team has since carried out all content review globally (covering apps for smartphones and tablets as well as watch apps), other than for apps released to the Chinese market.

Mr Le’s team comprises around 14 individuals, all of whom are local Vietnamese and were generally recruited as new graduates. Most of the team, including Mr Le, are software engineers by background. My impression is that none of them have international or European market experience, and that while all have some level of English they are not fluent English speakers. Mr Le accepted that they would have had little knowledge of luxury brands when they were recruited. Mr Le’s own knowledge of famous watch brands at the time appears to have been limited to what he had picked up from sports-related advertising.

As an indication of scale, in 2020 Mr Le’s team reviewed 319,000 apps, including 89,000 smartwatch apps. From January 2018 to March 2019, each member of Mr Le’s team reviewed 85 apps per day on average, working a 44 hour week. I note that that works out at a little over six minutes an app. Team members are assigned a number of apps each day and are expected to complete their review within three days. They are instructed to spend about 5-10 minutes reviewing each app against Samsung’s internal content review guide (the “CRG”, discussed below), using their own knowledge and judgment and referring to the internet (although see [69] below on the last of these points). Mr Le’s evidence was that 5-10 minutes was sufficient for experienced reviewers.

Mr Le’s team base their review on metadata, app names, tags, descriptions and screenshots provided by the app developer. They do not ordinarily actually review the apps as they would appear on a device. There is an ability to escalate borderline cases to be determined by the App Review Team in Korea, but based on Mr Le’s evidence this facility is rarely used.

Once an app has passed the review process it will go live in the SGA store within about two hours, unless the app developer had manually chosen to distribute it at a later date. The app developer would also determine the price (within permitted parameters) or whether the app was to be available for free.

The content review process has undergone an evolution since the commencement of these proceedings. Prior to February 2019 it was limited to a manual and visual review of app content conducted by Mr Le’s team with reference to the CRG. From February 2019 onwards an additional stage of automated “pre-review” was added to the process in the form of Samsung’s Validation Automation System (“VAS”). VAS has not been used to scan existing apps. It applies only when new apps are reviewed prior to being made available.

VAS is a computerised process that automatically cross checks a proposed app’s meta tags (source code information), title, app description and the developer’s name against a list of keywords, with the aim of identifying content that could be infringing or prohibited. As of February 2022 the database of keywords checked for by VAS comprised approximately 4,000 pre-defined trade marks, brands and phrases. It includes watch brands such as those owned by Swatch. However, VAS will only check for exact matches. Any app with a word or phrase that matches the keywords is flagged for manual review, in addition to the manual content review which continues to apply to all apps. Mr Kim gave evidence that Samsung has not received further claims of trade mark infringement since VAS was introduced, although Ms Aubert disagreed with this and pointed to an additional list of potentially infringing apps sent to Samsung in December 2020.

In addition to VAS, Samsung has also introduced a “Closed-Seller” model for watch face app developers, such that only pre-approved sellers, who have signed a specific pledge not to infringe intellectual property rights, can register to sell apps. It has also introduced an automated visual comparison of watch face app displays against conventional watch faces, although the recognition rate is quite low. Since 2021 app developers have also been required to submit any “brand phrase” (which I take to mean any Dial Branding, see [6] above) in the app description, which VAS can then review.

In considering whether an app should be admitted to the SGA, Mr Le’s team are expected to rely on the CRG.

The version of the CRG produced in evidence is 40 pages in length. It has four sections: 1) performance, which includes functionality and usability; 2) restricted content, which covers inappropriate material such as content that is sexually explicit, violent or drug-related; 3) legal; and 4) culture or market, which covers matters specific to certain countries.

The legal section covers 14 pages, of which most is devoted to intellectual property. It is notable that there is a detailed section covering the circumstances and context in which references to Samsung’s own protected marks and phrases are permitted, with a table of screenshots providing examples as well as relatively detailed guidance. The section devoted to third-party rights holders (referred to as “copyright owners”) contains much less detailed guidance. It provides pictorial examples of a number of well known sports club logos, cartoon characters, games brands and “copyrighted brands”, and describes the process for reporting infringing content. No watch brands are included and there is no specific guidance on apps for smartwatches. The substantive guidance amounts to little more than the following:

It can be seen from this that there was no clear instruction to do an internet search of any kind, at least in respect of what the reviewer did not regard as “well known contents”, and certainly no instruction to check any trade mark database.

Like other online marketplaces the SGA store operates a “notice and take down” process, whereby anyone can notify Samsung that a particular app infringes intellectual property rights. Samsung will investigate (suspending sales of the app while it does so) and, if appropriate, will remove the app from the store. There was no substantive dispute about the effectiveness of this process at trial.

Samsung smartwatch owners are first introduced to the SGA store during the set-up phase, with a “welcome” screen. By clicking on the icon the user is taken to a screen entitled “Legal information”, containing terms and conditions which the user must accept in order to proceed, and commencing with the words “All content made available to you by the Samsung Galaxy Store is subject to the terms and conditions set forth below”.

Once accepted, the user is presented with a range of icons on the watch face, one of which is the SGA store. On accessing the store different options are presented, one of which is “Watch Faces”. Within this users can scroll between different watch face apps, or make searches. During this process they will see the name of the app, a rating, a small icon previewing its design and an “install” button allowing for installation directly from that screen. No information about the seller of the app is provided at any stage of the process just described, and Samsung’s own apps and third-party apps are not distinguished. However, by clicking on the app name rather than pressing “install” at this stage a new screen will be shown which additionally includes a previously unseen name below the app name. That name is the name of the app developer, although that fact is not specified anywhere on the screen. (Footnote: ⁴)

Users may also search for watch face apps via their smartphones, if connected to a Samsung smartwatch. The phone version of the SGA store app allows users to select from different categories of apps, one of which is dedicated to watches and displays only watch apps. Again, they can scroll or search. In this case the details that will be displayed with the app include the name of the app developer as well as the other details shown on the watch, although again without specifying what the name represents. However, it is fair to say that, because more than one app can be previewed at once on a phone, the variety of different names, and the fact that on some of them Samsung’s own name might appear instead of a third-party name, does more to flag up that the name may relate to the source of the app. Clicking on the app name brings up a further screen which again contains the app developer’s name under the app name. It also separately contains “Seller info”, but that appears not to repeat the app developer’s name and to relate instead to version number and file size.

This section of the judgment considers the following elements of the tests for infringement under Article 9(2) of the EUTM Regulation and applies them to the facts:

A number of aspects of trade mark infringement are assessed by reference to the notional average consumer. Broadly, that concept is relevant to any assessment required of the perception of the mark or sign. This will include the question whether the sign and mark are identical or similar, whether there is a likelihood of confusion or a link, and what the sign or mark relates to or signifies (such as whether a sign is purely descriptive).

There was no dispute about the principles to apply. In brief summary, the relevant matter should be:

In London Taxi Corporation Ltd v Frazer-Nash Research Ltd [2017] EWCA Civ 1729, [2018] FSR 7 at [34], Floyd LJ (with whom Kitchin LJ agreed) referred to the purpose of a trade mark as being, broadly, to operate as a guarantee of origin, and said that the term average consumer “includes any class of consumer to whom the guarantee of origin is directed and who would be likely to rely on it, for example in making a decision to buy or use the goods…”. That passage was cited by Arnold J in Sky plc & Ors v SkyKick UK Ltd & Anor [2018] EWHC 155, [2018] RPC 5, [2018] ETMR 23 (“Skykick”) at [274], where he went on to make the point at [275] that the average consumer must both be familiar with the trade mark and exposed to, and likely to rely on, the sign.

The level of attention and care displayed by the average consumer is also likely to vary depending on the nature of the goods or services and how they are acquired: Case C-342/97 Lloyd Schuhfabrik Meyer v Kliysen Handel [1999] ETMR 690 at [26]. For example, the average consumer is likely to pay a higher degree of attention in relation to expensive purchases, as opposed to acquisitions of inexpensive goods that are purchased regularly.

Further, the concept of average consumer does not involve a statistical test or averaging of the class. In particular, infringement may be found “if, having regard to the perceptions and expectations of the average consumer, the court concludes that a significant proportion of the relevant public is likely to be confused such as to warrant the intervention of the court” (per Kitchin LJ in Comic Enterprises v Twentieth Century Fox [2016] EWCA Civ 41; [2016] FSR 30 (“Comic Enterprises”) at [34]). As Kitchin LJ also explained, the assessment must take into account the distinctiveness of the trade mark. There will be a greater likelihood of confusion where the trade mark has a highly distinctive character.

Samsung submitted that the average consumer in this case is a member of the general public seeking to purchase an app for use on a Samsung smartwatch that they already possess, and that they would be familiar with app stores, including that many of the apps are developed by third parties.

