1 This is the applicant’s (V V Technology Pte Ltd) appeal against the decision of the learned Principal Assistant Registrar Mark Lim (“PAR”) in Twitter, Inc. v V V Technology Pte Ltd [2022] SGIPOS 4 (“GD”). The PAR decided that the respondent’s (Twitter, Inc) opposition to the applicant’s registration of in Class 42 (“the Application Mark”) succeeded under both ss 8(2)(b) and 8(7)(a) of the Trade Marks Act (Cap 332, 2005 Rev Ed) (“the Act”). The applicant is appealing against both grounds of the PAR’s decision.

2 By way of brief background, the applicant had first applied to register the Application Mark on 10 September 2018 (“the Relevant Date”). According to the applicant, the Application Mark has a pictorial representation of the applicant’s initials (V V). It also depicts a hummingbird. The applicant had apparently chosen the hummingbird “given that symbolic parallels could be drawn between a hummingbird, which is nimble given its small size but also capable of travelling great distances, and the [applicant], which is committed to being responsive to market demand and to the wide-ranging needs of the communities it serves”. ¹ As evidence of how the Application Mark was derived, the applicant pointed to its Facebook posts dated 7 October 2020 and 4 November 2020. However, as the PAR noted, these Facebook posts were published more than two years after the Relevant Date.

3 The Application Mark had been registered for a very broad range of goods and services in Classes 3, 9, 21, 29, 30, 31, 32, 33, 35, 38, 39, 41, and 43. Some of these goods and services overlap with the goods and services for which the respondent’s mark has been registered, such as, “computer software applications, downloadable” in Class 9, “advertising” in Class 35, “providing online forums” in Class 38, and “entertainment services” in Class 41.

4 After the Application Mark was published for opposition purposes on 24 May 2019, the respondent filed its Notice of Opposition on 24 September 2019. The respondent relied on its prior registration for (“the Registered Mark”). The goods and services offered under the Registered Mark allow individuals or companies to provide and/or consume an extensive range of services and information. The PAR gave his decision in favour of the respondent on 11 March 2022.

5 After hearing the parties on 15 August 2022 and having had the benefit of the parties’ further submissions on 29 August 2022, I took some time to consider the matter. Having done so, I have decided that the applicant’s appeal should be dismissed. I set out my full reasons for this decision in this judgment. Due to the length of this judgment and the various issues raised, I have set out a detailed table of contents which I would suggest the reader to peruse first so as to form a broad picture of the structure I have adopted for this judgment. As with other cases, my hope is that a reader would be able to gain a broad understanding of the judgment simply by reading the table of contents alone. It is not practical to reproduce the table of contents (which runs into three pages) in the body of this judgment, but it is available at the beginning of the PDF version of this judgment.

6 For further ease of understanding, I provide some background of the parties before considering the substantive merits of the appeal.

7 The applicant is a technology start-up which was established to leverage on advancements in artificial intelligence, blockchain technologies, cloud computing and big data to drive the world’s transition to smart commerce. The applicant is part of the V V group of companies, along with two other companies based in China.

8 The applicant has been engaged in the development of a mobile application called “V V Life” (“the App”) since December 2018. By the applicant’s account, the App is meant to serve as a user’s smart personal concierge by operating as a single platform to host a wide range of products and services that cater to a user’s personal lifestyle needs. As of the date of the hearing before me, the App has not been launched. The applicant had intended for the Application Mark to be prominently displayed at the bottom of a user’s handphone screen when the App is being used. At the same time, the applicant’s company name () will also be depicted at the top left-hand corner of the screen in the following manner:

9 The respondent was founded on 21 March 2006 and was incorporated in the State of Delaware, United States of America on 19 April 2007. It has been a publicly listed company on the New York Stock Exchange since 2013. It owns and operates the Twitter platform, which is a microblogging and social networking service where registered users can post and interact with each other with messages known as “Tweets”. It is undisputed that Twitter is one of the largest social networks in the world today. The Twitter platform has been accessible in Singapore since 2006. According to statista.com, there were over 1.3 million Twitter users in Singapore as of 11 June 2020. After I heard the parties in August 2022, Twitter was bought over by Mr Elon Musk. However, nothing in the present case turns on this fact.

10 Between 2015 and 2019, the respondent’s approximate annual worldwide revenue under the “Twitter” brand ranged from USD 2.22 billion in 2015 to USD 3.46 billion in 2019. The respondent spends a large amount on advertising and promotions. Indeed, their worldwide marketing expenditure ranged from USD 717 million to USD 957 million a year between 2015 and 2019. Much of this expenditure is spent on digital marketing with a global reach.

11 The Twitter platform operates under the Registered Mark and its variations. Mr Jack Dorsey, the respondent’s co-founder, had said that “Twitter means a short inconsequential burst of information, chirps from birds” and this name was chosen as it best embodied the instantaneous nature of the platform. ² As such, the respondent’s brand strategy since its founding has been towards cultivating the symbol of a bird as synonymous with the respondent and its goods and services. In this regard, the respondent has used and promoted a variety of bird logos in connection with its products and services since 2006, including the following in various forms:

For convenience, I shall refer to these collectively as the “Twitter Bird Logos”. ³

12 While the respondent was not able to provide the precise dates on which each of the Twitter Bird Logos was introduced, an article in The New York Times Magazine entitled “Who Made That Twitter Bird?” (8 August 2014) ⁴ depicted the evolution of the bird logo as follows:

13 For present purposes, it does not appear that every single one of the Twitter Bird Logos has been used in Singapore. More specifically, screenshots of The Straits Times’ online Twitter profile every year from 2009 to 2020 show that the only following bird logos were used: (2009), (2010 and 2011), and (2012, 2013 and 2020). ⁵

14 Having set out the background which led to the PAR’s decision, I turn now to consider the two substantive grounds upon which the PAR’s decision was based on, and which the present appeal turns on.

15 The first ground on which the PAR’s decision rests is s 8(2)(b) of the Act. This section of the Act provides as follows:

16 As such, to succeed on the ground provided for in s 8(2)(b), it must be shown that (a) the competing marks are similar, (b) the competing specifications are identical or similar, and (c) if the first two conditions are satisfied, there exists a likelihood of confusion on the part of the public as a result.

17 In David Llewelyn, Ng Hui Ming and Nicole Oh Xuan Yuan, Cases, Materials and Commentary on Singapore Intellectual Property Law (Academy Publishing, 2018) (“David Llewelyn”), the learned authors write that the Court of Appeal in Staywell Hospitality Group Pty Ltd v Starwood Hotels & Resorts Worldwide, Inc and another and another appeal [2014] 1 SLR 911 (“Staywell”) endorsed a step-by-step approach to whether a trade mark registration should be refused or invalidated on the basis of s 8(2) (at para 07.104). This is in contradistinction to the “global appreciation approach” applied in Europe after Sabel BV v Puma AG, Rudolf Dassler Sport [1998] RPC 199. More specifically, under the step-by-step approach, the three cumulative conditions are assessed systematically: the first two conditions are assessed individually before the final condition is assessed in the round. I shall discuss some of these general principles in greater detail below.

18 Before I do so, it is helpful to ascertain the relevant issues that arise from the parties’ overall cases in relation to s 8(2)(b) of the Act. In this regard, the applicant’s case is that the PAR had erred in finding that (a) the Application Mark and the Registered Mark are similar overall; and (b) a likelihood of confusion between the competing marks was established. In response to the applicant’s case, the respondent’s case was simply that the requirements under s 8(2)(b) are satisfied as there exists a likelihood of confusion due to the parties’ abstract bird devices sharing numerous key features. The competing marks are also to be used in closely overlapping fields.

19 Given the parties’ respective cases, I will deal with the following two general issues:

(a) whether the competing marks are similar overall; and

(b)  whether there is likelihood of confusion.

I do not deal with whether the competing specifications are identical or similar since the applicant does not dispute the PAR’s finding that they are similar in the present appeal.

20 Obviously, there are many specific issues that lie beneath these general issues. Indeed, Ms Meryl Koh (“Ms Koh”), who appeared for the applicant, and Mr Aaron Thng (“Mr Thng”), who appeared for the respondent, raised several important points of principle in both their oral and written submissions. Given the extensive manner in which the present appeal was argued before me, I take the opportunity to set out my views on the related issues in some detail.

21 First, in relation to the general issue of whether the competing marks are similar overall, I will deal with the following specific issues:

(a) the correct approach to understanding and applying the concept of “distinctiveness” in the marks-similarity inquiry, including whether acquired technical distinctiveness should be considered at this stage;

(b) considering my conclusions on the correct application of the concept of “distinctiveness”, whether the competing marks are visually similar, and what the appropriate approach is in relation to animal marks; and

(c) whether the competing marks are conceptually similar, including the correct level of conceptual analysis, and the correct weight to attribute to conceptual similarity in cases where the marks are visually similar.

22 Second, in relation to the general issue of whether there is a likelihood of confusion, I will deal with the usual specific issues of (a) the similarity of the competing marks; (b) the similarity of the services; and (c) the likelihood of confusion on the part of the public given the similarities.

23 With the above outline in mind, I turn to the first general issue, which is whether the competing marks are similar overall.

24 The general principles for the evaluation of marks-similarity are not disputed. These have been set out in several Court of Appeal decisions, such as Staywell and Hai Tong Co (Pte) Ltd v Ventree Singapore Pte Ltd and another and another appeal [2013] 2 SLR 941 (“Hai Tong”). In this regard, I adopt the PAR’s succinct summary of these principles with the appropriate modifications (see the GD at [32]).

