Engineer.AI Global Limited v Appy Pie Ltd & Anor

This is a judgment after a two day liability-only trial of a trade mark dispute between two competitors in the ‘no code’ application development space. ‘No code’ or ‘code-free’ application development allows users to build or create an app without having to carry out coding, and so simplifies app development and makes it accessible to those without prior experience of creating or developing software. The dispute is about the Defendants’ use of the word “Builder” in conjunction with the trade in no-code app development products and services.

There are different types of ‘no code’ app development platforms. One type is a ‘drag and drop’ style app development product, which enables the user to drag the components they want and drop them into place to build the app on their own. Another type is a composable software platform, where the user specifies elements and functionality, and the platform provider builds the app for them.

The Claimant, Engineer.ai Global Limited is a UK incorporated early-stage software developer working in artificial intelligence and offers a suite of composable software platforms to the market. It was incorporated in 2017 but only obtained seed funding in October 2018 and achieved its first sales in March 2019. It is the proprietor of seven UK-registered trade marks in issue, as set out below (together, the “Marks”).

The Claimant says that all of the Marks have been in use in, and targeted at, the UK since 2019 (i.e. both before registration and afterwards). It says that it trades exclusively under BUILDER (as its house or ‘umbrella brand’), BUILDER.AI and associated sub-brands using BUILDER. It offers products and services through its website at www.builder.ai. It pleads that it is recognised and known within the industry and to its client base in the UK as BUILDER. Its family of products and services comprises:

The Second Defendant, Appy Pie LLP (“D2”), is an Indian LLP incorporated in January 2016. It trades in the development and supply of software, specifically ‘no code’ app builders of the drag and drop type, under the name APPY PIE, through its website www.appypie.com (“Appy Pie Website”) The two partners are Mr Abhinav Girdhar and his mother, Ms Savita Girdhar, who does not play an active role in the business of D2.

The First Defendant, Appy Pie Ltd (“D1”) is a company incorporated in the UK. Its sole director and shareholder is Mr Abhinav Girdhar’s wife, Ms Shagun Gupta, and Mr Girdhar acts as a manager of it.

Mr Howe, Kings Counsel, represents the Claimant and Mr Lomas represents the Defendants. I am extremely grateful to them for their very helpful skeleton arguments and oral submissions which were of significant assistance, and the manner in which they managed the trial so that a very great deal of ground could be elegantly and efficiently covered in a short space of time.

The Claimant claims infringement by the Defendants of certain of the Marks in two categories. What was pursued at trial has narrowed from its pleaded case.

The Claimant claims that the Defendants have been using the Marks (and/or portions of them) to direct consumers away from the Claimant, causing consumers to be confused, for example by the use of the word “Builder” in the nomenclature of the Defendants’ products.

Mr Howe KC for the Claimant in his closing submissions reduced the scope of this category of alleged infringements:

Paragraphs 11.1-11.3 of the APOC read as follows:

The Claimant pleads that this use by the Defendants is in the course of trade in the UK, as the Appy Pie Website is targeted at consumers in, inter alia, the UK. It pleads that the Marks are a family of trade marks comprising the Builder Word Mark along with additions and flourishes. It pleads that use of the word “BUILDER” by the Defendants in conjunction with one or more descriptive words in the field of computer programming, software or hardware will be perceived by the relevant public as one of the family of marks owned by the Claimant (T-194/03 Bainbridge), and so associate the Defendants’ use with the Claimant’s earlier marks forming part of the series, creating a likelihood of confusion.

Further, the Claimant pleads that by reason of the acts complained of, the Defendants infringed the Marks pursuant to s.10(3) TMA as they have, without the Claimant’s consent and in the course of trade, used signs that are similar to the Marks in relation to good and/or services, where the marks have a reputation in the UK and where their use, being without due cause, gives rise to an unfair advantage to the Defendants and causes detriment to the distinctive character of the Marks, sought to take an unfair advantage of the Claimant’s Marks.

The Claimant further claims s.10(1) TMA and s. 10(3) TMA infringement of the Builder.ai Figurative Mark and Builder.ai Word Mark by the use by D2 of Builder.ai in a video that the Defendants posted on LinkedIn called “7 of the best no code app builders in 2022” (“LinkedIn Post”).

The Claimant pleads that the LinkedIn Post:

The Claimant pleads that the Defendants are joint tortfeasors as the acts complained of have been carried out pursuant to a common design between them, in which each participated.

The Claimant seeks an injunction to restrain the Defendants from infringing, a declaration that the registered trade marks are vested in the Claimant and the Defendants have infringed them, damages or an account of profits plus interest, publication of the judgment, and costs.

In their Amended Consolidated Defence and Counterclaim, the Defendants aver that:

In relation to Category 1 infringement allegation, the Defendants admit that the Appy Pie Website is targeted at, inter alia, the UK and that Screenshots 1, 2 and 3 show the use of the signs complained of. The Defendants submit that only D2 has used the signs complained of and that none of the uses:

It further pleads that the Claimant has always traded under the name Builder.ai and never Builder solus, and denies that the public are likely to recognise Builder as a brand name as it is so generic and descriptive. It denies that Builder would be perceived as the Claimant’s house or umbrella brand for a family of marks.

In relation to the Category 2 infringement allegation, the Defendants admit that the LinkedIn Post was posted to D2’s LinkedIn page, and incorporates the Builder.ai Figurative Mark and the Builder.ai Word Mark. They deny that the LinkedIn Post (and the video embedded within it) is targeted at the UK. They deny that the video can be characterised as advertising at all, let alone comparative advertising and in any event say is not misleading and does not denigrate the Claimant’s product. If, contrary to its position, the Court finds that it is comparative advertising and targeted at the UK, it avers, therefore, that it complies with, and is permissible under, the 2008 Regulations.

To the extent that D2’s use might be considered infringing contrary to its case, it relies on a defence under s.11(2)(b) and (c) TMA, because: (i) such use is descriptive and/or necessary to indicate the intended purpose of D2’s product; and (ii) use in the LinkedIn Post is permitted referential use.

The Defendants counterclaim that each of the Marks is and always has been invalid in respect of the goods and services for which it is registered in classes 9 and 42. All seven Marks are challenged for non-compliance with s.3(1)(b) and s.3(1)(c) TMA and the Builder Word Mark and Builder House Mark are additionally challenged under s.3(1)(d) that they “consist exclusively of signs or indications which have become customary in the current language or in the bona fide and established practices of the trade”. At trial, Mr Lomas submits that although it relies on s. 3(1)(b), it does so as a backstop and that reliance does not extend the scope of the Defendants’ counterclaim under s. 3(1)(c) or (d).

The Counterclaim was also narrowed in scope at trial. Mr Lomas in his skeleton and opening submissions said that the Defendants were not seeking invalidation of the entirety of the Marks but were in each case seeking to remove from the relevant specification those parts indicated by him in his skeleton argument (and which specifications and indications I set out in the judgment below), which the Defendants submit are purely descriptive or highly allusive.

The Claimant denies the Counterclaim on the basis that each of the Marks are inherently distinctive and/or had acquired distinctive character by the date of their registration or alternatively had done so before the date of the Counterclaim and so are not liable to invalidation under s. 47(1) TMA.

HHJ Hacon, by a case management order dated 23 November 2023, ordered a split trial of liability and quantum, made directions, and set out in the schedule to the order a list of issues for determination at the liability trial as follows (amended to use the defined terms used in this judgment):

There is no dispute between the parties that the relevant date for the purpose of determining whether a registered trade mark has acquired distinctive character for the purposes of s. 47(1) TMA is the date of the application for invalidity, which in this case is the date of the original (unamended) Counterclaim: 2 February 2023 (per London Taxi Corp Ltd (t/a London Taxi Co) v Fraser-Nash Research Ltd [2016] EWHC 52 (Ch), [2016] ETMR 18 at [179]).

Equally, there is no dispute that this is distinct from the relevant date for assessing inherent and acquired distinctiveness for the purposes of determining whether a registered mark is validly registered, which is the date when the application was filed. That is 2 November 2020 for the Builder Home Mark (UK ‘066) and the Builder Word Mark (UK’010), and 17 December 2021 for the remaining Marks in suit. Evidence relating to facts and matters after that date are inadmissible for the purposes of assessment, save where it is of such a type as to support, or enable the drawing of conclusions about, what the position was as at the filing date (see Case C-192/03 Alcon Inc v OHIM [2004] ECR 1-8993 at [39] to [40]).

The relevant date for assessing whether a mark has a reputation is the date that the proprietor of the mark alleges infringing use began. In relation to Category 1 alleged infringement, no date of first use has been alleged, so that must at the latest be the date of the Claim Form: 17 October 2022. In relation to Category 2 alleged infringement, that is the date of the LinkedIn Post, which was a date unknown in July 2022.