I agree with Samsung that, when considering the offering of apps in the SGA store, the average consumer is a member of the general public seeking to acquire an app, and given that the apps were only available to those with a Samsung smartwatch, the average consumer must be in that category of person, and will be viewing the app either on their Samsung smartwatch or on a phone connected to their smartwatch. However, I am not persuaded that such a consumer will necessarily be particularly familiar with app stores, or have any clear understanding that the majority of apps available in the SGA store (in particular, watch face apps specifically designed for Samsung smartwatches) will originate from third parties unconnected with Samsung, and not from Samsung or, as discussed further below, from Swatch.

Further, in this case the concept of average consumer is also relevant in respect of the apps in their downloaded form. In that context, I agree with Swatch that the relevant average consumer is the general public, being persons who acquire watches and/or smartwatches. As explained by Birss J in Thomas Pink Limited v Victoria’s Secret UK Limited [2014] EWHC 2631 (Ch); [2014] ETMR 57; [2014] FSR 40 (“Thomas Pink”) at [118iii)], the average consumer also represents consumers at all levels of the market, ranging from low end to luxury.

In both cases, the average consumer must be assumed to be familiar with the relevant Swatch marks and exposed to, and likely to rely, on the sign (Skykick, above).

What constitutes “use” in an online environment was considered by the CJEU in joined Cases C-236/08 to C-238/08 Google France v Louis Vuitton Malletier [2010] ECR I-0000; [2010] RPC 19 (“Google France”), Case C-324/09 L’Oréal v eBay [2011] RPC 27; [2011] ETMR 52 and more recently in Case C567/18 Coty Germany v Amazon Services Europe [2020] ETMR 37(“Coty”).

Google France concerned Google’s “AdWords” service, which allows advertisers to identify keywords that, if entered by an internet user, will produce a prominently displayed “sponsored link” with a short commercial message, in return for a fee per click. In summary, the claims were brought on the grounds that the Claimants’ trade marks had been infringed by identical signs being used as keywords by competitors to promote their products.

The court concluded that, while the advertisers were using the relevant signs and could be liable if the internet user could not readily determine the origin of the goods, a service provider such as Google did not itself use them by storing the keywords and organising a display of advertisements. This was because “use” required, as a minimum, that an entity “uses the sign in its own commercial communications” (paragraph [56]). Google was simply allowing its clients to use the signs.

This approach was endorsed in L’Oréal v eBay. That was a reference from the High Court in proceedings brought by L’Oréal alleging trade mark infringement in respect of goods sold on eBay’s site that were either counterfeit, not intended for sale or not intended for sale in the EU (parallel imports). Among other things, eBay had used Google’s AdWords service to advertise some of the products for sale on its site, creating sponsored links to the eBay site.

The court held that, whilst an operator of an online marketplace such as eBay could be restrained from advertising via a keyword identical to the mark in question where the origin of the goods could not readily be determined (paragraphs [84] to [97]), insofar as the signs were used in offers for sale displayed on the operator’s own website it was not “using” them. The court said this:

Coty concerned a trade mark infringement claim brought by a perfumes distributor against Amazon in relation to its activities of storing infringing products and handing them to delivery companies on behalf of third parties. The court made the point at [37] that “use” involves “active behaviour and direct or indirect control of the act constituting the use”. After commenting that only a third party with control of the act was able to stop it, the court said this (citations omitted):

At [44] and [45] the CJEU pointed out that Article 9(3)(b) of the EUTM Regulation and its predecessor specifically covered offering of goods, their being put on the market, or their being stocked “for those purposes”, and that it followed that it was necessary for the economic operator providing the storage itself to pursue the aim of offering the goods or putting them on the market. It added:

The CJEU found that Amazon had not used the sign in its own commercial communication, because only the third party intended to offer them for sale or put them on the market (paragraph [47]). It therefore concluded that an entity storing goods for a third party without knowledge of infringement was not “stocking” them if it was not itself pursuing such an aim (paragraph [53]).

Samsung contends that it did not use the signs at all. It simply provided a vehicle, in the form of the SGA store, through which app developers provided apps.

Mr Alexander submitted that none of the individual acts carried out by Samsung amounted to use in a trade mark sense. Validating watch face apps for upload (including undertaking the content review) did not amount to use at all, let alone use in relation to relevant goods or services. It was the app providers, not Samsung, who placed the apps in the SGA store by uploading them. Downloads of apps were procured by customers for their own personal use, and not by Samsung, and it was the customer who affixed the watch face by installing it, having acquired it from the app developer. He placed particular reliance on the Advocate General’s opinion in L’Oréal v eBay at [119]-[120], which was referred to by the CJEU in that case at [103], to the effect that use and display of signs on the website of an electronic marketplace is not use by the operator, any more than their use in classified newspaper ads is use by the newspaper.

Mr Malynicz sought to distinguish Google France, L’Oréal v eBay and Coty on the basis that there was both active behaviour and control by Samsung, and use in its own commercial communications. He relied on Cosmetic Warriors Limited & Anor v Amazon.co.uk Ltd & Anor [2014] EWHC 181 (Ch), [2014] FSR 31 (“Lush”), where John Baldwin QC (sitting as a High Court judge) found that Amazon had infringed Lush marks in circumstances where typing “Lush” into the search facility on the Amazon site resulted not only in the display of goods for sale by third parties where sales would be fulfilled by those third parties, but also products sold by Amazon or where it provided fulfilment services. The judge concluded that no distinction should be drawn between the three categories, because they were mixed up and Amazon’s involvement in the transactions in the other categories was sufficient. The search engine was designed to maximise the sale of goods from the site, and Amazon was doing much more than enabling customers to display signs on its website (see paragraph [57]).

I have reached the conclusion that Samsung did use the signs in the course of its trade for the purposes of Article 9 of the EUTM Regulation. I should emphasise that, in reaching that conclusion, I have taken all the evidence relevant to this issue into account. Many of the features referred to below (including but not limited to assistance designed to make it easier for app developers to create apps, and providing an environment under which apps could be uploaded, stored and downloaded by customers) would not, or would be unlikely to, amount to use if taken in isolation. But that would not be the right approach. Instead, it must be correct to consider Samsung’s conduct as a whole.

An initial, and overarching, point relates to the Samsung smartwatches themselves and the way in which they are marketed, as “truly watch-like”, and also by reference to the wide variety of watch face apps available in the SGA store (see [45] above).

Samsung therefore specifically advertises the availability of a wide range of watch face apps in the SGA store. It clearly regards that as a feature worth promoting, with a view to making its products appear more attractive. It has made a commercial choice to design only a limited number of watch face apps itself, and instead to allow and indeed encourage app developers to develop the vast majority.

Mr Lee’s unchallenged witness evidence also showed that there are additional reasons why it is in Samsung’s commercial interests to host third-party apps, and in particular free apps (even though by definition such apps raise no revenue directly either for the developer or for Samsung). As I understood his evidence, Samsung regards it as desirable that free apps are developed and made available that can allow or highlight the use of functions or interfaces that are unique to Samsung devices, and in particular new models of them. In return, app developers can benefit from increased sales of paid-for apps (because the SGA store is configured to enable users to see other apps from the same developer) and promotional opportunities provided by Samsung.

I accept this evidence, which to my mind clearly indicates a symbiotic relationship between app developers and Samsung, going well beyond revenue derived from the sale of apps and benefitting both parties. The revenue raised is comparatively trivial and is clearly not a material motivating factor for Samsung.

No doubt reflecting its commercial interest, Samsung provides material assistance to developers of watch face apps, in the form of the Galaxy Watch Studio tool. That user-friendly tool was used to create all but one of the apps in dispute (the exception having been developed using an open source platform). Further, Samsung hosts developer conferences to inspire and showcase the capabilities of apps built on Samsung devices, with presentations available online so that app developers who did not attend in person can view them. Samsung enters into licensing arrangements with all app developers before they are permitted to use any of its development material or upload apps to the SGA store (see [55] above).

It is also relevant that the SGA store is a store that Samsung operates that is dedicated to apps for Samsung products, and that it reviews all apps for both functionality and content before they are made available in the SGA store. Although customers may be able to download watch faces for Samsung smartwatches from other sites, I find that the average consumer would understand that, unlike watch face apps in the SGA store, watch face apps available from other sites are not “official” and would not be understood to carry any assurance from Samsung that they would work satisfactorily on its products. In contrast, apps available in the SGA store would be understood by the average consumer to be made available by Samsung through its own store, and indeed to carry an implicit assurance that Samsung is satisfied with them and ought to provide assistance in the event that they prove problematic. Indeed, Samsung has specifically marketed its smartwatches by reference to the availability of watch face apps in its store.