25 First, the distinctiveness of the earlier trade mark is an integral factor in the marks-similarity inquiry. More specifically, the greater the earlier mark’s technical distinctiveness, the higher the threshold before a competing mark is considered dissimilar to it (see Staywell at [25]). This is because highly distinctive marks, unlike descriptive or laudatory marks, are strong badges of origins of the goods or services in question. These marks therefore deserve stronger protection to protect their strong source-denoting ability. The outcome is that, to overcome an opposition based on a highly distinctive mark, an applicant will need to prove “to a more compelling degree that his mark or sign is indeed dissimilar from the registered mark” (see Hai Tong at [30]).

26 Second, there are three aspects of the evaluation of marks-similarity, namely, visual, aural, and conceptual similarities. However, there is no requirement that all three similarities need to be made out before the marks or signs that are being compared may be found to be similar. Indeed, the Court of Appeal in Staywell (at [18]) held that a formulaic approach should not be adopted, and the similarity of marks is ultimately a matter of impression rather than one that can be resolved as a quantitative or mechanistic exercise (at [17]). Rather, the task is to “ultimately conclude whether the marks, when observed in their totality, are similar rather than dissimilar” and that the three aspects of similarity are meant to work as signposts to guide the court’s inquiry rather than as a “checkbox exercise in which a tick, however faint it might be, in any one box must compel the court to find that the marks are similar when a sensible appraisal of the marks as a whole would show otherwise” (at [17]). At the marks-similarity stage, the assessment is mark-for-mark (without consideration of any external matter), and thus the relative weight and importance of each aspect of similarity having regard to the goods/services should not be considered there. However, the court does not ignore the reality that the relative importance of each aspect of similarity might vary from case to case, and such considerations are properly reserved at the stage when the court evaluates whether there is a likelihood of confusion as that is when the court is called upon to assess the effect of the similarity between the marks on the perception of consumers (see Staywell at [20], citing Hai Tong at [40(b)]).

27 Third, the court assesses the two contesting marks or signs from the lens of the average consumer who would exercise some care and good sense in making purchases, and not that of an unthinking person in a hurry. However, at the same time, this average consumer has “imperfect recollection”. The two contesting marks or signs should not be compared side-by-side or examined in detail for the sake of isolating particular points of difference because “the person who is confused often makes comparison from memory removed in time and space from the marks” (see Hai Tong at [62(a)], citing the High Court decision of MediaCorp News Pte Ltd v Astro All Asia Networks plc [2009] 4 SLR(R) 496 (“MediaCorp”) at [33]).

28 Fourth, the similarity of two contesting marks or signs must normally be assessed by reference to the overall impressions created by the marks or signs, bearing in mind their distinctive and dominant components (see Staywell at [26]). As such, the test for marks-similarity is not one of substantial reproduction such as in copyright law (see the High Court decision of Polo/Lauren Co LP v United States Polo Association [2016] 2 SLR 667 (“Polo/Lauren Co”) at [17]). Further, where a mark lacks a dominant or distinctive component, it should be construed as a “unitary whole rather than just focus on one part of it” (see the High Court decision of Monster Energy Company v Glamco Co, Ltd [2021] 3 SLR 319 (“Monster Energy”) at [58]). In this respect, it bears reiterating that the assessment of marks-similarity is “mark-for-mark without consideration of any external matter” (see Staywell at [20]), and without regard to “any external added matter or circumstances” (see Hai Tong at [40(b)]).

29 I turn now to apply these general principles to the competing marks and will explore some of the legal issues more deeply while doing so.

30 I turn first to consider the concept of distinctiveness, since the PAR had framed this as a threshold enquiry in the marks-similarity inquiry. I should also note that the parties have made substantive submissions on whether acquired distinctiveness should be considered at the marks-similarity inquiry. These submissions on the concept of distinctiveness require me to consider some deeper conceptual issues before I can apply the law to the present facts.

31 To my mind, there are two unsettled conceptual issues in relation to the concept of “distinctiveness” in the marks-similarity inquiry. These may be discerned from the PAR’s reasoning process.

32 First, the PAR held that “[d]istinctiveness is integrated into the marks-similarity comparison, and is not a separate step”, citing Staywell at [30] (see the GD at [35]). However, the PAR then immediately proceeded to consider it separately as a “threshold enquiry”. This appears to be in contradiction to the Court of Appeal’s holding in Staywell. Indeed, as I will discuss below, there are other cases that have similarly treated “distinctiveness” as a separate threshold enquiry. As a matter of principle, it is helpful to clarify the correct approach in applying the concept of distinctiveness in the marks-similarity inquiry.

33 Second, following from his decision to consider distinctiveness as a threshold enquiry, the PAR then explained the non-technical and technical aspects of distinctiveness (see the GD at [36]–[38]). He further explained that technical distinctiveness may be inherent or acquired (see the GD at [39]). He then proceeded to consider the Registered Mark’s level of inherent technical distinctiveness (see the GD at [49]–[54]). Subsequently, the PAR declined to rule on whether it is permissible for him to consider acquired technical distinctiveness (see the GD at [58]). However, the PAR did not deal with the non-technical aspect of distinctiveness, which gives rise to several questions. For instance, does non-technical distinctiveness have both inherent and acquired aspects? And how are these to be applied in the marks-similarity inquiry? Again, as a matter of principle, it is helpful to clarify the correct understanding of the concept of distinctiveness in this regard.

34 I turn first to the meaning of “distinctiveness” in the marks-similarity inquiry. This is important because the exact meaning of “distinctiveness” feeds directly into the marks-similarity inquiry. Indeed, the Court of Appeal had held in Staywell that “distinctiveness (in both its technical and non-technical senses) is a factor integrated into the visual, aural and conceptual analysis as to whether the competing marks are similar” [emphasis in original in italics; emphasis added in bold underlined italics] (at [30]). The court’s characterisation of “distinctiveness” as a single factor suggests that it is a unitary concept within the marks-similarity inquiry. But is it?

35 In a helpful article, Paul McClelland in “Technical Distinctiveness and the Step-By-Step Approach” (2020) 32 SAcLJ 74 (“Paul McClelland”) points out that the introduction of the concept of “technical distinctiveness” into Singapore trade mark law was recent. Moreover, the expression is not found in the Act (at 74). The concept had originated from the English Court of Appeal decision of Budejovicky Budvar Narodni Podnik v Anheuser-Busch Inc [2002] EWCA Civ 1534 (“BUD”). The concept was then imported into Singapore law in Staywell. It is therefore useful to refer to BUD, where Lord Walker had said (at [39]):

36 However, as McClelland points out, Lord Walker’s “technical sense” was intended as an illustration of existing principles in his own words. Rather than setting out any new distinction, Lord Walker was likely using the “technical sense” of distinctiveness to show the right approach to the language of s 46(2) of the UK Trade Marks Act 1994 (c 26) (use of a trade mark in a form differing in elements that does not alter the distinctive character of a mark in the form in which it was registered). It is worth noting that Lord Walker did not mention this distinction between technical and non-technical distinctiveness again in his judgment. Indeed, McClelland’s research reveals that neither the passage in BUD nor the term “technical distinctiveness” appears to have been mentioned in subsequent decisions of the UK courts or those of other Commonwealth jurisdictions (see Paul McClelland at 76).

37 In Singapore, the Court of Appeal discussed the concept of “technical distinctiveness” in Staywell for the first time in our case law, and is thus a recent addition to the trade mark terminology. The court did this by relying on Lord Walker’s speech in BUD and then drawing a distinction between technical and non-technical distinction. For convenience, I shall term this the “Staywell Distinction”. The Court of Appeal first said that distinctiveness in the non-technical sense simply refers to what is outstanding and memorable about the mark. This “non-technical distinctiveness” and memorable components of the mark are those that tend to draw the consumer’s attention and tend to stand out in the consumer’s imperfect recollection. The court explained that that is why the court is entitled to have special regard to the distinctive or dominant components of a mark, even while it assesses the similarity of the competing marks as composite wholes (see Staywell at [23]).

38 In contrast, “technical distinctiveness” stands in contradistinction to descriptiveness (ie, the words do not describe the goods or services in question). Namely, “technical distinctiveness” refers to the ability of a mark to distinguish the goods or services of one particular trader from those of another, and functions as a badge of origin (see Staywell at [24]), ie, that which gives the mark the ability to perform its source-indicating function. It is in this context that the court said (at [24]):

It therefore appears that the distinction between inherent and acquired distinctiveness only applies to “technical distinctiveness”. The court did not draw a similar distinction between inherent and acquired “non-technical distinctiveness”. This is the understanding adopted by some local commentators as well (see Tan Tee Jim SC, Law of Trade Marks in Singapore (Sweet & Maxwell, 4th Ed, 2021) (“Tan Tee Jim”) at [8.065]).

39 The distinction between “non-technical distinctiveness” and “technical distinctiveness” pre-dates Staywell, albeit under different labels. The Staywell Distinction had merely recast this distinction into a different form.

40 Before Staywell, “non-technical distinctiveness” was simply analysed under the rubric of a dominant component within a mark. For example, in Hai Tong, the Court of Appeal, in dealing with visual similarity between the competing marks, considered that the textual elements of the marks in Lee Cooper Group plc v Levi Strauss & Co [1995] AIPR 457 stood out, or “talked”, because they were the dominant components of the marks in question (at [49]). This was also the case in Festina Lotus SA v Romanson Co Ltd [2010] 4 SLR 552 where the High Court noted that within the assessment of similarity between two marks, the court must consider the “overall impression” of the two marks, and in particular, their “distinctive and dominant components” (at [30]). In that case, a portion of the appellant’s submissions was also cited (at [29]) which attempted to explain which parts of the trade marks with crests and coat-of-arms would the consumers’ eyes be usually drawn to and what the average consumer would recall (ie, what is outstanding and memorable about the mark or distinctiveness in a “non-technical” sense). In Intuition Publishing Ltd v Intuition Consulting Pte Ltd [2012] SGHC 149, the High Court essentially adopted the “non-technical” distinctiveness analysis (without explicitly coining it as such) and considered whether the device component in the marks was displayed prominently, and consequently, whether it was “distinct and memorable” (which is similar to the phrasing used in Staywell at [23]) and could be recalled (at [64]).