The principles relating to establishing a family of marks were set out by Arnold J, as he then was, in W3 Ltd v easyGroup Ltd [2018] EWHC 7 (Ch) at [234] and [235]

The Claimant also draws my attention to Sir Anthony Mann’s consideration of families of marks in EasyGroup Ltd v Easy Live (Services) Ltd [2022] EWHC 3327 (Ch), [2023] E.T.M.R. 12, saying at [135] “This is a point capable of bearing on issues of confusion and links (under infringement) and misrepresentation (under passing off). Where a proprietor establishes that there is a family of marks, and the challenged sign reproduces some of the elements of the “family”, then the possibility of confusion is enhanced…”.

The Claimant relies on T-194/03 BAINBRIDGE ECLI:EU:T:2006:65 and R172/2013-2 CITY INDEX v CITI for the proposition that the use of the word “BUILDER” by the Defendants in conjunction with one or more descriptive words in the field of computer programming, software or hardware will be perceived by the relevant public as one of the family of marks owned by the Claimant.

A number of propositions as to distinctiveness were set out by Arnold J in Société des Produits Nestlé SA v Cadbury UK Ltd [2014] EWHC 16 (Ch); [2014] ETMR 17 (“Kit Kat I”):

The Court of Appeal considered on appeal Arnold J’s second judgment in the matter (Société des Produits Nestlé SA v Cadbury UK Ltd [2016] EWHC 50 (Ch)) (“Kit Kat II”), which followed his referral to the CJEU of a question on what an applicant for registration needed to prove in order to establish that a trade mark has acquired distinctive character, in which interpreted the answer that was returned. It was upheld ([2017] EWCA Civ 358) (“Kit Kat CA”) with Kitchin and Floyd LJJ providing the following guidance at [52] and [102] respectively (the Chancellor agreeing in a short concurring judgment):

The relevant factors for assessing whether a mark possesses a reputation include: (i) market share (ii) intensity, geographical extent, and duration of use of the mark; and (iii) size of investment in the mark. Reputation is a question of fact to be determined in each case, see for e.g. T-47/06 Nasdaq EU:T:2007:131.

I accept the Defendants’ submission that although distinctiveness and reputation are, strictly speaking, separate concepts, it is hard to conceive of circumstances in which a party that failed to show any distinctiveness can nevertheless ‘get home’ on reputation. This is because reputation requires a certain degree of knowledge of the earlier mark among the public, see C- 375/97 Chevy EU:C:1999:408, at [23].

S.10 TMA provides:

Regulation 4 of the 2008 Regulations provides:

I will consider the law in this area in more detail in my discussion of the issues.

S.47 TMA provides:

S.3(1) is relevant to the Defendants’ counterclaim:

Given the Defendants’ characterisation of s.3(1)(b) TMA as a ‘backstop’ position in this case, I will focus on s.3(1)(c) and (d). The underlying interest protected by s.3(1)(c) TMA is that of ensuring that descriptive signs relating to one or more characteristics of the goods or services in respect of which registration as a mark is sought may be freely used by all traders offering such goods or services: see for example C-51/10 P Agencja Wydawnicza Technopol v EUIPO EU:C:2011:139; [2011] ETMR 34 at [37].

The provisions of s.3(1)(d) TMA overlap with s.3(1)(c) to a certain extent and are directed to preventing the registration of those signs or indications which honest traders customarily use in trade (i.e., signs which are generic). The principles applicable to s.3(1)(d) are now settled by a range of cases such as Case C-517/99 Merz & Krell [2001] E.C.R. I-6959 at [21] to [25] and [35] to [41].

The Supreme Court in Lifestyle Equities CV v Amazon UK Services [2024] UKSC 8 has recently restated the law on targeting for trade mark infringement purposes. At [27] it confirmed that the question of whether an advertisement or offer for sale is targeted at consumers in the UK is to be considered from the perspective of the average consumer, and at [28] that the question is whether the average consumer would consider the website to be directed at him or her. That will require the judge to carry out a multifactorial assessment of all the circumstances, which at [29] are said to include “the appearance of the website, how it responds to the presence of the consumer, whether it is possible actually to buy goods and have them delivered, and how that is done. They may also include a range of other facts and matters such as (but by no means limited to) the Court of Appeal in Merck [Merck KGaA v Merck Sharp & Dohme Corpn [2017] EWCA Civ 1834, [2018] ETMR 10] at para 170, all of which may assist the court to assess the reaction of the average consumer and so answer the question whether the accused advertisements are targeted at consumers in the UK.” Those facts and matters in Merck include the appearance and the content of the website, which are said to be ‘of particular significance’, including whether it is possible to buy goods or services from it, the nature and size of the trader’s business, the characteristics of the goods or services in issue and the number of visits made to the website by consumers in the UK.

The Supreme Court in Lifestyle Equities confirmed at [30] that it is not necessary to establish that the trader had the subjective intention of targeting consumers in the UK, however, if such a subjective intention is established, it can “ease the path to a finding that the foreign trader’s activities, viewed objectively from the perspective of the average consumer, were and are targeted at consumers in the UK”.

The Supreme Court in Lifestyle Equities also clarified the law on joint tortfeasorship, holding that in order for persons to be held jointly liable with a tortfeasor for a tort, they must have knowledge of the essential facts which make the acts wrongful, whether or not the primary tort in question (such as trade mark infringement as it was in that case) is a strict liability offence and whether or not the accessory liability arises from procuring a tort or by a common design. As Lord Leggatt (giving the leading judgment and with whom the other members of the Court agreed) explained at [137]:

As is usual in IPEC, the statements of case stand as evidence. I heard from a single witness from each party at trial:

Ms Gupta has worked within the Claimant’s group of companies since 2019, initially in the role of Head of Marketing in India. She became Global Head of Marketing in August 2021 responsible for the worldwide marketing resource of 22 staff, and moved to London as soon as the pandemic allowed her to, in February 2022, when she became employed by the Claimant company itself. In cross-examination she said that even before joining the London office, she was managing the London marketing team of around 12 people remotely from India. She said she was responsible for all aspects of the Claimant’s brands, including print and event marketing, outside marketing such as billboards and posters, social media and certain parts of the Claimant’s website, including blogs.

Mr Girdhar gave his evidence in his capacity as the Manager of D1 and a Partner of D2.

Both witnesses attended trial (Mr Girdhar remotely over video link) and were cross-examined. Mr Lomas commented that both were strong cheerleaders for their respective companies and products, but made clear that was not a criticism of either, and I accept both points. Both witnesses presented in court as entirely credible people who came to court to provide honest evidence to the best of their recollection and ability, in my judgment. Ms Gupta’s written evidence was problematic in a number of ways which she fairly accepted in cross-examination, particularly in relation to financial and other figures, as I will come on to describe, and so I am unable to accept much of the financial or quantitative information as reliable. It also suffered from a lack of support from documentary evidence which she fairly accepted could have been obtained or provided and disclosed, but was not, in ways which I will also describe. This means that much of her evidence, particularly that which came from other sources such that she could not directly attest to them, could not be validated or tested. Under Mr Lomas’s cross-examination she, again fairly, made a number of concessions and further admissions such that her oral evidence was quite different from what she presented in her witness statement, and overall less supportive of the Claimant’s case. Accordingly although I accept Ms Gupta as an honest and credible witness and one who gave reliable evidence on matters within her own direct knowledge, much of the evidence she puts before the Court is not reliable. In contrast, Mr Girdhar’s oral evidence was entirely consistent with his written evidence, and he was really unshaken in cross-examination. I am satisfied it was both credible and reliable.

The Claimant describes itself as a highly successful software developer with a number of major global brands and innovative entrepreneurs amongst its client base, which has been recognised as such through its receipt of numerous industry awards, e.g. ‘Winner of the Hottest AI Start-Up’ at the 2020, ‘Best COVID-19 Innovation for Recovery’ at CogX 2020 and third place in the 2023 ‘Fast Company’ global list of ‘Most Innovative Companies in Artificial Intelligence’ (after OpenAI, the producers of ChatGPT, and DeepMind). The Claimant also ranked in a list of 5,000 of the fastest growing companies in America. Ms Gupta also notes that the Claimant won the category of ‘Scale Up of the Year’ at the 2022 Europas Hottest Tech Start Up Awards for UK and Europe, and the Claimant’s CEO Sachin Dev Duggal won the 2023 Ernst & Young UK Entrepreneur of the Year award, the Claimant being referred to as ‘Builder.ai’ in all of the announcements relating to the award. None of this evidence was challenged by the Defendants. The services for which the Claimant is known are provided and developed through the Claimant’s website, https://www.builder.ai/.