Further, in contrast to apps that obviously relate to goods or services provided by a third party and facilitate their provision (the Uber app was the example used at trial, but there are many others), watch face apps do not do so, and do not directly perform any such function. They are also grouped together in the SGA store with watch face apps designed by Samsung. They provide an optional alternative to the Samsung designed watch faces preloaded on the watch.

The function of watch face apps is not to provide third party goods or services, but to adorn the smartwatch, tell the time and allow interaction with other functions. When in use the watch face app is the starting point for user interaction with the smartwatch. It is the interface that provides the “default” screen, but it is clearly not a simple wallpaper or background: it has an active role. Put another way, it can be perceived as a key part of both the cosmetic design and working mechanism of the smartwatch itself, rather than simply as an app that can conveniently be accessed through the watch.

Further, the apps in question were designed exclusively for, and operated only on, Samsung smartwatches.

Whilst the average consumer would understand that, say, the Uber app was provided by a third party and not by Samsung (not least because the app name will proclaim that, and it would also be understood that the Uber app is available on multiple devices), that is much less clear in respect of watch face apps that are quite obviously designed to fit on, and function as a key part of, a Samsung smartwatch, and where the apps are grouped in the SGA store with watch face apps designed by Samsung.

Another aspect of content review is the means it provides to Samsung to ensure that only content that it considers to be appropriate for use on its products is made available. This is also linked to customer perception. The average consumer is unlikely to blame eBay itself for shoddy goods acquired through its market place. But a poor quality watch face app designed for the Samsung smartwatch and acquired through the SGA store would, in my view, reflect negatively on Samsung in the mind of at least a significant proportion of average consumers. A clear aim of the content review is to limit the risk of that occurring, and rather to ensure that the apps that make it through the review are good quality apps that will enhance, rather than diminish, customers’ perceptions of Samsung’s products. I note that the CRG specifically refers to the fact that apps should be “valuable, entertaining, unique or informative” (and should otherwise be rejected) and that they should be “designed by considering aesthetic factors”.

I also do not accept Samsung’s arguments that app downloads were simply procured by customers for their own personal use, such that there was no use in the course of trade. As with the rest of the process, their provision to customers was orchestrated by Samsung (albeit that without the other elements this would clearly be insufficient to amount to use by Samsung). Samsung would also field customer complaints about them and provide a level of customer support. It would also share in any revenue raised. In any event, Article 9(3) of the EUTM Regulation is not prescriptive as to the form of use. Samsung’s activities could be described as offering or stocking the goods in the SGA store under Article 9(3)(b), or (by arranging and permitting their download) affixing signs, included in any Dial Branding, to the smartwatch under Article 9(3)(a).

Similarly, the fact that, once an app had passed Samsung’s review, the developer could control when it became available to consumers, and that it did so at a price that was (within limits) of the developer’s choosing, is insufficient to alter the conclusion that there was use by Samsung, for the reasons already given.

Overall, Samsung’s role goes well beyond, for example, the can filler referred to in Coty at [41], and it is far from being the equivalent of a warehouse-keeper who provides storage services (Coty at [42]). It is not just providing the necessary technical environment to allow the apps to be uploaded, displayed in the SGA store and downloaded. It has a clear commercial interest in the watch face apps in the SGA store, their availability, presentation and use by consumers. As regards watch face apps, I do not consider that the SGA store is simply an online marketplace like eBay. Rather, Samsung’s activities go beyond that, just as (albeit with very different facts) eBay’s promotion of products for sale on its website via keyword advertising on Google went beyond the provision of an online market place that allowed customers to display offers for sale on the site.

In my view there was active behaviour by Samsung in relation to the apps and control by it of their availability, and Samsung was using the apps, and the signs contained in them, in its own commercial communications.

Appendix 3 to this judgment contains my findings in respect of identity and similarity in relation to the watch face designs and app names in dispute. This section sets the principles I have applied.

I note that Swatch conceded in its opening submissions that, where Samsung disputed a claim of identity, the case could be tried on the basis that the marks were highly similar rather than identical. I did not understand Samsung to object to this.

The test for identity of marks was helpfully discussed by Laddie J in Compass Publishing v Compass Logistics [2004] RPC 41 (“Compass Publishing”) at [15]- [21], by reference to Case C-291/00 LTJ Diffusion SA v Sadas Vertbaudet SA [2003] ETMR 83; [2003] FSR 34 (“LTJ”) and the Court of Appeal decision in Reed Executive Plc v Reed Business Information Ltd, Reed Elsevier (UK) Ltd [2004] RPC 40 (“Reed Executive”). In summary, a strict approach is taken. Either all elements constituting the mark must be reproduced, without modification or addition, or any differences must be “so insignificant that they may go unnoticed by an average consumer” (LTJ at [54]). In Compass Publishing at [20] Laddie J pointed out the advantage of a double identity claim as avoiding the need to prove likelihood of confusion, and said that that advantage “is only to be secured in cases where the mark and sign are so close that one could be considered a counterfeit of the other”, adding that small differences that ordinary members of the public would not notice save by a close comparison can be ignored.

In contrast, noticeable differences cannot be ignored. In Compass Publishing the difference between “Compass Logistics” and “Compass” could not be ignored, because “Logistics” would be perceived as part of a mark rather than as simply descriptive. It had trade mark impact (Compass Publishing at [41). Similarly, in Reed Executive “Reed Business Information” was found not to be identical to “Reed” because the use of capitals conveyed that “Business Information” was part of the name rather than being merely descriptive and really adding nothing, in contrast to (say) “Palmolive Soap” as compared to “Palmolive” (paragraphs [37]-[40]). Further examples are provided by Hasbro Inc v 123 Nahrmittel GmbH [2011] EWHC 199 (Ch); [2011] ETMR 25; [2011] FSR 21 at [194], where Floyd J concluded that “PLAY DOUGH” could not fairly be excised from “THE EDIBLE PLAY DOUGH”, “COLOURFUL EDIBLE PLAY DOUGH” and “PLAY DOUGH MIX”.

Therefore, the significance of deletions or additions must be taken into account. Alterations that are significant will prevent there being identity, provided they have trade mark impact.

However, another possibility is that the item objected to may comprise more than one separate sign: see for example Samuel Smith Old Brewery (Tadcaster) v Lee (t/a Cropton Brewery) [2011] EWHC 1879 (Ch); [2012] FSR 7 at [88]-[91]. Whether signs are separate will be determined through the eyes of the average consumer.

As regards similarity of marks and signs, in JW Spear v Zynga [2015] FSR 19 at [60] Floyd LJ summarised the approach to take as follows:

The assessment must therefore be based on the overall impression, normally based on the mark as a whole (see further the discussion of likelihood of confusion, below). I would add that phonetic, visual and conceptual similarity, are not always of equal weight (see the discussion in Kerly’s Law of Trade Marks and Trade Names, 16^th ed. at 11-093). While I have considered each type of similarity, I consider that visual similarity generally has the most practical significance in this case. In particular, Dial Branding and app names are designed to be seen rather than heard, and the conceptual significance of a name such as Hamilton or Jaquet Droz is limited.

Whether a sign is used in relation to identical goods or services will depend on the goods or services listed in the registration of the mark and relied on by the claimant in its pleaded case.

Where general terms are used to describe goods or services, they must be interpreted as including goods or services clearly covered by the literal meaning: Article 33(5) of the EUTM Regulation. Goods can be treated as identical where one is a subset of the other: Case T-133/05 Gérard Meric v Office for Harmonisation in the Internal Market EU:T:2006:247 at [29]. However, trade mark specifications are concerned with use in trade, so when construing a word used in a trade mark specification the court is concerned with “how the product is, as a practical matter, regarded for the purposes of trade” (British Sugar Plc v James Robertson & Sons Ltd [1996] RPC 281 (“British Sugar”) at p.289, per Jacob J).

Similarity is assessed by reference to all relevant factors, including the nature of the goods, their end users, their method of use and whether they are in competition with each other or are complementary: Case C-39/97 Canon KK v Metro Goldwyn Mayer Inc [1998] ECR I-5507; [1999] RPC 117 at [23]. Some guidance was provided by Jacob J in British Sugar at p.296, who said that the following factors were relevant:

One element of similarity is complementarity of goods. In Boston Scientific Ltd v Office for Harmonization in the Internal Market (Trade Marks and Designs) (OHIM) Case T-325/06 (“Boston Scientific”) at [82] the Court of First Instance described goods as being complementary where:

The fact that consumers regard a product as a complement of or accessory of another is insufficient, unless it was considered usual for the products to be sold under the same mark: Rossi v OHMI-Sissi Rossi (SISSI ROSSI) (Intellectual property) Case T-169/03 at [63], where aesthetic complementarity between shoes and handbags was held to be insufficient.