41 Indeed, at least one academic commentator has taken the view that the Court of Appeal in Staywell was referring to the dominant component of a mark when it referred to “non-technical distinctiveness” (see Rob Batty, “Conceptual confusing similarity and pictorial trade marks” (2022) QMJIP 409 at 417). This is discernible from a close reading of Staywell where the line between the concepts of “distinctiveness” and “dominant component” was blurred at times (eg, at [28]: “that component or element is found to be the dominant and distinctive element of the mark in the non-technical sense” [emphasis added]). The learned author further notes that the “non-technical distinctiveness of an element of a trade mark could depend on the size of the element or how the element is positioned within a trade mark” (at 417). The commentator’s explanation regarding the factors which influence the non-technical distinctiveness of a mark coheres with how some SGIPOS decisions have understood the meaning of “non-technical distinctiveness” as well, where reference has also been made to factors such as the size of the element, how the element is positioned and whether it was in bold font, etc (see, for example, Redsun Singapore Pte. Ltd. v Tsung-Tse Hsieh [2015] SGIPOS 1 at [29]).

42 Similarly, before Staywell, the “technical” sense of distinctiveness was what the courts referred to when they analysed the concept of “distinctiveness” (without any appellation) in the marks-similarity inquiry. For example, in Hai Tong, the Court of Appeal explained that “[d]istinctiveness … connotes the sense that the trade mark is capable of saying something meaningful not about the goods or services in question, but rather, about whence these have come” (at [28]). A distinction was drawn between marks which are descriptive in nature as opposed to those which can function as a badge of origin (ie, those which are “technically distinctive”) at [30]:

The exposition by the Court of Appeal in Hai Tong harks back to the definition laid down in Staywell, where the key to finding “technical distinctiveness” is in the assessment of the mark’s ability to function as a badge of origin. There is also a similar contrast drawn in Hai Tong between marks which are used to describe the goods and marks which are technically distinctive, just as how it was stated in Staywell at [24] that “[d]istinctiveness in the technical sense …, usually stands in contradistinction to descriptiveness”. The concept of “technical distinctiveness” had also been discussed in cases examining the requirement of whether a mark is “capable of distinguishing” its goods from another trader’s goods under the definition set out in s 2(1) of the Act. For example, in the High Court decision of Love & Co Pte Ltd v The Carat Club Pte Ltd [2009] 1 SLR(R) 561 (“Love & Co”) at [33], it has been said that an immediately registrable trade mark must possess “a ‘distinctive character’ status and it is immediately capable of distinguishing the goods or services of a trader from those of other traders, in the sense that it can immediately function as a clear badge of origin of that trader”, and the court may examine whether the mark has any “intrinsic or inherent features or characteristics that are sufficiently unique to enable the intended mark to immediately function (and not potentially function in the future through subsequent use by the promoter of the mark) as a readily obvious and reliable badge of origin” (at [53]) [emphasis in original] (see also, the High Court decision of Societe Des Produits Nestlé SA and another v Petra Foods Ltd and another [2014] SGHC 252 at [158] and [161]). As an aside, I also observe that the distinction between the concepts of inherent technical distinctiveness and acquired technical distinctiveness (through subsequent use) was recognised in Love & Co, though not coined as such.

43 As such, before the Staywell Distinction, the case law was quite stable in terms of the taxonomy employed to describe two well-accepted concepts in trade mark law, which are: (a) the dominant/outstanding and memorable component of a mark which stands out in the average consumer’s imperfect recollection (or what Staywell at [23] labelled as “non-technical distinctiveness”); and (b) the ability of a mark to distinguish the goods or services of one particular trader from those of another (or what Staywell at [24] labelled as “technical distinctiveness”).

44 Taking a step back, the taxonomical consequence of the Staywell Distinction is to recast the hitherto distinction between the “outstanding and memorable” aspects of a mark and the “distinctiveness” of a mark into “non-technical distinctiveness” and “technical distinctiveness”, respectively. Perhaps owing to the novelty of the distinction (in name but not in substance), the distinction between “non-technical distinctiveness” and “technical distinctiveness” has resulted in, as McClelland points out (see Paul McClelland at 77), some taxonomical uncertainty. First, in Staywell, “technical distinctiveness” was used in conjunction with, but also independently from, “inherent” distinctiveness. The Court of Appeal had said (at [25]):

While the court’s reasoning can be clearly understood, McClelland suggests that the passage would read more clearly if the words “technically” and “inherent” were included in the second and third italicised instances above respectively – matching the phrasing used in the first highlighted instance (see Paul McClelland at 78). I respectfully agree. While this may result in the sentence becoming rather unwieldy, a consistent set of terminology would provide greater clarity to the precise concepts being conveyed.

45 The taxonomical uncertainty introduced by the Staywell Distinction was further extended in Caesarstone Sdot-Yam Ltd v Ceramiche Caesar SpA [2016] 2 SLR 1129 (“Caesarstone (HC)”). In that case, the High Court seemed to regard the term “technical distinctiveness” as synonymous with “inherent distinctiveness” (at [26]):

Yet, the Staywell Distinction did not in any way equate “technical distinctiveness” with “inherent distinctiveness”.

46 Shortly after Caesarstone (HC) was decided, the case of Allergan, Inc and another v Ferlandz Nutra Pte Ltd [2016] 4 SLR 919 concluded. Once more, the High Court essentially equated the concept of technical distinctiveness to that of inherent distinctiveness (at [45]): “a newly-invented word (ie, a freshly coined word which bears no meaning) will carry a high degree of technical or what is sometimes called inherent distinctiveness” [emphasis added], albeit it is noted that the case was decided by the same learned judge who decided Caesarstone (HC).

47 Similarly, the High Court (again, under the same learned judge) in Rovio Entertainment Ltd v Kimanis Food Industries Sdn Bhd [2015] 5 SLR 618 (“Rovio”) seemed to have amalgamated the concepts of the dominant component of a mark with “technical distinctiveness” (at [76]):

Yet, this would appear to go against the Court of Appeal’s guidance in Staywell that the dominant component of a mark (or its non-technical distinctiveness) and its technical distinctiveness are different concepts.

48 A new term was further added to the mix in the case of Jamal Abdulnaser Mahmoud Al Mahamid v Global Tobacco Manufacturers (International) Sdn Bhd [2015] SGHC 42. In that case, the High Court (at [24]) first set out the guidance of the Court of Appeal in Staywell, but then made the comment that distinctiveness was a relevant consideration in Hai Tong “though there was no separate analysis of the difference between general and technical distinctiveness” [emphasis added]. Thus, a new term was added to the mix (ie, general distinctiveness) which might cause further confusion.

49 Further, more recent cases have focused mainly on “technical distinctiveness” without alluding to “non-technical distinctiveness”. This is notwithstanding the Court of Appeal’s holding in Staywell (at [30]) that “distinctiveness (in both its technical and non-technical senses)” should be considered in the marks-similarity inquiry. For example, in Monster Energy, the High Court referred simply to “technical distinctiveness” (at [47]), in the sense of “the capacity of a mark to operate as a badge of origin”. The court further acknowledged, following Staywell, that (at [47]):

The court in Monster Energy therefore recognised the distinction between inherent technical distinctiveness and acquired technical distinctiveness. However, the court in Monster Energy did not explicitly refer to “non-technical distinctiveness” although it did refer to the “dominant component” of the marks in relation to one approach for comparing the aural similarity of competing marks (at [61]).

50 In Dr August Wolff GmbH & Co KG Arzneimittel v Combe International Ltd [2021] 4 SLR 626 (“Dr August Wolff”), the High Court recognised the Staywell Distinction (at [18]). However, the court appeared to then use a new expression, “distinctive character”, to connote what is effectively technical distinctiveness. Thus, the court said that “invented words are considered to have a very high degree of distinctive character” [emphasis added] (at [21]). Subsequently, when analysing visual similarity, the court held that the prefix of the mark concerned “has weak distinctive character and, as such, is not likely to be perceived as the distinctive and dominant elements of the overall impressions conveyed by the marks” (at [30]). Here, the court appeared to be referring to “non-technical distinctiveness” but used the same expression “distinctiveness” to refer to both technical and non-technical distinctiveness.

51 In Digi International Inc v Teraoka Seiko Co, Ltd [2021] SGHC 165 (“Digi International”), the High Court acknowledged the distinction between distinctiveness in the technical and non-technical senses (at [87]), but made no further reference to the term “non-technical” distinctiveness throughout the rest of the analysis concerning distinctiveness (see Digi International at [91]–[119]). Instead, the visual elements of the marks were described as “eye-catching” (at [124]). This was similarly done in TMRG Pte Ltd and another v Caerus Holding Pte Ltd and another [2021] SGHC 163, where the High Court acknowledged that non-technical distinctiveness referred to what was outstanding and memorable about the mark (at [25]), but there was no further usage of the term “non-technical” distinctiveness in the rest of the analysis (at [27] onwards), albeit the court’s restatement of the plaintiff’s submissions which included references to what was the “most visually noticeable and memorable” parts of the marks (at [27]).