Mr Girdhar says that he developed an idea in 2010 for an app builder which allowed customers to build their own apps at low cost without coding knowledge, by dragging and dropping pre-built elements into templates to create personalised apps, while he was working in a previous business that he had founded, ONS Interactive Solutions PVT Ltd, which provided search engine optimisation, web development, web app services and Google AdWords management. He developed that idea with his father, and by around early 2013 had come up with a minimum viable product version of Appy Pie’s app builder. He said that from 2013 to early 2015, he and his father set up a company in India called Appy Pie Inc, and operated in partnership. He was involved with programming but otherwise continued to run ONS, and his father was responsible for day-to-day operations of Appy Pie Inc, including marketing and finances. However, Mr Girdhar’s father sadly died in June 2015, so Mr Girdhar took over the running of the business and set up D2 in January 2016 with his mother, who remains a joint partner in it, although playing no active role. It is D2, he says, who is now the operating entity of the Appy Pie brand; has around 200 employees, mostly in India; controls product development, advertising, the Appy Pie Website and all other activities except for payment processing in certain territories, including the UK.

Mr Girdhar’s evidence is that D2’s business is a very successful one, and ranked as an industry leader by a number of publications, including in December 2023 by G2 who rated it as the leading drag-and-drop app builder. His evidence is that the Appy Pie platform has over 10 million registered users and over 15 million apps have been created using it, by users based all over the world, who he referred to in cross-examination as “Citizen developers”. These may be corporate entities or other organisations, or individuals. In cross-examination he says he knows of users aged 6 and aged over 80 who use the Appy Pie platform to create apps. In the UK, he says, users include: the NHS which has used the D2’s app builder to create an app used by 1.2 million NHS staff; the Metropolitan Police, to create a counter terrorism app; and the University of Westminster (his alma mater) which uses it to teach app development to students.

Mr Girdhar says that D1 was set up by him and his wife in April 2016 to act as a pure payment gateway to D2, so that UK and European customers could pay for D2s products and services in pounds sterling and Euros. Although he was a director of D1 when it was incorporated, he resigned after less than a month on tax advice, and his wife is the sole owner and now the sole director of D1. He says D1 was set up because international bank and credit cards cannot be accepted by Indian banks.

In his witness statement Mr Girdhar says that D1 has no involvement in the development or marketing of D2’s products whatsoever, and this has at all relevant times been explained to users in the terms and conditions on the Appy Pie website. In answer to questions from Mr Howe, Mr Girdhar said that when a customer in the UK purchases from the Appy Pie Website, D1 issues a VAT invoice to that customer, which will show D1 as providing the goods or services. However, he said, that customer will have agreed with the terms of service on the Appy Pie Website, which provides that the customer is contracting with D2, and once they have paid, D2 will manage the relationship. He confirmed that after collecting payment and issuing the VAT invoice, D1 also accounts for VAT arising from D2’s trades to HMRC. However, subject to that, the rest of the revenues remain in D1’s bank account unless and until they are repatriated with approval from the Reserve Bank of India.

Mr Lomas asked Ms Gupta about a Gartner report of July 2021, entitled “Magic Quadrant for Multiexperience Development Platforms”. It says that its research focuses on 13 multi-experience development platforms (“MXDPs”) available “from vendors who are able to execute and have the vision to meet the demands of the market”.

Ms Gupta said that she had contributed to the production of that report by answering a number of questions posed of her by Gartner in March 2021, and said that the report was a roughly accurate description of the Claimant’s position in the first half of 2021. The Gartner report describes the Claimant as “…a Visionary… its MXDP product builds on common components seen in applications and uses AI and machines to assemble components and quickly build new applications… Builder.ai’s operations are mainly focussed on APAC and the US with expansion to Europe and the Middle East”. Ms Gupta also accepted as accurate Gartner’s description of the Claimant: “Overall viability. Builder.ai is a new entrant to the market and is relatively small”, but said that had to be considered in its context, which is that it was being compared to companies such as Salesforce which were multibillion dollar corporations.

Ms Gupta agreed in cross-examination that the Claimant’s presence in the UK when she arrived was smaller than it is now, saying that she had doubled the Claimant’s UK marketing efforts on her arrival in London, and had also increased the London marketing team from 22 people in 2022 to 27 people today. She said that the London marketing team worked on global sales of the Claimant’s products, not just UK sales.

In terms of quantifying those sales, Ms Gupta stated at paragraph 10 of her witness statement that the Claimant’s UK revenues between 2019 and 2022 were £6,208,000 in 2019/20, £10,723,000 in 2020/21, and £10,557,000 in 2021/22, with turnover split between Builder Studio and Builder Cloud as follows:

Ms Gupta does not state in her witness statement where those figures came from, but in cross-examination she said they were provided to her by the Claimant’s finance department. Whatever she received has not been disclosed and is not in the trial bundle. The more concerning issue is that these figures contradict those pleaded by the Claimant at paragraph 8A.1 of the APOC. There, the Claimant pleads that “since approximately March 2019 to the present [dated 3 May 2023], the Claimant had total UK sales revenues of approximately US $2.18 million.”. Again, the source of those figures has not been stated. It was put to Ms Gupta by Mr Lomas that this amounted to about half a million dollars a year, and that the revenues claimed in her witness statement had therefore gone up by a factor of about 20 to those contained in the APOC. Ms Gupta accepted that discrepancy, and said that she would give preference to the figures in her witness statement as being accurate, but accepted that the Court had no way of verifying them. I do not know what I can do with these unsupported and contradictory sets of figures and the answer is, I fear, that I am not able to rely on either of them as there is no documentary evidence to support either of them and Ms Gupta cannot assist me in assessing which are more likely to be correct on the balance of probabilities, as she was not involved in the production of either of them.

Of course without UK sales figures, it is not possible to assess the Claimant’s UK market share. Ms Gupta relies on a Gartner Report of July 2023 assessing the “Application Architecture, Development and Platforms” market, which quantified the market globally as being worth $76 billion in 2021 and $87 billion in 2022. She accepted in cross-examination that in the context of the global market identified by Gartner, the Claimant’s business was small, and further accepted that she had provided no information about the size of the UK market in order to assess the size of the Claimant’s UK market share. Mr Lomas put to her a document from the International Monetary Fund showing that the UK’s Gross Domestic Product based on purchasing power was c. 2.17% of the world GDP, which she accepted was a good guide to UK economic activity within the global context. That would put the UK market size at about $1.888 billion, and means that even if the much higher sales figures in Ms Gupta’s witness statement are correct, the Claimant’s UK market share (at least up to 2022) is quite insignificant, in my judgment, amounting to not much more than the rounding error if $1.888 billion were rounded up to $1.9 billion.

Nonetheless, Ms Gupta’s evidence is that the Claimant is one of the top 4 brands in the UK market for app development, the other main competitors in the low-cost app development space (the Builder Studio Store market) being Shopify, Fiverr and the Defendants and the main competitors for Builder Studio being Bubble, Appian, Shopify and Webflow. She says that the Claimant’s market is “anyone in need for an app”, which is mostly businesses but may be other organisations or individuals, and that the Claimant’s brand are “very well known” in the market because of their advertising and marketing that they do in the UK.

Ms Gupta was asked about this advertising and marketing in cross-examination. Once again, we run into problematic figures. She was taken by Mr Lomas to Annex B to the Claimant’s APOC, which is a table setting out payments made to Google for AdWords spend split out by word and by jurisdiction worldwide for each of 2021 and 2022. That appears to show that the Claimant had no Google AdWords spend for Builder.ai in the UK in 2021, and spent just under £30,000 in 2022, which was causative of 8,800 clickthroughs and 244 conversions. Ms Gupta agreed that was a small level of spend, but noted that in her witness statement she sought to disassociate the Claimant from those figures, saying that the Annex B APOC table was “obtained from the Claimant’s accounts department and had been misunderstood”. No further information was provided in her witness statement (or, indeed, in oral evidence) about how it had been misunderstood or why it should not be relied on.

Instead, Ms Gupta in her witness statement espouses a different Annex B, this time Annexed to the Reply and Defence to Counterclaim. This is a screenshot from a GoogleAds summary page. At the top, there is a reference to Builder.ai together with a contact name and their work email address at the Claimant, so it appears to be the Claimant’s account. It purports to show a spend of $391,000 on Google Ads from January 2022 – June 2022, but it is impossible to discern how that has been spent. The information about clickthroughs and conversions has been redacted so there are no success metrics before the Court either. The data seems to have been filtered using the somewhat ambiguous criterion “Campaign name contains uk” but Ms Gupta accepted that it does not tell us what AdWords it has been spent on, in what countries, and nor could she explain the discrepancy between the two sets of figures in the two Annex Bs. Once again, I regret I cannot rely on this information at all, when one document produced by the Claimant suggests the UK Google AdWords spend for Builder.ai is merely £30,000 over a full year, and the other puts AdWords ‘campaign with the name including uk’ at almost $400,000 over six months. The two are irreconcilable, and are not even directly comparable, in my view. I do not consider I can rely on either when Ms Gupta’s evidence is that the former is unreliable (or “misunderstood”) and I simply do not have enough information about the latter to know what it relates to, and no background documentation against which to verify it. Further, it seems unfeasibly high in my judgment, particularly when viewed against the table Ms Gupta displays at paragraph 19 of her witness statement.