Complementarity was recently considered by Joanna Smith J in Axogen Corporation v Aviv Scientific Ltd [2022] EWHC 95 (Ch) (“Axogen”) in the context of medical equipment. She rejected the proposition that the plausible use of goods alongside each other was enough, and concluded at [38] that Boston Scientific was concerned with whether the average consumer may think that responsibility for the goods and services lies with the same undertaking. I respectfully agree.

Swatch’s marks are relevantly registered in class 9 or class 14. Its pleaded case relied on marks registered under class 9 as covering (depending on the mark) “computers worn on the wrist” (Footnote: ⁵), “electronic apparatus incorporating a time display”, “smartwatches” or “smartphones in the shape of a watch”. Its pleaded case relied on marks registered under class 14 as covering (depending on the mark) “watches”, “horological and chronometric instruments”, “timepieces and chronometric instruments; smart watches with extended functionality and connected watches”, or “watches that communicate data to smartphones”. For marks registered in class 9 Swatch claims that there is identity of goods or services, and for marks registered in class 14 that there is similarity (Footnote: ⁶).

Samsung conceded in closing submissions that smartwatches are a form of watch, albeit a “very specialised sub-category”, but its case is that the relevant goods in this case are the apps, and not smartwatches or watches. Swatch had not registered any marks for apps for smartwatches or designs for smartwatch faces. Further, making an app available for download was a service and not a good. Swatch’s pleaded case was that the relevant goods were smartwatches, not apps. There was therefore no identity of goods.

For Swatch, Mr Malynicz submitted that there was identity of goods between the apps and smartwatches on the basis that watch face apps are smartwatch faces, and because they are interactive they are smartwatches in digital form. In any event there was a high level of complementarity between apps and smartwatches.

As I understood Mr Malynicz’s submissions in closing, he accepted that the app when on offer in store was not identical with a smartwatch (although there was a high level of complementarity) but once in operation on a smartwatch it was.

I have concluded that the apps are not identical goods to the “computers worn on the wrist”, “electronic apparatus incorporating a time display”, “smartwatches” or “smartphones in the shape of a watch” for which Swatch has registered marks and on which it relied in its pleaded case. Software was not specified in the registration of most of the marks in dispute and, even where it is specified, it was not pleaded. The description of goods pleaded are clearly tangible items that incorporate software or electronics. Software cannot properly be described as a subset of hardware and is not so regarded in practical terms.

However, I agree that the watch face apps in issue are similar goods to smartwatches, in particular on the basis of complementarity. They are obviously not only intended to be used together but are essential for each other’s operation. The apps were specifically designed for the Samsung smartwatch and appeared in the “official” Samsung store. They would appear to the average consumer to have a common commercial origin. Both could be obtained from Samsung, directly in the case of the apps in the SGA store, and either directly or indirectly via another retailer in the case of the smartwatch. As discussed further below, the display of the app provider’s name in the app store was not sufficient to displace this, whether for smartwatch owners using the SGA store or for consumers otherwise aware of the app’s origin in the SGA store. (See [154]-[155] below.)

This is very different to the example of wine glasses and wine referred to in Axogen at [40]-[41], where an average consumer would not assume a common source.

Further, smartwatches are, as Samsung conceded, a form of watch. They are effectively a subset and on that basis there is identity between smartwatches and watches. In any event smartwatches are at least highly similar to watches. Smartwatches and watches are both marketed to individual consumers (that is, end-users or those buying them as gifts). Further, whilst I saw market analysis suggesting that smartwatches and watches are not in “real competition” and are in separate “product categories”, I nevertheless conclude that they must to an extent compete with each other, since most consumers would choose to wear one or the other and not both, at least at the same time. Importantly, both smartwatches and watches are worn on the wrist, and the function of telling the time may fairly be described as their default function. This is so even if they also perform other functions, and indeed might primarily be used to perform those other functions. They also have a broadly similar outward appearance.

It is worth noting here that it is not only smartwatches that may be used primarily for purposes other than telling the time (a feature that Samsung relied on). In current usage a conventional watch may well be used more as a form of jewellery or as a fashion statement, rather than to tell the time (for which, for example, a mobile phone could equally be used). Of course, the watch face apps were also intended to produce a form of fashion statement when in use.

I would also be prepared to accept that the disputed watch face apps have a level of similarity with watches. When in use, a watch face app will perform the same function as a watch, telling the time on a wrist. The apps in dispute appear to have similar features to conventional watch faces, in particular analogue or digital dials. They were designed, like other Samsung watch face apps, to look like conventional watch faces.

It would also be artificial to describe the offer of the app in the SGA store as the offer of a service of making a watch face app available. What was on offer in the store was the app, which once acquired would function as a watch.

The signs have to be used “in relation to” the goods or services in question. This depends on how the average consumer would understand the marks to be used, taking into account the context: Schütz (UK) Ltd v Delta Containers Ltd [2011] EWHC 1712 (Ch) at [79]-[85]. To take an example used by Briggs J in that case, no one would suggest that using Pirelli trade marks on tyres amounted to use in relation to a Ford car to which they were affixed.

In Case C-17/06 Céline v Céline [2007] ETMR 80 the CJEU explained at [20]-[23] that use in relation to goods or services is use for the purposes of distinguishing them, meaning distinguishing them as originating from a particular undertaking: Case C–62/97 Bayerische Motorenwerke AG v Deenik [1999] 1 CMLR 1099 at [38] (see also Datacard Corporation v Eagle Technologies [2011] RPC 17 (“Datacard”) at [250]). This can include a situation where a sign is not affixed to goods but a link is established between the sign and the goods, rather than (for example) a case where use of a shop name is limited to identifying the relevant business.

Mr Malynicz submitted that where the sign was used in the app name that was use in relation to smartwatch face apps and/or smartwatch faces. However, where Dial Branding was used on the watch face it was used to signify the origin of a watch or smartwatch (for example, as being a conventional Tissot watch, or a Tissot watch in smartwatch/digital form). Further, the preview of the watch face that was available before downloading the app was also use in relation to a watch or smartwatch.

Samsung maintained that there was no use in relation to smartwatches or watches. The apps produced decorations for smartwatches and were purchased by consumers for private enjoyment. They were software, and Samsung only provided the hardware and the service of offering software for downloading. The average consumer would understand that the additional software generally originated from third-party app developers, and in this case the name of the third-party app seller was provided. The average consumer would appreciate that the app seller had produced the app and that the app name simply described the design (like “William Morris” on fabric manufactured by John Lewis). Further, no one would understand the preview (which was very small) as indicating the origin of a smartwatch, but rather that it simply showed what the face would look like. Once downloaded, the app just provided an image of a watch face. It was comparable to wallpaper on a phone or laptop and did not denote the origin of a smartwatch. If there was use in relation to goods at all, it was use in relation to apps or smartwatch face design.

Samsung also maintained that any use was purely descriptive, so that there was no use “in relation to” goods or services at all, even apps. That, linked, issue is discussed in the following section.

In my view the average consumer would understand the use of a sign in an app name seen in the SGA store through a consumer’s phone or smartwatch to be use in relation to that app and, in particular, what it will represent once downloaded. To take an obvious example, an app with the name “Tissot Watch Face” would be understood to denote that the app will produce a Tissot watch face when downloaded to the smartwatch. It is use in relation to the software, by which I mean the app, including in particular the watch face that the app will produce once downloaded (which after all is the point of the app).

The same analysis must apply to any preview of the watch face that could be viewed on the consumer’s phone or smartwatch before downloading the app. It showed what the watch face would look like once the app is downloaded. I do not agree with Mr Malynicz’s submission to the contrary.

In reaching these conclusions I bear in mind that the app name and preview would only be seen in a context where it would be obvious to the average consumer that the smartwatch app was just that, an app for a smartwatch, rather than as denoting the origin of the smartwatch itself.

Once downloaded and used, the position is more complex. The smartwatch owner knows that he or she is wearing a smartwatch and using an app. In contrast, a third party seeing the watch face may not do so.

In my view the Dial Branding on a downloaded watch face, which appears in the location that would be expected on a conventional watch and on a product which Samsung markets for its watch-like qualities, is not used solely or even primarily to denote the origin of the app (although it does that as well for the consumer who is aware of what it is). Rather, the Dial Branding is intended to denote, or at least to appear to denote, the origin of the watch. Further, the smartwatch owner aims to achieve that effect, as does the provider of the app. I note that it is well established that the post-sale context is relevant: see for example Thomas Pink at [142]-[145]; Arsenal v Reed [2003] EWCA Civ 696; [2003] ETMR 73; [2003] RPC 39 at [57]; and Datacard at [277]-[289] (in the context of likelihood of confusion).