52 These examples illustrate the taxonomical uncertainty that has resulted after Staywell. This is not merely an academic issue as it directly affects the correctness of the analytical process and the comprehensibility of decisions that may use inconsistent terms. In this regard, I agree with McClelland’s view that the taxonomical uncertainty in the cases may seem trivial, but it would ultimately help practitioners and decision-makers if clear guidance was provided to assist with both the formulation of coherent arguments and the identification of relevant considerations in the assessment process (see Paul McClelland at 79).

53 The taxonomical uncertainty can also be located in the leading local textbooks. In her latest edition of the textbook, Law of Intellectual Property of Singapore (Sweet & Maxwell, 3rd Ed, 2021) (“Ng-Loy Wee Loon”) at p 386, Professor Ng-Loy Wee Loon SC (“Professor Ng-Loy”) does not even refer to the distinction between the technical and non-technical aspects of distinctiveness in her discussion of the marks-similarity inquiry. Rather, the learned author simply defines “distinctiveness” to mean a “mark’s ability to function as the badge of origin of the goods or services in question”. Her analysis is worth citing in full, as follows (at p 386):

54 Indeed, Professor Ng-Loy discusses the dominant component of a mark without referring to the term “non-technical distinctiveness” (see Ng-Loy Wee Loon at pp 386–387). This shows that the learned author was quite intentional in not adopting the twin aspects of distinctiveness as laid down in the Staywell Distinction. While she does not explain why she has done this, it may well be inferred that the learned author no longer finds the distinction helpful.

55 In Tan Tee Jim, the learned author (“Tan”) does allude to the distinction between the technical and non-technical aspects of distinctiveness (at pp 384–385). Summarising the law, Tan writes (at p 386):

56 Yet, when discussing the applicable principles in detail, Tan appears to use “distinctiveness” only in its non-technical sense. For example, the learned author states that “the visual similarity of the marks in issue is to be assessed by reference to the overall impression created by the marks, bearing in mind their distinctive and dominant components” [emphasis added] (at p 389). Tan is here referring to “non-technical distinctiveness”, because he writes that this exercise is to be done by recourse to the following factors: (a) length of the marks; (b) structure of the marks; and (c) whether the same letters are used in the marks (at p 389). These factors only relate to “non-technical distinctiveness” and not “technical distinctiveness”. It is ultimately unclear if Tan intended, as Professor Ng-Loy likely did, to depart from the Staywell Distinction.

57 Finally, in David Llewelyn, the learned authors again refer to the Staywell Distinction (at para 07.108). They then say that the cases of Caesarstone (HC) and Rovio have provided further clarity on the issue of technical and non-technical distinctiveness (at paras 07.109 and 07.111). However, they do not then refer to the Staywell Distinction in their substantive discussion of the marks-similarity inquiry. It is therefore unclear whether the learned authors find the distinction a helpful one in the application of the relevant principles.

58 This brief survey of the leading local textbooks shows that the academic commentators have either abandoned the Staywell Distinction (as seems to be the case in Ng-Loy Wee Loon) or referred to it but do not actively engage with it (in both Tan Tee Jim and David Llewelyn). Ultimately, the lack of consistency in the discussion of the Staywell Distinction in the textbooks leads me to infer that: (a) it has been perceived as giving rise to taxonomical uncertainty; (b) it has not resulted in much practical utility given the previous stable taxonomy found in the case law; or (c) both. For completeness, I should mention that I did not refer to other textbooks (such as Susanna H S Leong, Intellectual Property Law of Singapore (Academy Publishing, 2013)) which pre-date Staywell. They would not have considered the Staywell Distinction, although I am certain that, had they done so, their insights would prove to be similarly helpful as the textbooks I have referred to above.

59 Having considered the relevant authorities, it would be opportune to provide some comments on the meaning of “distinctiveness” in the marks-similarity inquiry. In my respectful view, the Staywell Distinction, while founded on a sound distinction, has not resulted in conceptual clarity but has given rise to taxonomical uncertainty in the lower courts and tribunals. I am mindful of my position as a trial judge. Bound by the doctrine of stare decisis, I am compelled to apply the distinction between “technical distinction” and “non-technical distinction”, as that distinction has been laid down by a higher appellate court.

60 But if I were not so bound, I would respectfully have suggested that the word “distinctiveness” should simply mean “technical distinctiveness”, without the need to attach any appellation. I say this for a few reasons. First, Lord Walker in BUD likely never intended to introduce any taxonomical change into the law. Second, that Lord Walker did not intend any such change is also borne out by the case law following BUD: the expression “technical distinctiveness” does not appear to have been adopted elsewhere. Third, there was and remains a stable taxonomical practice of using “dominant component” to refer to “non-technical distinctiveness” and “distinctiveness” to refer to “technical distinctiveness”. This taxonomical practice has been stable in practice and the textbooks before and (arguably) even after Staywell. In my respectful view, there is simply no sufficient benefit to depart from that hitherto well-settled practice in favour of a new taxonomy that has not been followed in other jurisdictions. Indeed, the way to rationalise the Staywell Distinction may be that it did not purport to introduce any new taxonomy to this area of trade mark law, but simply provided an explanation to the existing application of the concept of “distinctiveness”. While not a perfect analogy, this is akin to how the Court of Appeal sought to explain Lord Hoffmann’s “assumption of responsibility” test in Transfield Shipping Inc v Mercator Shipping Inc (The Achilleas) [2009] 1 AC 61 as an explanation for the existing test in relation to remoteness of damages in Singapore contract law (see the decision of the Court of Appeal in MFM Restaurants Pte Ltd and another v Fish & Co Restaurants Pte Ltd and another appeal [2011] 1 SLR 150).

61 But bound as I am by higher authority, I can only suggest a more consistent use of the relevant expressions to aid in the analytical process and provide clarity to those seeking to understand our intellectual property jurisprudence. I would therefore propose a consistent use with the understanding that “distinctiveness” bears the following three meanings in so far as it applies to the marks-similarity inquiry: (a) inherent technical distinctiveness, (b) acquired technical distinctiveness; and (c) non-technical distinctiveness.

62 In so far as inherent technical distinctiveness and acquired technical distinctiveness are concerned, it appears that the Court of Appeal in Staywell was making this distinction in relation to technical distinctiveness. Indeed, the IP Adjudicator in The Polo/Lauren Co, L.P. v United States Polo Association [2015] SGIPOS 10 recognised this by holding that “inherent” and “acquired” distinctiveness are sub-sets of the broader concept of “technical distinctiveness” (at [60(a)]:

63 I noted above that the Court of Appeal in Staywell did not expressly allude to the inherent/acquired distinction in so far as non-technical distinctiveness is concerned (see [38] above). It is thus important to have regard to the underlying concept behind the concept of “non-technical distinctiveness”, which, pre-Staywell, was about the “dominant component” of a mark. It therefore makes no sense to speak of “acquired” non-technical distinctiveness. This is because the dominant component of a mark cannot be “acquired”. It is either there or it is not. A mark cannot acquire an outstanding and memorable component through prolonged use (other than the mark actually evolving over time through conscious modification). Therefore, if at all, non-technical distinctiveness can only be inherent. Since it is not helpful to distinguish between inherent non-technical distinctiveness and acquired non-technical distinctiveness, I would suggest that the expression “non-technical distinctiveness” is used only.

64 With the above clarification of the meaning of “distinctiveness” in the marks-similarity inquiry in mind, I turn to the second unsettled issue, which is how “distinctiveness” should be applied in this inquiry.

65 I start with Staywell. In my view, the key exhortation from the Court of Appeal can be seen in the following paragraph (at [30]):

It is clear from this that both technical and non-technical distinctiveness are to be integrated into the marks-similarity inquiry, without being a separate step (ie, becoming a two-stage process). Yet, cases after Staywell have consistently treated distinctiveness as a threshold enquiry, which is effectively a separate step.

66 The most relevant example for present purposes is the PAR’s decision. The PAR had reasoned as such (see the GD at [35]):

67 The substantive effect of the PAR’s decision is to treat distinctiveness as a separate “threshold enquiry”. This would directly go against what the Court of Appeal had warned of in Staywell. Indeed, despite the guidance laid down in Staywell, there are numerous cases, both at the Registry level and in the High Court, that have treated distinctiveness as a threshold enquiry.

68 I begin by examining the cases pre-Staywell. From a brief survey of the various cases, it appears that the first significant decision that actually applied distinctiveness as a threshold enquiry was the High Court decision of The Polo/Lauren Co, LP v Shop In Department Store Pte Ltd [2005] 4 SLR(R) 816 (at [26]–[27]), though not explicitly stated as such. I cite the relevant paragraphs for convenience, before explaining how this was so:

The High Court began by explaining that the distinctiveness of the earlier trade mark is an important factor in the marks-similarity analysis under s 8(2)(b) of the Act, which encompassed the three aspects of visual, aural, and conceptual similarity (at [26]). The court then went on to suggest that in proceeding with the analysis it would “first have to consider whether the plaintiff’s mark can be considered to be so distinctive” (at [27]), before then applying the assessment of distinctiveness and making the finding that “the plaintiff’s word mark is somewhat distinctive but not strikingly so” (at [28]). From this, what can be observed is that the High Court effectively applied distinctiveness as a threshold step, although not expressly couched in such a manner.