The paragraph 19 table was also said by Ms Gupta to have been produced by the Claimant’s account department, giving “details of all advertising spend for the Claimant aimed at the UK market from 2019 to 2023… broken down by type of marketing… [and] converted into pounds [sterling]”. In fact, the table does not show this. It only covers the period from 2021 to 2023. It purports to show the marketing spend rising from £683,081 in 2021 to £1,437,433 in 2022 to £1,506,733 in 2023. That comes to £3,627,247. That is, again, problematic, as it does not correlate with the figure for marketing spend pleaded in the APOC, which provides at paragraph 8A.2 that since approximately March 2019 to the present, the Claimant has spent approximately $3.18 million on marketing and advertising in the UK. This is significantly less than the total of £3,627,247 provided in Ms Gupta’s witness statement, particularly when the USD/GBP exchange rate is taken into account. Again, Ms Gupta accepted these figures were inconsistent in cross-examination, she said that she thought the figures in the APOC were probably not correct and those in her witness statement were, but could provide no real justification for that contention, and she accepted that no underlying information had been disclosed in relation to either set of figures to assist the Court in reaching a determination about which, if either, was right.

Even if I were simply to accept the advertising spend figures in Ms Gupta’s witness statement, those totals are broken down into just 6 categories (one of which is merely “Events”, another “Display” which Ms Gupta clarified was display marketing on third party websites and advertorial pieces on online publications). These are not further particularised and so they do not provide any useful information about exactly what pieces of marketing were carried out, at what cost or how effective any of it was in terms of impressions, conversions, leads, click-throughs, etc. All that is provided is a sub-total for each such category. I do not know into which category Google AdWords spend falls, but if it is in “Brand”, which seems the only real possibility, that category spend is £581,291 for 2022, of which the majority would have gone on Google AdWords in the first six months of the year, if the Annex B attached to the Reply and Defence to Counterclaim is right. I do not think it can be, as that suggests there was little other spend on “Brand” that year save for Google AdWords, and that there was little spend in the second half of the year at all. Both seem unlikely. Other bare figures include that in 2022 £271,244 was paid to Squirrel Media for online display and influencer advertising (but nothing was spent in this category in 2021), and in 2021 £190,163 was spent on ‘Events’.

In relation to the latter, Ms Gupta goes on in her witness evidence to say that the Claimant spent $76,000 at London Tech Week in 2021, but there is nothing at all to tell me where the rest of the £190,163 was spent that year nor, for example, how many people attended London Tech Week, or any success metrics arising from that event. Ms Gupta accepted in cross-examination that none of that sort of detail had been provided. She also accepted in cross-examination that many of the events from 2021 and 2022 were outside the UK, because, she said, the UK did not host many big tech community events in this time period (and even in 2023 and 2024) so the Claimant was forced to go abroad. There is no evidence before me to enable me to assess if the monies spent on events abroad are accounted for in whole or in part as UK advertising spend in either of the Claimant’s UK advertising spend figures, and if so, whether this is justifiable given the level of attendance at such events by UK consumers, for example.

In her witness evidence Ms Gupta refers to using LinkedIn posts for ‘free advertising’. She says that the main LinkedIn page is called Builder.ai and has over 240,000 followers, with posts going up about 4 times a week, mostly with the Claimant’s trade marks in them. A number of those have been exhibited to her witness statement and I have looked at them. Ms Gupta clarified in cross-examination that the 240,000 figure is the number of followers for the Claimant’s LinkedIn page at the time she wrote her witness statement, i.e. in 2024, but that had grown over time from “a very, very small number” in 2021. It has not provided any evidence of the number of LinkedIn followers in 2021, 2022 and 2023.

Ms Gupta accepted in cross-examination that the LinkedIn posts the Claimant relies on do not appear to have many interactions with them, in terms of comments or reposts, but said that she would not expect engagement on posts that were merely informative, for example those that simply inform the public that the Claimant will be at a certain tech event. She said that she considered it “lucky” if a LinkedIn post got more than 100 or 200 interactions, but many had considerably less. Ms Gupta accepted in cross-examination that the Claimant had not provided any disclosure of the geographical location of its 240,000 LinkedIn followers, although it could have provided data breaking that down, as Mr Girdhar had done in respect of the Defendants’ LinkedIn Post complained of.

Similarly, in relation to the blog posts relied on from the Claimant’s website, the Claimant has not provided any figures at all for how many of those who interact with or read those blog posts are from the UK. Ms Gupta accepted in cross-examination that many of those posts are reporting on tech events abroad and it seems to me that there may be little interest in the UK in reports of events in Helsinki or Dubai. I have no metrics in relation to those events, as previously mentioned.

The Claimant points to a marketing spend of almost £25,000 on perimeter branding or stadium edge advertising, at five Championship and 11 Super League rugby league games in October 2021, during the pandemic when attendance figures were limited to a quarter of the usual capacity. I have seen photographs which show the Builder.ai Figurative Mark was used around the pitch perimeters, and I accept Mr Howe’s submission that it was there in part to be picked up by the television broadcasts. The Claimant has not provided any evidence of attendance numbers at these closed games, although the Defendants have produced a print out from Wikipedia which suggests attendance of around 4,000 for one of the Super League games, which Ms Gupta was willing to accept. Ms Gupta confirmed in cross-examination that she knew they were all going to be closed games, but said that they were all televised on Sky Sports. Although I am willing to accept that is her honest recollection, this was not in her witness statement and the Claimant has provided no documentary evidence to verify that recollection. Nor has it provided any evidence about what viewing figures any such televised games achieved.

Ms Gupta’s evidence is that over six days in June 2022 the Claimant spent US$141,000 on outdoor billboards and posters in key business hubs, near WeWorks’ offices and at iconic sites, plus associated social media, for a Platinum Jubilee campaign over 6 days in June 2022. She says that the Claimant’s advertising agency provided figures suggesting that the outside advertising was seen by approximately 1.8 million people who each saw it on average about three times, being just under 6 million impressions. That underlying information from the advertising agency has not been provided by the Claimant. The social media campaign was across Facebook, Instagram and LinkedIn “targeting London entrepreneurs, start-ups, early tech-adopters and investors”. Ms Gupta said that she believes it reached about 1.79 million people who each viewed it approximately 4 times. The 552 social media posts of the campaign were engaged with, for example by ‘likes’ or reposts, 552 times, and resulted in about 6.600 clickthroughs to the Claimant’s website. She says that the Claimant viewed it as a very successful marketing campaign developing the reputation of the Claimant’s marks and brands. Again, no underlying information to support this data is provided.

I note here that the Claimant has provided two other examples of other marketing campaigns, but these date from November and December 2023 and so after the relevant date at which acquired distinctiveness falls to be assessed.

Ms Gupta’s written evidence is that as a result of the use and marketing of the Marks, the Claimant has seen considerable growth in media coverage of it. She annexes an Excel spreadsheet of UK media coverage in 2022 and 2023. Looking at the entries for 2022, the Claimant’s table shows that it had 35 pieces of UK media coverage, of which 11 were articles in the UK national press. The total ‘views’ of that coverage is said to be 1.7 billion, and the advertising value equivalent (“AVE”) of that coverage is put at £42.1 million. In 2023 this increased to 74 pieces of coverage (of which 36 were in the UK national press) said to have 6.5 billion views and an AVE of £161.5 million, but all of these 2023 entries, again, are after the relevant date at which acquired distinctiveness falls to be assessed for the purposes of the infringement claims, as Mr Howe for the Claimant accepted in closing submissions.

In cross-examination Ms Gupta said that this spreadsheet was not produced for this litigation, but is a maintained list. She said the figures it contains for ‘views’ were obtained from the publications themselves as the online and print ‘reach’ of their publication, and the AVE would have been produced from a mix of the publications themselves and from the Claimant’s PR and communications department. She accepted that no underlying information had been provided to enable those calculations to be checked or interrogated.

Ms Gupta says that as a result of the Claimant’s marketing efforts “the Claimant’s Builder family of brands is one of the best known app software development brands in the UK”. She says that she interacts with customers at events, where hundreds of people visit the stands. She says most of them have some idea who the Claimant is, and generally refer to the Claimant as Builder, as do the Claimant’s staff, or as Builder.ai initially and Builder later. In her written evidence she says that consumers use the word ‘Builder’ and ‘Builder.ai’ to search for the Claimant using search engines.