I therefore consider that signs that appear on watch faces produced by watch face apps are used in relation to smartwatches, which are at least highly similar to watches.

Double identity infringement under Article 9(2)(a) requires the use to be liable to affect the functions of the mark: see for example Case C-487/07 L’Oréal v Bellure [2010] RPC 1; [2009] ETMR 55 at [58]. The primary function is the origin function, that is the ability of a trade mark at act as a badge or guarantee of origin. Others include investment and advertising.

A use of a sign for purely descriptive purposes is not “use” for infringement purposes because it does not affect any of the interests which the mark is intended to protect (see for example L’Oréal v Bellure at [61]). This point is obviously not limited to claims for double identity infringement, but it is convenient to consider the points together given the way in which Samsung put its case.

In relation to its claim for double identity infringement, Swatch relied on the origin function. The origin function is affected if the average consumer is not able, or not able without difficulty, to ascertain whether the goods or services in question originate from the trade mark owner or an undertaking economically connected to it, or instead from a third party (Google France at [84], discussed by Arnold J in Datacard at [263]-[264], and L’Oréal v eBay at [94]). In both Google France and L’Oréal v eBay the CJEU also pointed out that Article 6 of the e-Commerce Directive requires that the person on whose behalf a commercial communication which is part of an information society service is made must be clearly identifiable (Google France at [86]; L’Oréal v eBay at [95]).

Mr Alexander submitted that the signs were not used to indicate that the apps originated from or had any connection with Swatch. The signs indicated the nature of the designs rather than the origin of the apps. There was no indication of anyone taking the use as being anything other than that the design was similar to one appearing on Swatch’s watches. Where the sign was used in an app name it was clear that it was being used as such. Where it appeared on the watch face design it was just part of the design.

There is significant overlap between these issues and the question of what goods or services the signs were used in relation to, and reference should be made to the discussion of that issue. As regards the points specifically raised by Samsung on function and descriptive use, I found them to be unconvincing.

The apps did not simply produce a depiction of a watch face, like a poster of a Ferrari showing the well-known logo on the car that Mr Alexander used as an example. When downloaded to a Samsung watch, the apps produced a watch face that looked like, and functioned as, a watch. The fact that the app might also allow other functions to be accessed cannot affect that. Any Dial Branding would not be understood by the average consumer as descriptive of the design of the app. Rather it was, and was clearly intended to be, perceived as branding. Similarly, where used in the app name it was not like a label on the poster in Mr Alexander’s example stating “Ferrari”. It was not the name of something that would be understood as, for example, a picture or other depiction of a watch face, but rather as something that was, and would function as, a watch face.

Similarly, I do not accept that the signs would not be understood by the average consumer as an indication of the origin of the app in question. Bearing in mind the apps’ availability on the “official” Samsung site, the average consumer either viewing the app in the SGA store or aware of its origin in the SGA store would be likely to assume that there was some licencing or other economic arrangement with the watch maker that permitted use of the sign, just as with (say) the Uber app there would be an assumption that Uber had permitted the use of its name and app.

The fact that the third party developer’s name appeared in the app details is insufficient to displace that assumption. That name was neither necessarily seen before the app was downloaded (see [72] above), nor was it made clear who the person named was intended to be. In particular it was not made clear that the person named was the app seller rather than, say, a designer or design house. I also find that a significant proportion of average consumers would view and download watch face apps directly from the smartwatch (being the obvious place to view them and determine what app or apps they wish to acquire) as described at [72] above, and would be unlikely to see the developer’s name, or to appreciate its significance if they did. If they focused on a name at all, they would have been more likely to focus on the name of the app.

Combining this with the fact that the app being viewed was available on a Samsung site, I conclude that a reasonably observant user would not have been able to ascertain without difficulty that Samsung did not provide the app under an arrangement with Swatch, or that there was otherwise no economic link between the provider of the app and Swatch. The requirement for clear identification emphasised by the CJEU in Google France at [86] and L’Oréal v eBay at [95] was not met.

As with my findings on similarity of marks and signs, Appendix 3 to this judgment contains my findings in respect of likelihood of confusion in relation to the watch face designs and app names in dispute. The principles that I have applied are set out below.

In order to establish infringement under Article 9(2)(b) Swatch must demonstrate the existence of a likelihood of confusion. The relevant legal principles to apply were summarised by Kitchin LJ in Comic Enterprises (in the registration context) and set out as follows by Arnold LJ in the context of infringement in Liverpool Gin v Sazerac [2021] EWCA Civ 1207; [2021] ETMR 57 (“Liverpool Gin”) at [8]-[9]:

Arnold LJ went on at [10]-[12] to consider the two main kinds of confusion, namely direct confusion (where the sign is mistaken for the mark) and indirect confusion. He pointed out at [12] that indirect confusion is not limited to cases where there is no such mistake but there is a belief that the goods or services come from the same or an economically linked undertaking. He gave a further example, being a situation where consumers are led to believe that there has been co-branding, such that there is an economic link between the proprietor of the sign and the proprietor of the mark.

There is a particular aspect of this that is relevant in this case. Swatch are concerned not only about cases where a Samsung smartwatch might be confused with a watch made by Swatch (or a Swatch version of a smartwatch), but also about consumers potentially being led to believe that Swatch companies have entered into licensing arrangements with Samsung. The latter is relevant to smartwatch owners who may view apps in the SGA store and download them, but also to others who may see or be shown the apps in use and understand that they are apps on a Samsung product.

In Liverpool Gin at [13] Arnold LJ said that “there must be a proper basis for concluding that there is a likelihood of indirect confusion given that there is no likelihood of direct confusion”. As previously remarked by James Mellor QC in Cheeky Italian v Sutaria (O/219/16) at [16], a finding of a likelihood of indirect confusion is not a consolation prize for those who fail to establish a likelihood of direct confusion.

The marks the subject of the claim have all been used, and as discussed further below have acquired a reputation as, branding for watches. I have found that smartwatches are at least highly similar to watches (see [132] above). These are relevant factors in determining likelihood of confusion: see in particular paragraphs (g) and (h) of Kitchin LJ’s summary, and the relevance of context emphasised in Liverpool Gin at [9]. A further important point is that post-sale confusion can be relevant (Datacard at [277]-[289], referred to at [145] above).

In my view the context, including the post-sale context, is of particular importance in this case, in the following respects:

I also remind myself that evidence of actual confusion need not be established. The test is one of likelihood of confusion. It was submitted on behalf of Samsung that the absence of actual evidence of confusion in this case was material in demonstrating that there was no likelihood of confusion. I see that point in principle – indeed in Compass Publishing at [22] Laddie J stated that it is often a rule of thumb that where there is no actual confusion in the market-place there is no registered trade mark infringement – but it is not fatal. Instances of confusion might well not have come to Swatch’s attention. Further, and importantly, Mr Alexander supported his submission on this point by reference to the argument that the average consumer would understand that online platforms provide apps from sources not commercially related to the app provider. In the context of the watch face apps in issue I have rejected that argument: see in particular [101]-[105] and [154]-[155] above.

The evidence relied on by Samsung in this connection included Facebook entries posted by two members of a group called Hamilton Watch Owners, where one posted a picture of their “Hamilton smart watch” and another responded “its not a Hamilton watch. Its a Samsung smart watch with a Hamilton face”. As Ms Aubert pointed out, that was a group dedicated to Hamilton aficionados (and therefore not necessarily representative of the average consumer), but in any event the fact that the smartwatch was identified for what it is does not rule out indirect confusion, in the form of an economic arrangement between Hamilton and Samsung.

Further, I am not persuaded by Samsung’s arguments that it would be very unlikely that a person who was not a Samsung smartwatch owner would see and/or be confused by any of the apps complained of, given that they would not be readily visible on someone else’s wrist and/or that the smartwatch would not be displaying the watch face when they did see it. First, the watch face app would obviously be shown when checking the time, whether for the benefit of the smartwatch owner or for the benefit of anyone else. Secondly, given that the watch face app provides the interface, it would be the screen first seen when activating the smartwatch, including (at least on some models) by motion (such as lifting an arm or flicking the wrist to tell the time or use another function). Thirdly, users who had gone to the trouble of downloading an app with Dial Branding were in my view also likely to have wanted to show others what they were, or purported to be, wearing, whether physically or via social media, and to have done so in fact.

Swatch also seeks findings that the apps complained of breached Article 9(2)(c) on the basis that the use of the signs took unfair advantage of, or was detrimental to, the distinctive character or repute of the relevant marks. Swatch seeks those findings irrespective of any finding that there has been infringement under Article 9(2)(a) and (b), on the basis that it may be relevant to quantum, and also because in its view Article 9(2)(c) encapsulates the real mischief in this case (Footnote: ⁷).