69 This manner of treating the assessment of distinctiveness of the trade marks as a separate step was then perpetuated in several subsequent High Court cases such as Ferrero SPA v Sarika Connoisseur Cafe Pte Ltd [2011] SGHC 176 at [47] (citing the High Court decision of Ozone Community Corp v Advance Magazine Publishers Inc [2010] 2 SLR 459), which described the marks-similarity inquiry as a “two-step approach in analysing the similar of marks”. The High Court in that case (at [47]) undertook two separate inquiries in the following manner: (a) whether there was visual, aural and conceptual similarity between the allegedly infringing sign and the registered trade mark; and (b) whether the registered trade mark was distinctive. When this case later went on appeal, the Court of Appeal reviewed this “two-step approach” (see Sarika Connoisseur Cafe Pte Ltd v Ferrero SpA [2013] 1 SLR 531 (“Sarika”) at [20]):

The comment stemmed from the fact that the “two-step approach” may “inadvertently give rise to the misconception that the distinctiveness of the registered trade mark is either another aspect of, or an element in, the determination of similarity, separate from the three aspects of similarity” [emphasis in original].

70 But even after the Court of Appeal’s statements in Sarika (at [20]), and then later in Staywell (at [30]), that the distinctiveness of a mark is a factor integrated into the visual, aural and conceptual analysis, and is not a separate step within the inquiry, a good number of cases post-Staywell have continued to treat the distinctiveness assessment of the marks as a separate threshold inquiry. The earliest of which are possibly the SGIPOS decisions of Taylor, Fladgate & Yeatman Limited v Taylors Wines Pty Ltd [2014] SGIPOS 11 (“Taylor, Fladgate”) and Intel Corporation v Intelsteer Pte Ltd [2015] SGIPOS 2 (“Intel Corporation”).

71 Interestingly, in Taylor, Fladgate at [47], the learned AR relied on a statement from Hai Tong where the Court of Appeal suggested (at [26]) that the distinctiveness inquiry can be highlighted as a separate step for analytical clarity: “for the purpose of elucidating the analytical process, we have highlighted it [ie, the distinctiveness assessment] here as a separate step”. It is perhaps unfortunate that this throwaway line in Hai Tong (a case decided before Staywell) had been taken to mean that it was permissible for distinctiveness to be analysed as a separate step. The AR in Taylor, Fladgate then proceeded to analyse distinctiveness under a separate heading (at [48]–[63]), before moving on with the analysis of visual similarity (at [64] onwards).

72 Turning then to the SGIPOS decision of Intel Corporation, the learned PAR also made findings on the distinctiveness assessment as a separate threshold inquiry despite being “mindful” of the caution in Staywell, and the PAR had found it necessary to do so for “ease of analysis” due to the “voluminous evidence tendered and extensive submissions made in relation to this issue” (at [19]).

73 This manner of assessing the distinctiveness of the marks as a separate threshold continues to be done post-Staywell in the most recent decisions at the High Court. For example, in the decision of Combe International Ltd v Dr August Wolff GmbH & Co KG Arzneimittel [2022] SGHC 78, while the court was mindful that the assessment of distinctiveness was not a separate step in the overall analysis, it appeared to have treated it as a threshold inquiry. Indeed, the court had analysed the distinctiveness of the marks under a separate heading altogether (at [22]):

74 This has likely become common practice as it is probably simpler to deal with the assessment of distinctiveness of the marks as a separate threshold question, as opposed to truly integrating it into each step of the visual, aural, and conceptual marks-similarity analysis. Thus, it is done as a matter of practicality. This was candidly acknowledged by the High Court in Digi International (at [88]):

This concern for practicality was also apparent from the High Court decision of Dr August Wolff at [19]: “as a matter of practicality, I shall also start with this [ie, the distinctiveness assessment of a mark]”.

75 I appreciate that in all these cases that have been cited above, they are usually prefaced by or caveated with the statement that one is “mindful” of the Staywell guidance that distinctiveness is a factor integrated into the visual, aural, and conceptual analysis, instead of being a separate step within the marks-similarity analysis. Thus, it was not as if these cases were completely oblivious to the prohibition in Staywell. However, with that said, it seems substantively inconsistent with Staywell’s guidance to then proceed with applying the distinctiveness assessment as though it was a separate threshold inquiry (in contrast, see for example, the decisions of Societe Des Produits Nestle S.A. v Liwayway Marketing Corporation [2014] SGIPOS 5 and Time Inc. v Li San Zhong [2014] SGIPOS 14, where the Staywell guidance was faithfully adhered to by integrating the distinctiveness inquiry into the analysis as the learned PAR considered the issue of distinctiveness when dealing with each of the various aspects of similarity as necessary rather than considering distinctiveness as a separate threshold assessment).

76 In my respectful view, it is important for courts and tribunals to strictly follow the Court of Appeal’s holding in Staywell that distinctiveness is not a separate step within the marks-similarity inquiry. Rather, it is integrated into the visual, aural, and conceptual analysis. As such, and with respect, reasons of “convenience” and “practicality” are not good enough for courts and tribunals to unilaterally decide to depart from what is clearly a binding authority emanating from the Court of Appeal. However, to be fair to the adjudicators, I appreciate that the permissibility of treating the consideration of distinctiveness as a distinct step may have emanated from a line in Hai Tong where the Court of Appeal stated as such (at [26]): “Distinctiveness is considered within the assessment of similarity, as noted by this court in Sarika. However, for the purpose of elucidating the analytical process, we have highlighted it here as a separate step.” That said, that line must be treated with caution after the later Court of Appeal’s guidance in Staywell against doing this, even if for “ease of analysis” (see Staywell at [30]).

77 Quite apart from stare decisis, there are, in my view, good substantive reasons why distinctiveness should not be analysed as a separate step. The first of these reasons is the need for conceptual certainty. It will be recalled that the Court of Appeal in Staywell endorsed a step-by-step approach, in contradistinction to the “global appreciation approach”. As McClelland explains, the intended effect of dividing the assessment into three progressive steps is that “the issue of resemblance between the competing marks is distinct from the question of the effect of such resemblance” [emphasis in original] (see Paul McClelland at 80, citing Staywell at [20]). This is based on the trite law, established in previous cases such as Hai Tong at [40(b)] and Sarika at [17], that the marks-similarity inquiry should be carried out without reference to external factors. In my view, the danger of treating distinctiveness as a “threshold” inquiry is to mar the very conceptual clarity that the Court of Appeal emphasised in Staywell as the threshold consideration of distinctiveness may allow for the very permeation of extraneous factors into the marks-similarity inquiry (as will be explained further below).

78 Secondly, as McClelland points out, considering the step-by-step approach, the concept of “technical distinctiveness” will be found in more than one of the steps. As such, its precise meaning will depend on the context in which it is being applied. This applies to both technical distinctiveness as applied separately to the visual, aural, and conceptual comparisons within the first step, as well as to each of the three steps in the overall approach. Thus, to coalesce the issue of distinctiveness into a single “threshold” issue risks obscuring the context to which it is meant to be applied in. This very risk can be seen in the present case, where the PAR had considered inherent technical distinctiveness as a threshold enquiry (see the GD at [40]–[54]). However, in his analysis, he was really in fact considering inherent technical distinctiveness in relation to the competing marks’ visual similarity. In a case where more than visual similarity is relevant, the PAR’s approach risks shading the distinctiveness enquiry between the various separate considerations at a threshold stage.

79 For all these reasons, it is important for courts and tribunals to return to the Court of Appeal’s exhortation in Staywell and not treat distinctiveness as a separate, threshold enquiry.

80 Having considered what I had thought were unsettled issues in the case law, I come to the related problem that was argued extensively before me, that is, whether acquired technical distinctiveness, which would require the consideration of external matter, should be considered in the marks-similarity inquiry.

81 The problem is well summarised by PAR Gabriel Ong (“PAR Ong”) in GCIH Trademarks Limited v Hardwood Private Limited [2021] SGIPOS 6 (“GCIH”). Having set out the Court of Appeal’s holding in Staywell that the step-by-step approach means that the mark comparison is undertaken mark for mark without consideration of any added matter, PAR Ong elaborated on how distinctiveness played into all of this. In my respectful view, his explanation of the problem is extremely clear, and I set it out fully as follows (at [106]):

I shall have occasion to rely on PAR Ong’s analysis later in this judgment.

82 In a recent article, Vignesh Vaerhn and Avery Yew, “Distinctly Confusing: Clarifying the Applicability of Acquired Distinctiveness under Singapore Trade Mark Law” (2022) 34 SAcLJ 621 (“Vaerhn and Yew”), the learned authors (at 622) likewise refer to the two apparently contradictory propositions that were cited with approval in Staywell, namely (a) that distinctiveness is an integral factor in the marks-similarity inquiry (see Staywell at [25]); and (b) the assessment of marks-similarity is mark-for-mark without consideration of any external matter (see Staywell at [20]). The authors say that it is this apparent contradiction that has created confusion over the rightful point at which acquired technical distinctiveness is to be considered in the step-by-step approach.

83 Reduced to its essence, the key question is this: at which stage of the step-by-step approach should a trade mark’s acquired technical distinctiveness be considered – should it be considered at the marks-similarity stage, or the likelihood of confusion stage?