On that latter point, Ms Gupta has provided some evidence from Google AdWords statistics showing the number of times that Builder.ai was the search term, and the number of click throughs to the Claimant’s website from such searches. That is quite interesting, given Ms Gupta’s evidence of the limited marketing which took place in 2019, and her decision to double the 2021 marketing spend when she arrived in London in 2022, because it shows surprisingly little change in brand search volumes over the March 2019 – May 2023 time period it covers. For the 9 months of 2019 that it covers, it shows 2,821 searches and 61% of those clicking through to the Claimant’s website (being 1,733 ads). If that is uplifted proportionately for a full year, that would be 3,761 searches. 2020 was a depressed year with a significant dip in brand search volumes, which perhaps could be explained by the pandemic. However, 2021 and 2022 show an almost identical volume of searches for Builder.ai as in 2019, at 3,730 and 3,800 searches respectively, and the conversion rate is significantly lower than in 2019, at 49% and 47% respectively. In 2023 there is an increase in the volume of brand searches at 2,821 over 5 months, which uplifted proportionately over a full year would be 6,770 searches, but the conversion rate continues its descent, to 36%.

Ms Gupta also exhibits a Google Trends screenshot relied on by the Claimant, which tracks as a graph the search terms ‘Builder.ai’ (in blue) and ‘Builder ai’ (without the full stop separating the two elements) (in red) in the UK from 1 January 2018 to 31 December 2023. This has been annotated with a comment, that “We’ve observed an uptrend in searches for the ‘Builder ai’ [I note, no full stop] brand beginning in November 2022. Since that time the visibility and recognition of our brand have been on a consistent upward trajectory, closely aligned with the growth and development of the brand itself”. Ms Gupta confirmed that was not her comment, but that of someone from the Claimant’s digital team. That screenshot shows both the blue and red lines bumping along the X axis, barely troubling the Y axis which is labelled “interest over time” from 0 – 100, until the end of 2022 when the red line (being ‘Builder ai’ with no full stop) suddenly takes off and hits peak popularity in the second half of 2023. The blue line does not behave in the same way at all, showing only a very slight increase over the same time period. Ms Gupta was not really willing to accept this suggestion when it was put to her in cross-examination, but nonetheless it seems to me more likely than not that the dramatic increase in searches was not because of a sudden interest in ‘Builder ai’ (as then it would likely to be replicated at least in some form in the ‘Builder.ai’ searches, and is not) but in AI generally. As Ms Gupta did accept in cross-examination, Chat GPT was launched in November 2022, which is about when that red line takes off in the screenshot, causing much excitement about AI with the general public since that time. Mr Girdhar also made that point in his cross-examination, saying “Everything is becoming AI”. For that reason I put the red line to one side and focus on the blue.

Ms Gupta accepted that the blue line in that screenshot showed small numbers of searches for ‘Builder.ai’, but said that others would also be searching for “Builder Studio” and “Builder Now” etc. She accepted that the Claimant could have provided data for those search terms, but had not done so. It has provided brand search volumes for “Builder”, which show that of those who had searched for the term “builder” between 1 August 2021 and 31 May 2023 in the UK, only 15 had clicked on the Claimant’s “builder.ai” ads as a result of the search. That is about 2 searchers every 3 months. Ms Gupta agreed that was very low, saying “Yes, that is not the best number”.

On balance, therefore, I find that the quantitative data before me does not support Ms Gupta’s written evidence that consumers use the word “Builder” to search for the Claimant using search engines, and it provides only very limited support to the contention that ‘Builder.ai’ has strengthened as a brand in the UK from 2019 to early 2023.

The Claimant’s main complaint in relation to Category 1 infringement is the Defendants’ use of “Builder” as a part of names for their products and services as set out in paragraph 11.2 of the APOC. Ms Gupta says “Consumers recognise references to “Builder” as references to the Claimant, so these uses are likely to suggest to consumers that these products come from or are associated with the Claimant contrary to the true position. The other references to “Builder” throughout the Defendants’ website reinforce this message. More generally, in calling to mind the Claimant, these references enable the Defendants to take the benefit of the positive brand image associated with the Claimant’s trade marks”. This is a very lawyerly passage. When this was put to her in cross-examination, Ms Gupta confirmed she has had no legal training, but said those were her words. She accepted that these comments were just an expression of her opinion. They are, in my judgment, trespassing into the realms of advocacy which is not the function of a witness of fact.

Ms Gupta further stated in her written evidence that “There is no reason for the Defendants to be using the word “Builder”. It is not a recognised term in the app development market… the sole reason for the Defendants using “Builder” in the way they do is to try to confuse consumers into thinking their products are the products of or connected to the Claimant.” Ms Gupta resiled from all of these statements in cross-examination as I will come on to explain.

Mr Girdhar’s evidence is that when he started D2 and launched his app builder and website builder products, he did not associate the term “builder” with any particular company or brand. He chose to use the term “app builder” in marketing from the very start of the business to describe the products that D2 offered and their core function, including on the Appy Pie Website since inception, and has supported this with a screenshot showing use of the term “app builder” on D2’s website in 2013. He says the term “builder” has been used descriptively in the software industry for as long as products that enable users to build websites and apps have existed. He says it is a descriptive term and used by consumers and the industry as a whole with other descriptive terms to describe the functionality of such software tools e.g. “website builder”, “app builder” or with a higher degree of specificity, “retail app builder”. He thought these sorts of terms, using “builder” or “Builder”, had been used since about 2010 but relies on screenshots of use of “website builder” “app builder” and “chatbot builder” in articles and press releases from various third party sources dating between 2013 and 2020. He says those reflect his recollection of how those terms were used by third parties throughout the industry to describe such software during that period. Mr Howe suggested to Mr Girdhar that the word “builder” without a descriptor in front did not have an accepted meaning in the software context, but he denied it. He said, “So Microsoft has just launched AI Builder. So if you Google it, you know, AI builder, the entire first page is filled with Microsoft pages. Google just launched Gen AI app builder. Now everything is becoming AI: AI app builder, Gen AI, generated AI… I see app builder everywhere in this category… the broader category is no-code app builder or drag and drop app builder”.

His evidence is, in my view, supported by documents from the Claimant’s own website. Those include a blogpost entitled “A little bit about AI (and more, straight from Builder’s mouth!” of 3 August 2019. Some extracts (with my emphasis) include:

Another blog on the Claimant’s website answers the question “What are online app builders?” with the following definition: “Online app builders and online platforms let you build your own app from scratch without coding. They more often than not utilise a drag and drop interface. They offer a range of stock templates and features to suit different types of apps, including restaurant apps.” (my emphases).

The Claimant uses the term “builder” in a generic and descriptive sense (referring to, for example, “app builders”) in multiple places on its own website, in examples contained in the trial bundle, as does Ms Gupta in both her witness statement and in her oral evidence. Ms Gupta accepted in cross-examination that by 2018 at least, “app builder” was “absolutely” such an accepted and generic term. I find that ‘builder’ was a descriptive term used widely in the software industry as described by Mr Girdhar, i.e. as a tool to enable the creation or development of software, whether used with a further descriptor or not, since at least 2013. There remains a dispute about the significance of capitalisation of the word “Builder” which I come onto now.

Mr Girdhar said neither he nor D2 has attached any significance to whether the initial “b” at the start of the word “builder” is capitalised or not, but that they generally followed the norm in India of capitalising words in headings on their website, and also usually capitalised product names when they appeared in bodies of text on webpages.

In respect of the Defendants’ use complained of, Ms Gupta in cross-examination:

I am satisfied from the evidence I have seen that the word “Builder” with an upper-case ‘B’ was equally also a descriptive term used widely in the software industry as described by Mr Girdhar, i.e. as a tool to enable the creation or development of software, whether used with a further descriptor or not, since at least 2013 and interchangeably with use of the word “builder” with a lower-case ‘b’.

It is convenient to note here that Mr Girdhar says that both “Cloud” and “Pro” are two other very common terms in the software industry which are used descriptively: “Cloud” since the early 2010s when cloud hosting by Amazon Web Services had become a very popular alternative to dedicated server hosting, and so was used to describe a product which involves virtual rather than local hosting of data; and “Pro”, which was commonly used in the technology industry to describe “professional” or superior hardware or software to appeal to businesses or customers requiring greater functionality. Again, he has provided support in the form of articles and website printouts showing such use, but his evidence on this point was not challenged and I accept it.