There was no submission at trial that the signs were used with “due cause”, so the only requirements to consider that are specific to Article 9(2)(c) are the reputation of the marks, link and injury.

Arnold J provided guidance on the reputation requirement in Skykick at [307], as follows:

Whether use of the sign gives rise to a link to the trade mark in the mind of an average consumer depends on a global assessment of all relevant factors: Case C-252/07 Intel v CPM [2009] RPC 15 (“Intel”) at [41]. The fact that the sign would call the mark to mind is sufficient: Skykick at [309]. Obviously, the greater the similarity between the mark and the sign and between the relevant goods and services the more likely this is to occur. Other relevant factors include the strength of the reputation of the mark in the context of the goods or services for which it was registered, and its degree of distinctiveness (Intel at [42]-[58]). Provided a connection is made between the mark and the sign, it does not matter that a likelihood of confusion is not established: Intel at [66].

For EU trade marks (and their new UK comparables), reputation in respect of periods before 31 December 2020 is assessed by reference to the position of the corresponding EUTM in the European Union, including the UK: Article 54(5)(c) of the Withdrawal Agreement (Footnote: ⁸) and paragraph 10 of Schedule 2A TMA 1994. The mark must be known by a significant part of the public in a substantial part of the EU, although this may comprise a single Member State: Skykick at [308].

I am satisfied that Swatch has provided sufficient evidence to establish a reputation in respect of each of the marks in issue in the relevant location (namely the EU for all marks other than Tissot Navigator UK141, and the UK for that mark). I reject Mr Alexander’s submission that no such evidence has been provided, or that evidence needed to be provided that was specific to the EU. A significant amount of evidence was provided under a Civil Evidence Act notice covering, among other things, histories of the brands, brand rankings and market analysis, none of which was challenged by Samsung. Further, both Mr Steiger and Mr Dolla provided evidence of reputation which I accept. Mr Dolla’s evidence focused on Tissot, but Mr Steiger’s was more general and encompassed all Swatch brands. Whilst the estimated turnover figures provided in the materials were worldwide rather than for the EU or UK specifically, I am satisfied from the remainder of the evidence, including the witness evidence summarised in the following paragraphs, that sufficient evidence was provided to meet the reputation requirement, which as Arnold J remarked is not particularly onerous.

Mr Steiger’s witness statement explained the significance of the Swiss-made heritage and the development of the Swatch brand as a turning point for the Swiss watch industry, following its reputation risking disappearance after the introduction of quartz watches such as Seiko. He set out details of each of the Swatch group brands and described their target markets. He emphasised the group’s policy not to grant licences for its brands or undertake co-branding, to avoid the risk of dilution. He described the group’s approach to control and management of its brands and the carefully-built images they project, including exclusivity in particular, as being at the core of its business model.

Mr Steiger also explained the group’s approach to marketing, including that all the brands use advertising and PR events (including carefully selected sporting events for some brands, such as Omega for the Olympics), and that some of them use “brand ambassadors” or other methods such as an association with charitable activities. He gave some details of Swatch’s very significant marketing budget at over CHF 1 billion, amounting to 11 to 12% of turnover. He also referred to visual merchandising at the point of sale, and described Swatch’s significant distribution network. He further explained that around 25% of the group’s sales were in Europe, with a similar share of the marketing budget being devoted to Europe (of which around 10 to 12% was spent in the UK). Mr Steiger also explained that marketing in Europe was particularly important because of its association with luxury brands and the fact that Europe is a significant travel destination for travellers from elsewhere, with tourists making a significant proportion of purchases of luxury goods in Europe.

Mr Dolla’s evidence emphasised the significance of brand image, describing this as the main foundation of Tissot’s success, and the differentiation Tissot aims to achieve through an emphasis on “Swissness”, craftmanship and heritage. He explained how Tissot is promoted through sporting events, particularly those associated with technical innovation such as cycling, and Tissot’s heavy investment in marketing, in particular in Europe.

Marks such as those associated with Omega, Longines, Tissot and Swatch are very well-known. Their reputation is obvious. As regards others, I would point out that there is no requirement that some threshold percentage of the public must know the mark. This is particularly relevant to the most exclusive prestige or “niche” marks. It is obviously not the case that the more exclusive the brand the less likely the relevant mark is to have established a reputation for Article 9(2)(c) purposes. Further, I would observe that the developers of apps using identical or similar signs to the marks clearly intended to imitate or at least to pay homage to the relevant marks, including exclusive marks such as Jaquet Droz. That rather demonstrates the existence of a reputation.

I am also satisfied that in respect of all signs which I have found to be identical or similar to the marks the necessary link would be established. I particularly bear in mind the context of the use of the signs, on the Dial Branding of watch faces produced by watch face apps and in the names of watch face apps, and the reputation and distinctiveness of the Swatch marks as watch brands. It is also clear that it would have been the app developers’ intention to create a link.

Swatch relies on all forms of injury under Article 9(2)(c), namely detriment to the distinctive character of the marks (“blurring”), detriment to the repute of the marks (“tarnishing”) and unfair advantage being taken of the distinctive character or repute of the marks (“free riding”).

The existence of injury depends on an assessment of all relevant circumstances: Intel at [66]. Guidance about the different types of injury was provided by the CJEU in L’Oréal v Bellure at [39]-[41]:

The onus is on the trade mark proprietor to establish either actual injury to the mark (being evidence of a change in the economic behaviour of the average consumer), or a serious risk or likelihood that injury will occur in the future: Intel at [38]-[39], [71] and [77]. Samsung maintained that no credible evidence had been provided to establish this, and that any reputation of the brands was unsullied.

The requirement for a change in economic behaviour was considered by Arnold J in Jack Wills Ltd v House of Fraser (Stores) Ltd [2014] EWHC 110 (Ch); [2014] FSR 39 at [81]-[83]. He explained that there was no requirement for evidence of a change in the economic behaviour of consumers of the trade mark proprietor’s goods or services, but there would still need to be some change in economic behaviour (or a serious likelihood of it), in that case a change in behaviour of the defendant’s consumers. Arnold J concluded at [83] that the correct way to approach the question was to proceed by analogy with the approach in the following passage of the CJEU’s judgment in WOLF HEAD (C-383/12):

I also bear in mind that, although it is not permissible for the court to speculate, it is legitimate to draw inferences from the facts: Enterprise Holdings Inc v Europcar Group UK Ltd & Anr [2015] EWHC 17 (Ch) [2015] FSR 22 at [222].

I am satisfied that the use of signs identical or similar to the marks in watch face apps (whether in Dial Branding, or in the app name or the preview of the watch face that could be seen prior to downloading the app) was capable of having a detrimental effect on the character or reputation of the marks and of amounting to unfair advantage for the purposes of Article 9(3).

For brands in the prestige, luxury and high ranges (see [2] above), the impact or at least potential impact on brand reputation resulting from use of signs identical or similar to the relevant marks seems to me to be obvious, and supported by the witness evidence of Mr Steiger and Mr Dolla. They are intended to be, and are marketed as, exclusive, beautifully crafted Swiss-made products, produced (in the case of prestige products) in limited numbers and with an emphasis on heritage and longevity. In contrast, the Samsung smartwatch is a commoditised gadget with a short life and, despite work on its aesthetics, in a wholly different league as a product. The damage can be regarded as a form of dilution or tarnishment.

For Swatch brands in the middle range, similar points apply. Those watches still cost significant sums. Tissot is marketed for its Swiss quality and reliability, Mido as a benchmark for Swiss-made fine watches at affordable prices, and Hamilton as producing watches combining “American spirit with true Swiss precision”. Further, a specific aspect of their reputation is that they are intended to be long-lived, and indeed “timeless”, in sharp contrast to the limited life of a smartwatch. The association of these brands with smartwatches could inevitably have an impact on their reputation, as Mr Steiger explained. It could taint their perceived value and exclusivity.

In the basic range, Swatch can obviously not be regarded as an exclusive product, and it is relatively inexpensive. However, it has been an extremely successful brand which has positioned its watches as fresh and innovative lifestyle and fashion accessories, again Swiss made and exclusive to Swatch. Use of an identical or similar sign on a product that does not conform to that description could have the effect of tarnishing the mark. Alternatively, it could have had a dilutive or blurring effect.