84 Before me, and in further submissions after the hearing, the applicant submits that acquired technical distinctiveness should not be considered at the marks-similarity inquiry for two reasons. First, the conceptual clarity heralded by the Court of Appeal in Staywell should be preserved. The applicant argues that acquired technical distinctiveness, which is based on the consideration of external matters, should not be considered in assessing marks-similarity, as it is meant to entail a clear and objective mark-for-mark comparison. The applicant submits that the Court of Appeal in Staywell had drawn a clear distinction between the issue of resemblance between the competing mark (ie, the marks-similarity stage) and the question of the effect of such resemblance (ie, the likelihood of confusion stage). The issue of acquired technical distinctiveness impinges on the latter question of the effect of such resemblance, and hence should not be considered in the marks-similarity stage. The applicant further says that this is especially dangerous in the present case where any evidence of acquired technical distinctiveness of the Registered Mark only related to microblogging and social networking, and not for the broad range of goods and services asserted. Thus, the applicant argues that it would be illogical for acquired technical distinctiveness to be considered at the marks-similarity inquiry if the actual goods and services for which the Registered Mark has acquired technical distinctiveness in are different from the goods and services for which the opposed mark is applied for. ⁶

85 Second, the applicant further submits that deploying acquired technical distinctiveness at the marks-similarity stage is conceptually unsound due to “cognitive dissonance” produced upon application. The applicant points out that the finding of acquired technical distinctiveness should lead to greater protection of a mark (an applicant would have to show to a more compelling degree that the mark it seeks to register is dissimilar), however, the same evidence of use going towards acquired technical distinctiveness may instead show that consumers are so familiar with the mark that there is less likelihood of confusion (the recollection of the mark becomes better than perfect). The effect of the reputation of a mark depends on the external factors that may reduce or increase the likelihood of confusion as a matter of fact, for reasons that have nothing to do with the mark as it objectively appears. In essence, there is no reason why acquired technical distinctiveness should only lower the threshold for the finding of similarity between competing marks when the reputation of a mark may have a different effect under different circumstances. ⁷

86 As such, the applicant submits that acquired technical distinctiveness should not be considered at the marks-similarity inquiry but should be subsumed in the inquiry as to how the reputation of an opponent’s mark affects the likelihood of confusion, that is, at the last stage of the Staywell test. This, the applicant further argues, would allow future courts and registrars to make logical deductions about how reputation might change the threshold for the finding of confusing similarity in either direction, thereby avoiding the “‘reputation-therefore-confusion’ (or the reverse)” approach. Instead, the exact effect of the reputation should differ based on the specific facts of a case. This approach would also avoid the need to consider evidence of an opponent’s reputation twice at both the marks-similarity and likelihood of confusion stages, preserving the integrity of the step-by-step approach as mandated in Staywell. ⁸

87 The respondent rejected the views of the applicant in its main submissions and in further arguments after the hearing. First, the respondent submits that decisions in Singapore have considered both a mark’s inherent and acquired technical distinctiveness in assessing technical distinctiveness as part of the marks-similarity inquiry, citing, inter alia, Sarika at [19] and [36], Hai Tong at [31]–[33], Staywell at [24]–[25], and Ceramiche Caesar SpA v Caesarstone Sdot-Yam Ltd [2017] 2 SLR 308 (“Caesarstone (CA)”) at [28]–[31]. The respondent therefore urges me to apply these Court of Appeal cases because I am bound by them as a matter of stare decisis. ⁹

88 Second, the respondent argues that acquired technical distinctiveness should be considered at the marks-similarity stage because the marks-similarity assessment does not entail only an objective mark-for-mark comparison and requires consideration of the senior mark’s degree of distinctiveness to calibrate the threshold of protection. ¹⁰ Indeed, if the prohibition to consider external matters in Staywell were to be read as precluding consideration of acquired technical distinctiveness, it must follow that inherent technical distinctiveness similarly cannot be considered at the marks-similarity stage (which cannot be true since the Court of Appeal in Staywell at [31] considered the inherent technical distinctiveness of the mark in relation to hotel and hospitality services). ¹¹ The respondent argues further that the Court of Appeal did not mean to impose a blanket exclusion to all external matters at the marks-similarity stage and only sought to exclude external matters in so far as they do not relate to the issue of resemblance between the marks. ¹²

89 Lastly, the respondent submits that considering acquired technical distinctiveness will not result in cognitive dissonance (as alleged by the applicant). The respondent argues that the applicant had erroneously equated the concept of reputation with acquired technical distinctiveness which are distinct legal concepts: reputation refers to how widely recognised a mark is (the object of comparison is the public’s perception), while acquired technical distinctiveness simply examines how well a mark is relied upon as a badge of trade origin (practically, the comparison is more focused on the registered goods and services). A widely recognised mark may well be regarded as having no acquired technical distinctiveness – there is no cognitive dissonance. Since they are distinct concepts, the fact that reputation may weigh against a finding of confusion does not mean that it is internally contradictory for acquired technical distinctiveness to bolster a finding of marks-similarity. ¹³

90 In my judgment and having regard to my consideration of the two conceptual issues above, acquired technical distinctiveness should not be considered at the marks-similarity inquiry. I say this for three reasons based on precedent, principle, and policy.

91 First, I am of the view that precedent compels me to decide (or is at least ambiguous such that I could decide) that acquired technical distinctiveness should not be considered at the marks-similarity inquiry.

92 I turn first to examine the case law at the Registry level, which I consider to be split and therefore of no direct assistance as a matter of precedent. In this regard, I am indebted to the excellent summary of the prevailing case law on this issue by PAR Ong in GCIH. In his careful analysis of the case law, PAR Ong considered that previous cases at the Registry level could be divided into two broad categories.

93 The first category comprises cases which have applied what PAR Ong termed the “contextual” approach (see GCIH at [108]). This was apparently first formulated by the IP Adjudicator in Monster Energy Company v NBA Properties, Inc. [2018] SGIPOS 16 (“NBA”). The essence of the contextual approach is set out in NBA at [46], as follows:

Thus, under this contextual approach, parties should be permitted to adduce evidence of surrounding circumstances that are relevant towards establishing the general knowledge possessed by the average consumer. To do otherwise would be to adopt a “highly contrived, and artificially blinkered, approach towards mark similarity assessment that is incompatible with the commercial realities within which the trade mark regime operates”.

94 In contrast, the second approach from the cases at the Registry level, which PAR Ong has termed the “mechanical approach”, is that it would be inconsistent with Staywell to take into account contextual evidence, or any evidence of use, at the marks-similarity inquiry (GCIH at [111]). This approach is well-represented by the decision of the IP Adjudicator David Llewelyn in Valentino S.p.A. v Matsuda & Co [2020] SGIPOS 8 (“Matsuda”), where he had said at footnote 1:

The mechanical approach is characterised by such decisions taking the position that it would be inconsistent with Staywell to take into account evidence of use at the marks-similarity stage (to guard against adducing irrelevant evidence), and that such evidence may be adduced only at the likelihood of confusion stage. I shall revisit some of these decisions later. But for present purposes, the decisions at the Registry level are divided between whether acquired technical distinctiveness can be considered or not.

95 I move to consider the decisions at the High Court level. It appears that while there have been decisions touching upon the issue, there has been no clear answer derived as well.

96 In Polo/Lauren Co, the High Court expressed some reservations on the propriety of considering acquired technical distinctiveness at the marks-similarity stage in general based on the Court of Appeal’s comments in Staywell. The court was also unsure whether earlier marks which were already inherently technically distinctive can become more distinctive through use and whether this “enhanced” acquired technical distinctiveness can be taken into consideration at the marks-similarity stage (at [28]):

97 There has also been at least one High Court decision which considered acquired technical distinctiveness at the marks-similarity inquiry. This is the decision of Digi International. There, the court found that the opponent’s trade mark had a low degree of inherent technical distinctiveness (at [91]–[102]), and then went on to evaluate whether the trade mark had acquired technical distinctiveness through use. The court considered the opponent’s evidence of use and applied a set of factors taken from Hai Tong (at [33]) to determine whether a mark had acquired technical distinctiveness through use (at [103]–[119]). The court found that the competing marks were similar and made the observation that (at [125]): “It also bears emphasising that the Respondent’s Mark (Singapore) [ie, the Opponent’s mark] has inherent and acquired distinctiveness. All the more, the differences in respect of the triangle devices are not so substantial as to distinguish the Application Mark visually” [emphasis added]. Whilst not stated explicitly, in the course of the decision, it appears that the court did not consider itself to have contravened the prohibition in Staywell in considering external matters by evaluating acquired technical distinctiveness within the marks-similarity inquiry.

98 However, as Professor Ng-Loy notes, the issue of whether acquired technical distinctiveness can be so considered was not actually argued or properly raised before the court. Instead, the parties were content to proceed on the basis that this was a relevant consideration (see Ng-Loy Wee Loon at p 386, footnote 159). Accordingly, I do not consider that there has been a High Court authority which has taken a definitive view on the subject, and in any event, these are not binding on me.

99 I turn now to the Court of Appeal decisions. In this regard, the IP Adjudicator in Combe International Ltd. v Dr August Wolff GmbH & Co. KG Arzneimittel [2021] SGIPOS 10 (“Dr August Wolff (IPOS)”) opined that there was Court of Appeal authority which compelled him to decide that acquired technical distinctiveness should be considered at the marks-similarity inquiry. With respect to the IP Adjudicator, I disagree with this conclusion upon a careful examination of the relevant authorities. For context, I set out the relevant portion of his decision (at [30]):

The IP Adjudicator was cautious in holding that the earlier rulings of the Court of Appeal were no longer good law because they appeared to contradict more recent pronouncements in Staywell, unless the position is explicitly clear.

100 It would be fruitful to look at the Court of Appeal cases both before and after the seminal decision in Staywell to determine if acquired technical distinctiveness was meant to be considered at the marks-similarity stage. A close examination of these cases would reveal that while there may have been rather strong obiter comments or suggestions on the issue, none of these statements actually formed the ratio decidendi of the case on the basis that the controversial issue of acquired technical distinctiveness did not have to be decided in the end.

101 I begin with the case of Sarika, which was decided before Staywell. The Court of Appeal appeared to endorse the view that the analysis of acquired technical distinctiveness should be considered at the marks-similarity stage. Under the heading of “General approach to the similarity analysis”, the Court of Appeal stated that the distinctiveness of a trade mark is an important factor to bear in mind and that the “question of whether a sign and a mark are similar will often be dependant [sic] on the inherent or acquired distinctiveness of the trade mark” (at [19]) [emphasis added]. This statement was made as part of the court’s attempt to set out the relevant principles to clarify the general approach to be adopted in the marks-similarity analysis (at [16]), and does indeed suggest that acquired technical distinctiveness should be considered in the marks-similarity stage. However, the Court of Appeal ultimately did not consider the acquired technical distinctiveness of the mark. It was unnecessary to do so, considering that the court had found that the prior registered mark had a “considerable degree” of inherent distinctiveness (at [35]–[36]).