There is no dispute that the LinkedIn Post uses the sign “Builder.ai” in a form identical to both the Builder.AI Figurative Mark and the Builder.AI Word Mark, to refer to the Claimant’s goods or services, as part of a list entitled “7 Best App Builders”. The LinkedIn Post lists the D2 first and the Claimant third. It deals with each in a few seconds, showing the brand, and a screenshot from the relevant company’s website showing pricing. It provides no detail of specific features or characteristics of the differing products. At the end it says, “Which do you prefer”?

Ms Gupta’s written evidence is that all posting on LinkedIn, especially by companies, is really advertising or marketing, and said, “It is clear that the Defendants’ LinkedIn [P]ost was promoting their product and trying to say that it was better than other similar products, including the Claimant’s”. She said “The Defendants would have wanted to cash in on the good brand name of the Claimant. They would also want to try to show they were cheaper than the Claimant, but not to show the differences in the product and would not have shown the amount of effort their customers would have to put in to produce their app as they have to develop it themselves using the Defendants’ tools. The Claimant develops the apps for their customers unlike the Defendants who solely provide tools to the customer for that customer to develop its own app itself.”

She describes the LinkedIn Post as “materially inaccurate” because it said the price quoted for the Claimant’s product featured, Builder Studio Store, is “from £295” a month when there were lower prices available, including a starter package at £150 per month for 24 months. She complains that the LinkedIn Post does not compare like with like, because Builder Studio Store has different functionality, features and support than the Defendants’ product, and those differences serve to explain the price difference highlighted in the LinkedIn Post.

In cross-examination, however, Ms Gupta accepted that the LinkedIn Post was just a list of seven products in the no-code application development space, rather than any sort of qualitative review; that the pricing screenshot on the Builder.ai entry came from the Claimant’s own website and so was not inaccurate, albeit she maintained her complaint that lower prices could have been showed; and that the post was not encouraging anybody to buy a particular product at all, but rather presenting different products available in the market place each with a price from the providers website, and asking “Which one do you like?”. She further accepted that the Claimant had posted a similar list on a blog on its own website, saying “With so many app builders out there, it can be difficult for beginners to know which platform to go for. Below, we explore some of the most widely used app builder platforms, giving you all the information you need to choose a builder and make an app, even if you’re a complete novice”. In that list, the Claimant is listed first, and D2 second.

Mr Girdhar does not himself manage the marketing function at D2, including the posts it makes on D2’s LinkedIn page or its GoogleAds spend. However he says that he has investigated the circumstances of the creation of the LinkedIn Post complained of, and has spoken to D2’s Head of Marketing who was aware of the circumstances of its creation. He said that it was one of several similar posts published by D2 on its LinkedIn page for marketing purposes. In cross-examination he said that the marketing purpose was not particularly to drive sales of D2’s app builder product, but had the wider purpose of seeking to promote the growth of the whole “no-code ecosystem”. He said that it was part of the marketing work that D2 does to grow that industry as a whole, so it did not focus just on Appy Pie but on multiple no-code development platforms or no-code app builders. He said “We want no-code to be the biggest industry and the way it is growing, no-code is the future. We want everyone to be aware of the technology, so people do not have to write code, basically”. Given that the LinkedIn Post does not encourage the purchase of one product over the other, as Ms Gupta accepts, I accept his evidence that its primary purpose was not to drive sales but to drive interest in the no-code app builder market.

Mr Girdhar’s written evidence is that “In each case, junior members of [D2’s] marketing team conduct Google searches for specific relevant keywords and compile a list of the top natural search results for those keywords. We then publish a post that lists those results. There is no intention to rank the products shown in the posts in any particular order. The products or companies listed in these posts will all relate to the search term, however we do not limit the posts to direct competitors: we simply include products responsive to the keywords.”. In oral evidence he clarified that D2 did not change or reorder the positions in which the Google search returned the results, so the order that the different products were presented in the LinkedIn Post video is the order they were returned on the search. I accept his evidence. He said his team did, however, go onto the websites of each of the products returned by the search to find pricing, and took a screenshot of what they found and added it to the video. Mr Howe put it to him that there were other pricing options on the Claimant’s website which were rather cheaper than the price chosen by his team, and he said that he had been told that his colleagues’ intent originally was not to show any pricing. He said, “Whatever screenshot we got, we just took the screenshot and pasted it.”. He said it can be seen from the web archive that “this is exactly what we were seeing on the [Claimant’s] website at the time.”.

Mr Howe put it to him that even if it was not D2’s intention to do a price comparison, that is how it would appear to people who watched the LinkedIn Post video. Mr Girdhar did not accept that. He said that there was no attempt at any comparison between the products, and it was never meant to be perceived as a comparison, because it did not, in fact, compare the products. It was a “listicle” of products available in the no-code application development he said, similar to those that that appear on the Claimant’s own blog.

In cross-examination, Mr Girdhar said that the LinkedIn Post was directed at the Citizen developer no-code community across the world, as D2 has customers across the world, but said that these posts didn’t get much traction on LinkedIn where people were mostly looking for jobs, so it was not a good marketing platform. His evidence on these points is supported, in my judgment, by the significant amount of analytic data provided by Mr Girdhar in relation to D2’s LinkedIn page and the LinkedIn Post. That supports evidence in his witness statement that typically less than 10% of D2’s LinkedIn traffic is from the UK (the evidence being D2’s LinkedIn profile analytics for the year 2023, which he said has been consistent for as long as he can recall); and to support the analysis he provides in para 26 of his witness statement:

I find that the LinkedIn Post makes no attempt to draw any comparisons between the products (save that they are all no-code development platforms or app builders) because it was a post created for the purposes of education and not marketing, as Mr Girdhar says. It informs viewers of options in the market, and provides indicative pricing which ranges from $8 to $340, but does not attempt to put forward any of them as better or preferred options, compare the features and functionality of the products, review them or address the different pricing structures that each provider may have in place. It asks viewers merely “Which do you prefer?”, but in order to answer that question, they will have to carry out their own research and assess the products against their own needs and finances. I also accept Mr Girdhar’s evidence that the LinkedIn Post was not targeted at any particular jurisdiction, in the lay sense, but made available to anyone around the world who wished to access it on D2’s LinkedIn page. The question remains whether it was targeted at the UK in the legal sense for the purposes of the Category 2 alleged infringement claim, and I will address that in Issue 8 below.

The Claimant has set out in the APOC and Reply and Defence to Counterclaim what it says are uses by the Defendants of the signs which are similar or identical to the Marks, by reference to screenshots from the Appy Pie Website. Ms Gupta’s written evidence is that this “shows them using Builder.ai although they have removed the full stop between “Builder” and “ai”” in February and November 2022. She was taken to those examples in cross-examination and accepted that these did not show that a full stop had been removed from “Builder.ai”, but they were simply instances where the word ‘builder’ was next to ‘AI’, such as in reference to one of D2’s products: “Based on our response, our app builder AI…”.

Similarly, Ms Gupta in her witness statement says that the Defendants have also been using the Builder Now Mark, but she accepted in cross-examination that the examples she relies on do not show the words “Builder Now” in the trade mark sense but as sentence fragments where the Defendants are exhorting people to use their own products, such as “Try Android Builder Now” and “Try DIY on-demand delivery App Builder Now”.

Issue 1, 15 and 16 really fall to be considered together. Before considering those, I must first determine who is the average consumer, as the question of inherent distinctiveness is considered through his eyes. This does not appear to be disputed. The parties agree that the average consumer is anyone who wishes to create an app without coding, whether businesses, other organisations or individuals. Those may be organisations of any size and individuals of any age. The Defendants submit that there are no special features of the average consumer to weigh in this case, save that the level of attention would be higher than average given that the consumer is selecting a product that will enable them to build something of value. I accept that submission.

The Claimant’s case is that all the Marks are inherently distinctive. Mr Howe in closing submits that if one were to rank them on an ascending scale of inherent distinctiveness, then: the Builder Word Mark would be at the low end of the scale, but he submits just enough to be inherently distinctive; the Builder Home Mark is more capable of being distinctive, particularly because of the blocked in initial letter ‘B’, which is not just in a particular font, but catches the eye; Builder.ai Word Mark is next, with additional capacity to distinguish by joining two words which is not, as such, a descriptive phrase that is previously recognised, and then the Builder.ai Figurative Mark because the stylisation further adds to the capacity to distinguish.

The Claimant further submits that:

The Defendants submit that the Marks are inherently non-distinctive with respect to material parts of their specifications. I will need to deal with these by mark or pair of marks, and I will determine each as I go, always bearing in mind the Claimant’s submissions already recorded.