I am also satisfied that the use of signs identical or similar to the marks in watch face apps was capable of amounting to free-riding. Swatch has invested very significant sums in building and maintaining the reputation of its brands, and cultivating their “Swissness”. Samsung markets its smartwatch as “truly watch-like” and indeed has promoted it as a “true watch”, also pointing out the “always on display” feature as adding to the “real watch” aesthetic. It has even collaborated with a Swiss watch designer in working on its designs, and has publicised that fact. Slides from the 2018 Samsung Developer Conference (available online) feature apparently Swiss watch faces, including in one instance an Omega watch face. Although Samsung sought to dismiss the relevance of this on the basis that the presentation was intended to cover technical aspects of watch designer tools and that the presenter made a one-off mistake, this amounted to implicit encouragement to designers to use such features.

Samsung’s approach to promoting its smartwatch was not lost on consumers, as demonstrated by the fact that the watch face apps in dispute were admittedly downloaded around 160,000 times. The availability of images of “Swiss” watch faces on Galaxy smartwatches, including some of the apps in question, on the internet and social media would clearly have been to Samsung’s commercial benefit in encouraging sales of Samsung smartwatches. For example, in one thread on a Samsung forum there was a comment in 2018 by someone who said they loved Omega Speedmaster Moonphase watches, could not buy one at present but had created a replica face using the Galaxy design tool, which led to requests to share it. Another Instagram post in 2020 showed an Omega branded face on a Samsung smartwatch with a comment that “this would do for now” because they could not afford the real thing.

However, there is a separate question as to whether Swatch has proved the existence of some change in economic behaviour.

Mr Steiger fairly accepted in cross-examination that “one-time” infringements that were addressed immediately did not present major issues in terms of protecting brand reputation, but pointed out that some of the apps were downloaded a significant number of times, when compared to sales of watches of the brand in question. Taking Tissot as an example, its reputation was harmed by a smartwatch showing a copy of a Tissot watch face which was not of the same quality as a Tissot watch. Although Samsung criticised Mr Steiger’s calculations as having an incorrect basis, the thrust of Mr Steiger’s point remains. I would add that what to Samsung is tiny or immaterial is not necessarily so for Swatch.

Further, I disagree with Mr Alexander’s submission that Mr Dolla’s evidence expressing concern about consumers choosing to purchase £300 smartwatches instead of a £900 Tissot watch was inconsistent with the evidence that purchasers of Swiss made watches were making an aspirational purchase, buying into heritage and craftsmanship. First, I have already concluded that users who had gone to the trouble of downloading an app with Dial Branding were likely to want to have shown others what they were, or purported to be, wearing, whether physically or via social media. In cases where the smartwatch was recognised as a Samsung product, it is likely that this would have had a positive effect on the perceived attractiveness of a Samsung smartwatch to consumers who could afford that but felt that they could not, or could not yet, afford a desirable Swiss made watch. Secondly, and importantly, Mr Dolla expressed a second concern, namely that consumers would consider that there was an association or partnership between Tissot and Samsung, when there was not, to the detriment of Tissot’s reputation. I agree that this is a valid concern.

However, whilst some of the apps were downloaded in large numbers, others were not. Further, Samsung has taken down all the apps and there is no suggestion that there is any risk of them reappearing in the SGA store. Although existing downloads have not been deleted, smartwatches to which they have been downloaded are in practice likely to be in use for a limited period: see [47] above. In these circumstances I do not consider that Swatch can demonstrate a serious risk or likelihood of future injury, if actual injury cannot be shown.

It is true that, when they were available, the apps are likely to have been viewed in the SGA store by persons who did not download them. However, no numbers are available in respect of that and it is reasonable to infer that there would have been a correlation between the number of downloads of a particular app and the number of times it was viewed – or at least viewed in the detail required to appreciate the existence and nature of the relevant sign – in the SGA store.

I have concluded that injury did occur in respect of apps 7, 8 and 14 in Appendix 2, to the extent that they were downloaded and used. Apps 7 and 8, which were “Tissot” apps offered for free, made up the vast bulk of the downloads (67,444 and 75,175 respectively). App 14, a “Hamilton” app, was downloaded 12,569 times.

In contrast, other apps were each downloaded on far fewer occasions. The total for all those other apps was 2,527 downloads, with by far the largest number being the 1,147 downloads of app 10. Taking a pragmatic approach, and having regard to Mr Steiger’s evidence in particular, I have concluded that these numbers are too low to justify a conclusion that any injury occurred.

I should also clarify that in the light of Mr Steiger’s and Mr Dolla’s evidence my finding relates to the signs as they appeared on watch faces, rather than app names.

In view of the conclusions reached it is not necessary to address Swatch’s alternative case that the app developers were primary infringers, whether under general principles or by virtue of carrying out “preparatory acts” under Article 10 of the EUTM Regulation, and that Samsung was liable as a joint tortfeasor.

Samsung relies on Article 14 of the e-Commerce Directive (“Hosting”), which it says provides a complete defence.

The e-Commerce Directive, which was transposed into English law by the Electronic Commerce (EC Directive) Regulations 2002, SI 2002/2013, was as the name suggests intended to facilitate e-commerce. It was introduced at a relatively early stage of the development of modern internet commerce. Article 14 is part of a group of “safe harbour” provisions in Section 4 of Chapter II of the Directive, entitled “Liability of intermediary service providers”, the legal effect of which is not limited to intellectual property. The other provisions comprise exemptions in Article 12 relating to transmission (“Mere conduit”), Article 13 relating to temporary storage in transmission (“Caching”), and Article 15, which prohibits the imposition of any general monitoring obligation.

Article 14(1) provides as follows:

The corresponding provision of the UK regulations is regulation 19, which provides:

It is also worth setting out certain of the recitals to the Directive:

For the purposes of these proceedings, Swatch has accepted that Samsung is the provider of an information society service, and that the apps in question are information for the purposes of Article 14. It also did not contend at trial that Samsung failed to remove the offending apps expeditiously, although it was pointed out that no steps have been taken by Samsung to seek to ensure the removal of previous downloads of the apps. The remaining disputes related to whether what Samsung did went beyond the storage of information, and whether it had awareness of facts and circumstances for the purposes of Article 14(1)(a).

Article 14 has been considered by the CJEU in Google France, L’Oréal v eBay and (since the UK’s withdrawal from the EU) in Cases C-682/18 C-683/18, Peterson v Google ECLI:EU:C:2021:503.

As already discussed, in Google France the CJEU found that a provider does not “use” the signs by storing the relevant keywords, but it went on to consider Article 14. After commenting on the role of a provider such as Google in transmitting and holding information, it said this at [112] to [114]:

The court went on to state that the fact that Google received payment, set payment terms or provided general information would not deprive it of protection, and that concordance between a keyword and search term did not suffice to create knowledge or control of the data. In contrast “the role played by Google in the drafting of the commercial message which accompanies the advertising link or in the establishment or selection of keywords is relevant” (paragraph [118]). It concluded that Article 14 must be interpreted as applying where a service provider “has not played an active role of such a kind as to give it knowledge of, or control over, the data stored”.

In L’Oréal v eBay, as already discussed, the CJEU also found that there was no “use” of signs by an online marketplace displaying them in offers for sale, but went on to consider Article 14. It said this about the scope of Article 14(1):

The CJEU concluded at [123] that Article 14(1) applies to the operator of an online marketplace where that operator:

The CJEU also considered the knowledge related conditions in Article 14(1)(a) and (b), and said this:

The court concluded as follows at [124]:

Peterson v Google (which does not form part of EU retained law but to which the court may have regard (Footnote: ⁹)) concerned alleged breaches of copyright in respect of the YouTube platform and another platform operated by Cyando. It applied the principles set out in Google France and L’Oréal v eBay, making it clear at [107] and [108] that contribution that extended beyond “merely making the platform available” would not fall within Article 14. However, it noted that the platforms in question do not “create, select, view or monitor” content, and added that the fact that technological measures were implemented aimed at detecting content that may infringe copyright did not amount to playing an active role giving knowledge and control (paragraph [109]). Further, the court explained at [111]-[114] that “abstract knowledge” that protected content was being made available was insufficient to satisfy the condition in Article 14(1)(a). Rather, knowledge needed to relate to specific activities. There was no general obligation to monitor or search for information, and automatic indexing and the provision of recommendations based on users’ profiles and preferences were insufficient to amount to specific knowledge.

Mr Alexander sought to distinguish earlier case law, and L’Oréal v eBay in particular, on the basis that this case concerns an app store, rather an online marketplace where it is possible for the operator to adopt a passive role. This is simply not possible with app stores and would be undesirable. It was inherent in an app store that there had to be reasonably close involvement between the operator of the platform and those developing apps for it. Apps were developed for a particular operating system and had to interact with specific software and hardware. Some level of assistance was essential to ensure that apps work properly and did not damage equipment or affect other software.