102 Sarika was then endorsed in the latter case of Hai Tong, which was also decided before Staywell. Within the analysis of marks-similarity, the Court of Appeal cited Sarika and affirmed that distinctiveness was to be considered within the assessment of marks-similarity and that distinctiveness encompasses both inherent and acquired technical distinctiveness (at [31]–[33]). The Court of Appeal in Hai Tong (at [33]) then went one step further to set out a non-exhaustive list of factors to guide the assessment of how acquired technical distinctiveness could be established (stating factors such as the market share held by the registered mark, the nature of its use, the amount invested in promoting the mark, etc). But nevertheless, and once again, having set out these factors, the court did not ultimately have to consider whether the registered trade mark had acquired technical distinctiveness as the trade mark garnered sufficient inherent technical distinctiveness (at [35]): “The juxtaposed words were therefore not devoid of distinctive character; and in that light, it really did not fall for consideration whether these words had also or separately acquired a measure of distinctiveness through use” [emphasis added].

103 I end the survey of authorities with what appears to be the only Court of Appeal authority post-dating the case of Staywell discussing the issue in some detail, and that is the decision in Caesarstone (CA). The PAR at first instance had considered both the inherent and acquired technical distinctiveness of the registered mark and held that it had a medium level of distinctiveness, and consequently enjoyed only a “normal threshold” before a competing mark would be considered similar to it (see Ceramiche Caesar S.P.A. v Caesarstone Sdot Yam Ltd [2015] SGIPOS 9 at [42]–[43]). That finding was not disturbed by the High Court (see Caesarstone (HC) at [40]). On appeal, the Court of Appeal had the opportunity to consider the PAR’s findings in relation to acquired technical distinctiveness at the marks-similarity stage but ultimately did not because it found that nothing in the appeal turned on that issue (at [28]–[31]):

From the above passage, we can see that the issue of acquired technical distinctiveness was not considered by the Court of Appeal as: (a) the High Court below did not make any express finding in relation to acquired technical distinctiveness; and (b) the appellant in that case did not challenge the High Court’s findings on distinctiveness (but only contested whether a higher threshold of protection would apply).

104 Taking a step back, what can be observed is that whilst the Court of Appeal authorities pre-dating Staywell seem to suggest, by way of obiter, that acquired technical distinctiveness could be considered in the marks-similarity stage, the only decision post-dating Staywell was undecided on the issue as: (a) there was no explicit endorsement that acquired technical distinctiveness could be considered at the marks-similarity stage (unlike Sarika and Hai Tong); and (b) the Court of Appeal did not actually have to decide the issue in the end and undertake the exercise of considering the evidence of alleged acquired technical distinctiveness at the marks-similarity stage (similar to Sarika and Hai Tong). For completeness, even in the case of Staywell itself, the Court of Appeal did not consider the acquired technical distinctiveness of the mark in question and had only affirmed the High Court’s finding that the mark enjoyed a substantial degree of technical distinctiveness (at [31]).

105 While it is true that the Court of Appeal had considered acquired technical distinctiveness at the marks-similarity inquiry before Staywell, it bears remembering that Staywell is the determinative case, being decided after those earlier cases, and that further, the only Court of Appeal case arising after Staywell is also ambiguous. In this connection, the learned PAR Ong in Clarins Fragrance Group f.k.a. Thierry Mugler Parfums S.A.S v BenQ Materials Corp. [2018] SGIPOS 2 observed that in both Staywell and Caesarstone (the two most recent apex court decisions on the issue at the time), there was no consideration of any evidence of acquired technical distinctiveness at the marks-similarity stage, and whether intentional or not, this actually conduced towards conceptual clarity (at [25]):

106 Given the inconclusiveness of the other apex court decisions, we must therefore turn to Staywell itself. In this regard, it is also true that Staywell does not explicitly say that acquired technical distinctiveness should not be considered at the marks-similarity inquiry. But the important parts of judgments, as with many things in life, sometimes lie within the parts when the court says nothing at all. At least, not directly. This much is clear from the tenor behind the important paragraph of Staywell (at [20]), where the Court of Appeal held that “the assessment of marks similarity is mark-for-mark without consideration of any external matter”. The court then defines acquired technical distinctiveness as words that “acquire the capacity to act as a badge of origin through long-standing or widespread use” (at [24]). The implication of this is that the Court of Appeal in Staywell likely regarded the inquiry of acquired technical distinctiveness to include considerations of “external matter”, since this would typically require the court to consider external matters such as: the market share held by the registered mark, the amount invested in promoting the mark, the proportion of the relevant class of individuals who identified goods sold under the mark as emanating from a particular source, etc (see Hai Tong at [33]). These would go beyond the mark-for-mark comparison at the marks-similarity stage and would involve considering evidence external to the trade marks.

107 Accordingly, I respectfully disagree with the IP Adjudicator in Dr August Wolff (IPOS) that the older decisions of Hai Tong (and others), in so far as they allowed the consideration of acquired technical distinctiveness at the marks-similarity inquiry, live on beyond the important pronouncements of Staywell (see [99] above). In my respectful view, that is a very constrained reading of Staywell that does not give effect to its broader intended effect, which is equally binding on lower courts as its more direct pronouncements. I therefore regard that I am bound by Staywell to hold that acquired technical distinctiveness should not be allowed at the marks-similarity inquiry. This, as the learned IP Adjudicator Llewelyn noted in Matsuda, is the only way to remain faithful to the Court of Appeal’s holding in Staywell. IP Adjudicator Llewelyn’s views in Matsuda have been endorsed in GCIH (at [112]) and separately, other cases also opine that whether a mark has acquired technical distinctiveness is not a relevant factor in assessing marks-similarity having reference to the decision of Staywell (see Damiani International BV v Dhamani Jewels DMCC [2020] SGIPOS 11 at [24]; cf Tata Sons Private Limited v Tata’s Natural Alchemy, LLC [2022] SGIPOS 15 (“Tata Sons”) at [53]–[55]).

108 For similar reasons, I also respectfully disagree with the learned view expressed in Vaerhn and Yew (at 628) that, when paragraphs 23, 24 and 25 of Staywell are read together, it is clear that the Court of Appeal in Staywell held that acquired technical distinctiveness should be considered at the marks-similarity stage of the step-by-step approach. In my respectful view, drawing a conclusion from these two isolated paragraphs is akin to reading the Court of Appeal’s judgment in Staywell as though it is a statute. Judgments are not meant to be read like statutory instruments. There will be points of inflexion within judgments that are not as present in statutes (which are necessarily drafted more technically). Accordingly, in reading judgments, it is always important to bear in mind the overall picture. I do not think that isolating two paragraphs in Staywell and attempting to read them like a statute assists in figuring out what the big picture from Staywell is.

109 The Court of Appeal’s recognition in Staywell that technical distinctiveness comprises both inherent and acquired technical distinctiveness (at [24]) was meant to be descriptive and not prescriptive. In pronouncing that “technical distinctiveness is an integral factor in the marks similarity inquiry” (at [25]), the Court of Appeal likely did not intend to contradict its own statement that the mark-for-mark assessment must be conducted without consideration of any external material (at [20]). Instead, it is likely that the Court of Appeal intended that only inherent technical distinctiveness is integral to the marks-similarity inquiry. That could be one way to reconcile the seemingly contradictory statements made by the Court of Appeal in Staywell. Accordingly, as a matter of precedent based on Staywell, I am bound by Staywell to hold that acquired technical distinctiveness should not be considered at the marks-similarity stage (but should be left for the likelihood of confusion inquiry).

110 Quite apart from precedent, I am of the view that not considering acquired technical distinctiveness at the marks-similarity inquiry is also consistent with principle.

111 I deal first with the argument in favour of considering acquired technical distinctiveness at the marks-similarity inquiry stage. The argument made by the IP Adjudicator in NBA appears to be that, if the relevant contextual background is not considered at the marks-similarity inquiry, it would result in the adoption of a “highly contrived, and artificially blinkered, approach towards marks-similarity assessment that is incompatible with the commercial realities” (see NBA at [46]). As such, the IP Adjudicator advocated for the admission of evidence in relation to the knowledge of the average consumer, as well as how words are pronounced or heard by the average consumer and the connotations that are conjured up in the mind of the average consumer perceiving the mark for the purposes of assessing visual, aural, and conceptual similarity (at [48]). On the specific facts of that case, he would have allowed the average Singaporean consumer’s general knowledge of how sports teams are conventionally named (that is, to preface and pair a team’s informal name with its city of origin) (at [47]). The admission of such evidence could potentially lead to the consideration of acquired technical distinctiveness at the marks-similarity inquiry.

112 With respect to the IP Adjudicator in NBA, his advocated approach risks blurring the distinction between truly general knowledge and knowledge that is specific to the marks in question. In relation to general knowledge, it goes without saying that context is needed for any interpretative exercise. For example, to take an extreme example, to recognise a sign of the Moon requires the contextual understanding that the Moon is Earth’s natural satellite that orbits our planet once every (about) 27 days. Or that the average consumer knows how to read English to understand the letters being used in a word mark. I do not think anyone is seriously disputing the relevance of that kind of general contextual evidence. Nor is there a specific need to adduce evidence of such general knowledge.