In relation to the Builder Word Mark and Builder Home Mark, the Defendants submit the word “Builder” is prima facie descriptive of at least the following emphasised parts of their (identical) specifications because:

I have found that ‘Builder’ and ‘builder’ are descriptive terms used widely and interchangeably in the software industry as a tool to enable the creation or development of software, whether used with a further descriptor or not, since at least 2013. I am satisfied that the average consumer, who is deemed to be well informed reasonably observant and circumspect, would have known that, and would immediately perceive the Builder Word Mark as descriptive, and as such inherently unable to identify the goods and services for which the Builder Word Mark is registered (as the Defendants limit their invalidation case above) as originating from the Claimant and so distinguish them from goods or services of other undertakings, as at the registration date.

In relation to the Builder Home Mark, I do not consider that the average consumer would perceive that as memorable and striking merely because of the stylisation of the capital ‘B’ as the Claimant submits. Although I have found that the average consumer pays a higher degree of attention because he is looking to create something of value, I consider the level of stylisation in the blocked-in ‘B’ and conventional font to be weak and not memorable even to the reasonably observant average consumer. The wide-ranging and long use of ‘Builder’ as a descriptive term before the Claimant registered the Builder Home Mark limits the capacity of such weak and unmemorable stylisation inherently to distinguish the Builder Home Mark from the Builder Word Mark, in my judgment, as I consider that the average consumer would still immediately perceive it as descriptive. Accordingly I consider the Builder Home Mark is also inherently unable of acting as a mark of origin to identify the goods or services for which it is registered (as the Defendants limit their case above) as being those of the Claimant and distinct from goods or services of other undertakings, as at the registration date.

In terms of the parts of the specifications that Mr Lomas has emphasised, Mr Howe in closing submissions indicated that even if I was not with him on the question of inherent distinctiveness, he objected to the second marked passage being excised from the Class 9 specification for each of the Builder House Mark and Builder Word Mark, because he submits that there is no evidence that they are descriptive of “downloadable software for content management, content creation, content aggregation…and content searching”. Similarly, he objected to the excision of the last passage from Class 42 to the extent it relates to content management, creation, publishing etc. I am satisfied that depending on the content and functionality of the apps created, development of apps could fall within this category and that it is appropriately emphasised. I would also draw the Defendants’ attention to “software design enabled by artificial intelligence” which has not been emphasised, but I think should be.

In relation to the Builder.ai Word Mark and Builder.ai Figurative Mark, the Defendants submit that they are prima facie descriptive of at least the following emphasised parts of their (identical) specifications, because:

I have found that the average consumer would immediately perceive the Builder Word Mark and Builder Home Mark as descriptive, and as such lacking in distinctive character. The only difference between those marks and the Builder.ai Word Mark and Builder.ai Home Mark respectively, is the ‘.ai’ element. In my judgment, the average consumer in the context of the software industry at the time of the registration of the marks, in December 2021, is likely to be aware first, that ‘ai’ was short for artificial intelligence, and second, that it was used as an adjective for anything autonomous, as the Claimant itself explained on its own website several years earlier.

Mr Howe submits that these marks are distinguished by joining two words in what is not a descriptive phrase previously recognised. However, they are joined by a ‘dot’, and in my judgment, the average consumer at the relevant time would perceive this as a top-level domain, not a phrase. I am satisfied that at least a class of the well informed and reasonably observant and circumspect average consumer (being adults, not children) would be familiar with the use of allusive domains, such as ‘.tv’, which although a country code top-level domain for Tuvalu, has been widely used in the television industry since the mid-1990s and has since been categorised as generic. Accordingly, I accept the Defendants’ submissions that the average consumer is likely either to ignore the ‘.ai’ element, particularly where it is used in a URL where his focus will be only on the descriptive word ‘builder’, or if he does not ignore it, perceive it as a top-level domain, but one which is allusive of artificial intelligence and not as the country code top-level domain for Anguilla (as in fact it is). In the latter case I am satisfied that he will easily recognise it as being descriptive of a builder, i.e. a tool to enable the creation or development of software, which utilises artificial intelligence to introduce an autonomous element to the build. As such, I am satisfied that the Builder.ai Word Mark and Builder.ai Figurative Mark are both inherently unable to identify the goods and services for which they are registered (as the Defendants limit their invalidation case above) as originating from the Claimant and so distinguish them from goods or services of other undertakings, as at the registration date.

In relation to the Builder Studio Pro Mark, which has an identical specification to the Builder Word Mark and Builder Home Mark, the Defendants submit that the Studio Pro element adds little to the average consumer’s overall impression of the registration because: (i) STUDIO is allusive of the platform or means of building; and (ii) PRO is a laudatory term akin to PLUS, EXTRA or similar, as a contraction of PROFESSIONAL, and indicates that the tools of the studio or the results of the build are of a superior quality, as was the tenor of Mr Girdhar’s evidence. The Defendants submit that BUILDER STUDIO PRO is therefore descriptive of the same parts of the specification as highlighted in respect of the Builder Word Mark and the Builder Home Mark for the same reasons, i.e. that it does nothing more than designate the kind, intended purpose and broad characteristics of the goods and services and is therefore descriptive and non-distinctive.

In my judgment, the average consumer who is well informed, reasonably observant and circumspect, and aware of popular suffixes in the naming of software products, will perceive the Builder Studio Pro Mark as having no material additional distinctive character to the Builder Mark. These are generic, descriptive words which I am satisfied would be perceived by the average consumer as lacking distinctive character alone or together. STUDIO adds nothing to BUILDER save that it might suggest to the average consumer that there are a number of different tools in the variant, and PRO is generically used to indicate, as was Mr Girdhar’s evidence, that it is a variant which will appeal to business or professional users or those who need some additional functionality or features, which I am satisfied the average consumer will know, being well informed. I accept the Defendants’ submission that the average consumer would consider the Builder Studio Pro Mark to do nothing more than designate the kind, intended purpose and broad characteristics of the goods and services, and is therefore descriptive and non-distinctive. I find that it is inherently unable to identify the goods and services for which it is registered (as the Defendants limit their invalidation case) as originating from the Claimant and so distinguish it from goods or services of other undertakings, as at the registration date.

In relation to the Builder Now Mark, the Defendants submit that the “NOW” element adds nothing to the average consumer’s overall impression of the registration as it is simply a call to action or reference to the immediacy of BUILDER. It is also smaller text within a button motif and could just as easily be disregarded if not forming part of the sign. For those reasons, the Defendants submit that the Builder Now Mark is prima facie descriptive of at least the following emphasised parts of its specification:

I am with the Defendants. In my judgment, the average consumer will immediately perceive the Builder Now Mark as having no distinctive character, for the reasons the Defendants give. As such, I am satisfied that the Builder Now Mark is inherently unable to identify the goods and services for which it is registered (as the Defendants limit their invalidation case) as originating from the Claimant and so distinguish it from goods or services of other undertakings, as at the registration date.

In relation to the Builder Cloud Mark, the Defendants submit that the “CLOUD” element adds nothing to the average consumer's overall impression of the registration as it is a generic term in the marketplace referring to specific hosting arrangements. It is also in smaller (all caps) text, in a “button” motif and could just as easily be disregarded as not forming part of the sign;

In the circumstances, they submit, BUILDER CLOUD is prima facie descriptive of at least the following emphasised parts of its specification:

I have found that CLOUD is a descriptive term used widely in the software industry as described by Mr Girdhar, i.e. to describe a product which involves virtual rather than local hosting of data. I take judicial notice that it has been so used and understood widely amongst the general public, and by implication by the average consumer in this case, since at least 2012, when Apple launched iCloud.

I accept the Defendants’ submissions that the average consumer would consider that the ‘Cloud’ element of the Builder Cloud Mark adds no material distinctive character to the Builder Home Mark because of the descriptive, generic nature of it and because the button motif in which it set is easily disregarded, the ‘Builder’ element being the dominant element of the mark. As such, I am satisfied that the Builder Cloud Mark is inherently unable to identify the goods and services for which it is registered (as the Defendants limit their invalidation case) as originating from the Claimant and so distinguish it from goods or services of other undertakings, as at the registration date.

The Claimant submits that even if the Court is not with it on inherent distinctiveness, the Marks are not so descriptive that they are incapable of acquiring distinctiveness through use, and it submits that all of the Marks have acquired a distinctive character through substantial promotion and use by the Claimant.