Mr Alexander submitted that there were strong public policy reasons to facilitate co-operation. It was clear from case C-298/07 Bundesverband der Verbraucherzentralen und Verbraucherverbände at [15] and [19] that the e-Commerce Directive was required to be interpreted in the light of its objectives, and that it had the aim of contributing to the development of information society services and maximising the opportunities afforded to e-commerce. The Commission had criticised practices making it harder for third party providers to produce and market apps in competition to those provided by the app store provider, and a similar point had been made by a House of Lords select committee. App store providers should not be dissuaded from providing assistance to create apps or from promoting apps of third parties. That context should be taken into account in interpreting Article 14.

Mr Alexander relied on the extensive recitals to the e-Commerce Directive, including recital (41), which makes clear that its aim is to strike a balance between different interests, and recitals (42) to (44) which he submitted suggested that the operative provisions applied to situations where the platform operator did not modify the content complained of (see [201] above).

Further, L’Oréal v eBay indicated that for Article 14(1)(a) to operate a specific alert was needed. Samsung had acted promptly once alerted to the possibility of an issue. It was not reasonable, in the context of the international nature of e-commerce, to expect Samsung to pick up every possible risk of confusion with specific Swiss watch brands, particularly niche ones. Swatch was well-placed to identify any issues itself and use the notice and take-down procedure.

In summary, the court should conclude that Article 14 should not be disapplied as a consequence of co-operation that was reasonably necessary to facilitate the creation and distribution of apps, or by an app store provider seeking to identify and filter out potential issues with apps, as long as it did not design the relevant app or make editorial changes to it. In addition, the court should find that if there was an adequate notice and take-down procedure it would not be “apparent” that an app was problematic for the purposes of Article 14(1)(a) if issues were not identified in advance.

The protection conferred by Article 14 and regulation 19 is limited to activity that “consists of the storage of information”. Further, that concept has been interpreted in Google France and L’Oréal v eBay, which both form part of retained law, in a relatively restrictive manner, as being limited to conduct of a “technical, automatic and passive” nature, rather than an active role that would give the relevant person “knowledge of, or control over” the data. This approach reflects the wording of the recitals, in particular recital (42) (albeit that in terms that recital refers to transmission and caching). It is worth noting that the reference to knowledge and control here is separate from the knowledge-related condition in Article 14(1)(a).

Thus, steps such as optimising the presentation of, or promoting, offers by an online market place would preclude exemption (L’Oréal v eBay at [116] and [123]). Further, whilst there is no obligation to monitor, a failure to act expeditiously following awareness of facts or circumstances that would have alerted a diligent economic operator would also lead to a denial of protection (paragraph [124]).

Based on the approach taken in the existing case law, it is far from clear that what Samsung did in this case was limited to acts of a “mere technical, automatic and passive nature” such that it lacked knowledge of or control over the data. Rather, in addition to the functional and safety review, the SVMC performed a relatively detailed content review with reference to the CRG. As already discussed, only a relatively small part of that guide relates to third parties’ intellectual property rights. Further, in other respects Samsung also took active steps, including in relation to facilitating and encouraging the design of apps, and marketing its smartwatches by reference to the availability of watch face apps in the SGA store. The commercial benefit it derived from doing so was from sales of its products, rather than any form of charge for storage. It has also promoted some of the apps.

Further, a watch face app has relatively limited content. Each app relates to a different watch face (plus, potentially, a related “always on” version of that face). The evidence indicates that whilst a reviewer would not generally download the app on to a watch, they will look at the details, including the app name and screenshots of the watch face. Reviewers are able to, and do, reject apps that are not regarded as meeting Samsung’s policies. As a result, it can be inferred that Samsung, via the reviewer in question, would have obtained “knowledge” of signs that appeared on the watch faces and in the app name, even if it is the case that a gatekeeper role such as that adopted by the SVMC should not be regarded as amounting to “control” of data in the manner contemplated by the CJEU.

However, Samsung has a fair point that the case law does not consider app stores, and it is possible that the different factual context could have some impact on the approach to take to the “knowledge or control” test that the CJEU has applied, a test which is not reflected in the express terms of Article 14(1). I would add that the first part of Article 14(1) should also not be interpreted to deny practical effect to Article 14(1)(a) (see L’Oréal v eBay at [121] and [122] in particular).

I agree that Article 14 and regulation 19 should be interpreted in the light of the objectives of the legislation. I also accept that some level of interaction is required between an app store and third party app providers to ensure that apps can function properly and do not cause problems with equipment or other software. Further, I can see a strong argument that it would be unattractive to penalise operators solely by virtue of having additional review procedures in place that are designed to weed out inappropriate or illegal content. I note in particular that recital (10) to the e-Commerce Directive emphasises the principle of proportionality, and also refers to the need to ensure protection of objectives of general interest, including the protection of minors and human dignity.

In the light of these points I would prefer to base my conclusion on Article 14(1)(a), as interpreted in L’Oréal v eBay. I cannot accept Mr Alexander’s interpretation that the condition in Article 14(1)(a) is not met as long as an app store provider does not design the relevant app or make editorial changes to it. It is clear from L’Oréal v eBay that the test is one of whether a diligent economic operator should have identified the illegality by reference to facts or circumstances of which it is (actually) aware. The existence of notice and take-down procedures does not itself provide a defence.

I am satisfied that the content review process would have resulted in the reviewer, and through it Samsung, becoming aware of the app name and the appearance of the watch face, including any Dial Branding on it.

At the trial, emphasis was placed on behalf of Samsung on the modifications it has made to its procedures since Swatch made its claim (although, no doubt mindful of its case, there was also a more or less explicit complaint by Samsung that it should not have been required to make those improvements and that the Closed-Seller system in particular (see [65] above) could harm free access and competition). I accept Swatch’s case that Samsung’s procedures were inadequate before the claim was brought. The fact that Samsung has made changes since then rather underlines that conclusion, but is not necessary to it.

In the case of those apps that I have found did infringe Swatch’s marks, I have concluded that the awareness of the app name (where relevant) and/or the appearance of the watch face, and specifically any Dial Branding on it, that would have been gained during the content review process would have alerted a diligent economic operator to the infringement.

In particular:

In my view this approach provides an appropriate balance between the different interests. In essence, having determined for its own (perfectly legitimate) reasons to undertake a content review, Samsung is required to behave as a diligent operator would in respect of the content of which it becomes aware in the process. The changes Samsung has made to its processes since the dispute started appear to go some way towards meeting that requirement, although as indicated above there appears to be a bit further to go with the review process.

I should also mention the fact that some of the apps were promoted as “Editor’s Picks” or similar. These were chosen by reference to numbers of downloads and recommendations from an advertising agency acting for Samsung. Promotion is an activity that the CJEU has said is outside Article 14(1), although Samsung maintains that the promotion was automatic, based on download frequency. Given my conclusion on Article 14(1)(a) I make no finding about whether this activity would otherwise have taken Samsung outside the protection of Article 14(1).

One final point I should mention is that Samsung made a number of complaints at trial about what it asserted was a lack of co-operation by Swatch in assisting in identifying and resolving issues, including failing to provide a list of trade marks, and its failure to monitor the SGA store before 2018. None of these matters appear to me to be strictly relevant to liability, and in particular to what steps Samsung itself should have been taking before 2019.

In conclusion Swatch has established infringement of its marks, under Article 9(2)(a) or (b) of the EUTM Regulation, in respect of apps 1 to 18, 20-22 and 26-30, as set out in more detail in Appendix 3. In respect of apps 1, 4, 12, 15, 16, 18, 20-22 and 26-28 the finding of infringement relates to the relevant sign on the watch face only, as opposed to any app name. In respect of apps 11, 13, 17, 29 and 30 it relates only to the sign in an app name. For other apps the conclusion relates to both.

I have concluded that use in app names (or in previews of the watch face) is not use in relation to smartwatches. Therefore, findings of double identity infringement under Article 9(2)(a) are limited to the use of signs on watch faces produced by watch face apps, which does constitute use in relation to smartwatches (or in relation to the “computers worn on the wrist”, “electronic apparatus incorporating a time display” or “smartphones in the shape of a watch” for which Swatch has registered marks and on which it relied in its pleaded case under Part 1 Appendix 1 of the APoC).

I have however concluded that watch face apps are similar goods to smartwatches and that they have some level of similarity with watches, such that infringement is established under Article 9(2)(b) where the other requirements are met, including likelihood of confusion.

Swatch has also established infringement under Article 9(2)(c) in respect of apps 7, 8 and 14, in respect of the relevant sign as it appeared on the face of the watch.

Samsung does not have a defence to the infringements under Article 14 of the e-Commerce Directive.

APPENDIX 1: THE TRADE MARKS IN ISSUE

APPENDIX 2: THE APPS IN ISSUE

APPENDIX 3: CONCLUSIONS ON INDIVIDUAL MARKS AND SIGNS