Mr Howe asks me to assess the evidence of use and acquired distinctiveness in the context of these marks being presented to the market as a linked ‘family’ of marks with BUILDER at the core, reinforcing the message that BUILDER is being used as an indication of trade origin. However, I have found that BUILDER is descriptive and non-distinctive, and I accept the Defendants’ submission that a family of marks cannot be established for the purposes of supporting the existence of a likelihood of confusion, when the common component, ‘Builder’ in this case, is itself a non-distinctive, descriptive word. As the Court of First Instance in Miguel Torres v OHIM, cited at [234] of W3 Ltd v easyGroup and set out in my summary of the law, made clear at paragraph 80, “According to that case law [BAINBRIDGE and Il Ponte Finanziaria], there can be said to be a series or a ‘family’ of marks when either those earlier marks reproduce in full the same distinctive element with the addition of a graphic or word element differentiating them from one another, or where they are characterised by repetition of the same prefix or suffix taken from an original mark…” (my emphasis). BUILDER is neither a prefix nor suffix but is the entirety of the Builder Word Mark or the Builder Home Mark, so in order to found a family of trade marks, it must be a “distinctive element” to which a differentiating graphic or word element is added. The reason for this is found from a full reading of paragraph 80, which finishes: “…When there is ‘family’ or a series’ of trade marks, the likelihood of confusion results more specifically from the possibility that the consumer may be mistaken as to the provenance or origin of goods or services covered by the trade mark applied for and considers erroneously that that trade mark is part of that family or series of marks (Il Ponte Finanziaria, paragraph 63).” If the common component is not distinctive itself, by definition (per Arnold J in Kit Kat I at [40]) it cannot be effective to identify the origin of the goods or services to which it is applied, and the purpose of establishing the family of marks simply falls away.

The Claimant seeks to rely on the UK sales figures and UK advertising spend figures that Ms Gupta adopts in her witness statement, but I have found those cannot assist me because of the discrepancies that I have identified. Accordingly I have no evidence of sales at all under or by reference to the Marks, I am unable to calculate the market share held by goods bearing the Marks, and I have very little reliable evidence of the scale of investment by the Claimant in promoting the Marks. I do have some evidence of marketing spend, as I am satisfied with the figures that Ms Gupta has given me for the Rugby League sponsorship and the Platinum Jubilee campaign.

The Claimant submits that sums spent on marketing as a proportion of the overall market is an irrelevant criterion in establishing acquired distinctiveness, which I accept, as Arnold J at [43] of Kit Kat I makes clear it is the investment in marketing that matters, not how that investment compares to the size of the market. However the market share held by goods bearing a trade mark is relevant evidence to the Court’s assessment of whether that mark has acquired distinctive character.

Even if I was able to accept Ms Gupta’s sales figures, which I have found that I cannot, she only gives those in respect of BUILDER CLOUD and BUILDER STUDIO. As the Defendants submit, and I accept: the Claimant does not assert a trade mark for BUILDER STUDIO which is different to BUILDER STUDIO PRO; those figures provide no evidence at all of sales under or by reference to BUILDER or BUILDER.AI, whether as a word mark or in stylised form; and the lack of documentary evidence relating to those figures means the Court cannot assess what marks or signs those sales were made under. I have also found that on Ms Gupta’s sales figures, the UK sales achieved is an insignificant amount in the context of the overall UK market for no-code app development software. The Defendants submit that this is insufficient to demonstrate that BUILDER CLOUD has acquired a distinctive character through use, but of course if I did have reliable evidence of market share, which I do not, that is only one of the factors which would go into the assessment of acquired distinctiveness (per Arnold J at [43] of Kit Kat I).

Mr Lomas produced in advance of trial annexed to his skeleton argument a document entitled “Annex 3 – analysis of evidence”, which describes the documentary evidence of use contained in Ms Gupta’s first exhibit to her witness statement, against each of the Marks, which Mr Howe also referred to, and augmented, in closing. I have considered their submissions in relation to this, which were detailed.

The first thing to be said is that this document highlights how much of the documentary evidence relied on by the Claimant relates to use after the relevant date for acquired distinctiveness. In closing, Mr Howe accepted that I should only look at those that are dated up to 2 February 2023.

Secondly, it highlights that the Claimant has put forward no evidence of use of the Builder Studio Pro Mark, the Builder Now Mark or the Builder Cloud Mark.

Thirdly, it highlights that a number of social media and website posts of the Claimant relate to events which have taken place out of the UK, without any metrics to enable an assessment of whether they were seen within the UK, or any evidence that they were targeted at the UK. Accordingly they are of little assistance to me when seeking to assess how intensive, geographically widespread and longstanding the use of the remaining Marks has been in the UK.

In relation to the press coverage spreadsheet disclosed by the Claimant, the Defendants submit that the ‘reach’ and AVE figures it contains are unrealistically high. Mr Howe for the Claimant submits that these may be surprising, but are conventionally used in the advertising industry and amount to evidence of how the Marks were presented, which I accept. He submits there is always the problem in assessing the impact of advertising, of knowing what has actually been seen and read, by how many people, and where. I will come back to that.

Mr Howe submits for the Claimant:

Turning to my overall assessment of the relevant evidence, then, and looking at the categories of evidence set out at [43] of Kit Kat 1 so far as they are relevant:

My conclusion, after stepping back and taking all relevant matters into account, is that I find that none of the Marks have acquired distinctiveness. The evidence is simply lacking. The effect of my findings is that I find that all of the Marks are invalid for lack of distinctive character pursuant to s.3(1)(c) to the extent pursued by the Defendants, i.e. as set out in the emphasised specifications above. The Builder House Mark and Builder Word Mark are also invalid to the same extent pursuant to s.3(1)(d).

With regards to the balance of the specifications, I accept the Defendants submissions, and find that the Marks are prima facie not incapable of acting as an indication of origin for the goods and services described therein, but do so as ordinary English words with, in the case of the stylised Marks, weak and unmemorable stylisation.

Given the extent of my findings that the Marks are devoid of inherent or acquired distinctive character, and given that reputation requires a certain degree of knowledge of the earlier mark per C-375/97 Chevy, the Claimant has not satisfied me on the evidence that the Marks or any of them have a reputation with a significant part of the public in the UK or an enhanced distinctive character capable of impacting on the assessment of infringement. That is because I have no reliable evidence on market share, very little reliable evidence on the size of investment in the Marks in the UK, and insufficient information on the intensity and geographical extent of the use of the Marks, and the duration of use is even shorter than for assessing acquired distinctiveness, the relevant dates being earlier, in July 2022 for Category 2 infringement and 17 October 2022 for Category 1 infringement.

I have already found that they would not.

My finding that the Builder Word Mark and the Builder Home Mark lack inherent or acquired distinctiveness and are invalid to the extent set out in the emphasised specifications above, means the Category 1 alleged infringement claim must fail and be dismissed. Mr Howe in closing appeared to accept that there was only room for a finding of infringement if I found the Builder Word Mark to be valid, which I have not in respect of the uses complained of. The Claimant has accepted by Ms Gupta’s evidence that it has no complaints about D2’s use of ‘Builder’ save when it is capitalised and not used in a title or heading. Although D2 has used ‘Builder’ in the UK as complained of, which is identical to the Builder Word Mark and highly similar to the Builder House Mark (given my findings about the weakness of the stylisation of that mark), I am satisfied that it has only used it together with preceding words (‘App Builder’, ‘Android App Builder’ etc) (and never solus), as an ordinary and descriptive word to describe the purpose of those products. Accordingly: (i) the average consumer would perceive that use as merely descriptive and not use in a trade mark sense as an indicator of the origin of goods and services, and so would not be likely to be confused for the purposes of s. 10(2)(b) TMA; (ii) s. 10(3) TMA would fail for lack of distinctive character of the Marks; and (iii) as Mr Howe acknowledged in closing if the Court was satisfied that the D2’s use was descriptive, D2 has the benefit of the s. 11(2)(b) TMA defence. For those reasons, I will not go on to consider Issues 4 – 7.

My finding that the Builder.ai Figurative Mark lacks inherent or acquired distinctiveness such that they are invalid to the extent sought by the Defendants means the Category 2 alleged infringement claim must fail and be dismissed, as that requires the admitted use of identical signs by D2 to be in relation to goods and services identical to those for which they are registered. Even if I am wrong about the distinctiveness of the Builder.ai Figurative Mark, I would find that the Category 2 alleged infringement claim cannot succeed. I will deal with it briefly.

The question, per Lifestyle Equities, is whether the average consumer would consider the website (or in this case D2’s LinkedIn Post) to be directed at him or her, as a consumer in the UK. I am satisfied that on an assessment of all the circumstances of the case, the average consumer would not. That is because:

For that reason even if the Builder.ai Figurative Mark was valid across the Class 9 and Class 42 specifications, the Category 2 alleged infringement claim would fall at the first hurdle and be dismissed. Accordingly I will not go on to consider Issues 9, 10 and 11.

I have already dealt with Issue 12 (whether the Defendants have satisfied the Court of a s. 11(2)(b) TMA defence) in [137] above. My determination set out above mean that the final issues fall away, those being Issue 13 (due cause) and Issue 14 (joint tortfeasorship).

The Claimant’s claims for infringement of the Marks are dismissed.

The Defendants succeed on the counterclaim. All of the Marks are invalid for lack of distinctive character to the extent indicated in the emphasised specifications set out in the body of this judgment.