113 However, to extend that conclusion to the potential relevance of other more specific contextual evidence that paves the way for acquired technical distinctiveness to be considered at the marks-similarity inquiry is to invite travel down the proverbial slippery slope. To be fair, the IP Adjudicator in NBA did not state that acquired technical distinctiveness can be considered at the marks-similarity inquiry, although I do note he seems to have said so in the recent decision of Tata Sons at [58]–[59]. Indeed, in Tata Sons, the learned IP Adjudicator repeated the point (at [56]) that the “average consumer” in Singapore (which he refers to as a “legal fiction” given the difficulties of constructing such a character) needs to be fleshed out and raises questions such as whether he or she “has general knowledge based on what is taught in mainstream schools, what is available on the shelves of neighbourhood retail outlets, what is reported in local media outlets and advertised on local television channels or, in the digital era of the present, what is trending in cyberspace”. While I appreciate the learned IP Adjudicator’s point, in my respectful view, to ask these questions is really to descend into the minutiae. While it is not academically attractive, I think we need to accept that, in this imperfect world, the laws that are passed by the hands of humans will necessarily never be perfect. The “legal fiction” of the average consumer is merely a heuristic tool within the analysis. As such, if one begins to ask questions such as whether the average consumer knows what is “available on the shelves of neighbourhood retail outlets”, then why not ask if he or she knows what is available on the virtual shelves of Amazon? The inquiry would be endless. It would also be unnecessary. Laws would not function in the real world if one were needlessly fixated on the minutiae.

114 Indeed, as PAR Ong notes in GCIH, the broader point is that one must be very careful when considering factors which may impact on the actual knowledge of average consumers. This is especially the case where such knowledge arises because of a trader’s market activities, such as sales or advertising in relation to the mark. This would be, to adopt the parlance used in Staywell, to do with the effect of such resemblance on average consumers, rather than the issue of resemblance as between the competing marks (see GCIH at footnote 31; Staywell at [20]). Although often framed as an “issue of resemblance”, acquired technical distinctiveness in reality impinges on the question of the “effect of such resemblance” (see GCIH at [110]):

As such, I do not think that the contextual approach takes us very far in relation to the consideration of acquired technical distinctiveness at the marks-similarity inquiry.

115 I turn then to the mechanical approach, which I agree with. I have already alluded to how I think Staywell prohibits the consideration of external matters at the marks-similarity inquiry, which would include acquired technical distinctiveness. It is crucial that the broader rationale behind the position taken in Staywell is borne in mind – to distinguish clearly between the issue of resemblance between the competing marks and the question of the effect of such resemblance (at [20]). When this rationale is borne in mind, it becomes clear that extrinsic evidence that pertains to the effect of the resemblance between the marks, such as acquired technical distinctiveness, is not allowed. However, I agree with PAR Ong that this should not mean that all and any contextual evidence is inadmissible. Like him, I think a suitable “carve out” can indeed be drawn for general contextual evidence that is independent of the trader’s public activities (which would usually not be related to the effect of the resemblance between the marks) that can be admitted at the marks-similarity inquiry. This would still be consistent with not considering acquired technical distinctiveness in the marks-similarity inquiry, since such general contextual evidence would likely not amount to evidence in support of acquired technical distinctiveness.

116 I also agree with part of the applicant’s arguments that not considering acquired technical distinctiveness at the marks-similarity inquiry would fully comport with the position in Staywell that the question of the effect of resemblance between the marks on the perception of consumers should only be considered at the likelihood of confusion stage. Further, the same evidence of use normally adduced to show acquired technical distinctiveness may then be considered alongside evidence of reputation as well – such that the assessment of acquired technical distinctiveness and reputation are considered together in the confusion inquiry. Indeed, as noted by academic commentary (see Vaerhn and Yew at 644), the factual bases and evidence that would be proffered to show that a trade mark has acquired technical distinctiveness (such as evidence of sales, advertising campaigns, and so on) have also been used to prove that a trade mark has a reputation (see, for example, the Court of Appeal decision in McDonald’s Corp v Future Enterprises Pte Ltd [2005] 1 SLR(R) 177 (“McDonald’s Corp”) at [52]). There are thus a number of decisions which considers the same evidence for both (see also, Caesarstone (CA) at [21] and [53]; Rolex S.A. v FMTM Distribution Ltd [2020] SGIPOS 6 at [28]–[36] and [71]).

117 Therefore, by considering acquired technical distinctiveness at the likelihood of confusion stage (as opposed to the marks-similarity stage), there is the benefit of avoiding the evaluation of the same kind of evidence twice over at different stages – and this would be the more economical and analytically coherent way forward.

118 Finally, I agree with the IP Adjudicator’s view in Matsuda (at footnote 1) that not considering acquired technical distinctiveness at the marks-similarity inquiry would stop the adducing of much irrelevant evidence with the attendant cost consequences on all parties. While I recognise that the IP Adjudicator in Matsuda would prohibit the consideration of surrounding circumstances that are relevant towards establishing the general knowledge possessed by the average consumer, and that I would allow for such evidence, I must emphasise that, in line with the pronouncement in Staywell at [20], such evidence must go only towards understanding the resemblance of the competing marks and not the effect of such resemblance. Therefore, in most cases, such background knowledge is likely to be commonly assumed to exist between parties with no adjudication on their admissibility required.

119 This has been a long section. I therefore summarise my conclusions in relation to the concept of “distinctiveness” in the marks-similarity inquiry.

(a) First, I would suggest the consistent use of the following expressions when discussing the concept of distinctiveness at the marks-similarity inquiry: (a) inherent technical distinctiveness; (b) acquired technical distinctiveness; and (c) non-technical distinctiveness. The consistent use of these expressions would aid in the formulation of coherent arguments before decision-makers, as well as enhance the comprehensibility of our intellectual property law jurisprudence.

(b) Second, I would suggest a faithful return to Staywell and not consider “distinctiveness” as a “threshold” enquiry (even for reasons of convenience or ease of analysis), as this is in reality a separate step to the analysis that is not permitted by Staywell. Treating distinctiveness as integrated within the step-by-step approach would ensure that it is properly applied in the right context.

(c) Third, I hold that acquired technical distinctiveness should not be considered at the marks-similarity inquiry based on reasons of precedent, principle, and policy. The issue of acquired technical distinctiveness should be considered at the likelihood of confusion stage of the inquiry to preserve conceptual clarity.

120 In ending this discussion, I should say that I have ventured these views not because I think I have the correct answers. Far from it, although the doctrine of stare decisis would imbue that characteristic to my views with respect to lower courts and tribunals. Instead, I have done so because I perceived, with respect, genuine uncertainty in the various legal issues at both the Registry and High Court levels. At the very least, if my views can result in an opportunity for the Court of Appeal to clarify the issues, even if entirely contrary to my views, then this would have been all worthwhile.

121 Returning to the present case, the two immediate consequences of my conclusions above are that, first, I do not consider “distinctiveness” as a threshold enquiry, and second, I do not consider acquired technical distinctiveness at the marks-similarity inquiry at all. I turn now to the marks-similarity inquiry proper, considering first the visual similarity of the competing marks.

122 As I mentioned earlier, it is undisputed that the competing marks should not be compared side-by-side as it is assumed that the average consumer has an “imperfect recollection” and views each mark separately in time and space (at [27] above). It is also not disputed that registration in black and white confers protection on the mark in all colours (see Fox Head, Inc. v Fox Street Wear Pte Ltd [2018] SGIPOS 8 at [56]). As such, as the PAR did, I reproduce the Registered Mark in a similar shade of yellow as the Application Mark side-by-side, purely, for ease of comparison:

123 The PAR found that the competing marks are visually similar because: (a) both marks depict a bird in flight; (b) both depict the side profile of a bird; (c) both appear to depict a relatively small bird; (d) neither mark depicts features such as eyes; and (e) neither mark corresponds to an identifiable species of bird (see the GD at [79]–[80]). In doing so, the PAR applied the “useful guidelines” suggested by the respondent, which are “the general shape, movement, features and composition of the animals depicted in the competing marks”. The PAR reasoned that this assisted in “reaching a more objective determination as to whether the marks are more similar or dissimilar overall” (see the GD at [78]).

124 I consider first the inherent technical distinctiveness of the Registered Mark as considered from a visual perspective. In this regard, I accept the respondent’s argument that the Registered Mark enjoys considerable inherent technical distinctiveness.

125 First, the Registered Mark, which depicts a two-dimensional profile of a small bird in flight, is arbitrary and meaningless in relation to the goods and services claimed under the mark. The use of a mark that is either meaningless or has no discernible correlation to the product or service in question or its key elements generally serves no purpose other than as a mark or badge of the origin or source of that product or service (see the Court of Appeal decision of The Singapore Professional Golfers’ Association v Chen Eng Waye and others [2013] 2 SLR 495 at [29]), and it will thus be regarded as distinctive (see Hai Tong at [29]) or, in present parlance, inherently technically distinctive. The present circumstances are similar to that in Abercrombie & Fitch Europe SAGL v MMC International Services Pte Ltd [2016] SGIPOS 6 (“Abercrombie”). In that case, the PAR found that a standalone moose logo enjoyed a high level of inherent technical distinctiveness as it was “meaningless in relation to the goods and services claimed” (see Abercrombie at [99]).

126 Second, and accordingly, I reject the applicant’s argument that the respondent’s mark is not completely arbitrary in relation to the services claimed. The applicant argues that the bird mark alludes to the fact that birds make short chirping noises akin to the short messaging service provided by the respondent. The applicant thus submits that as the Registered Mark was partially descriptive of the goods, and lowers its inherent technical distinctiveness. In my judgment, this argument is too far-fetched, and many leaps must be made to arrive at the applicant’s assertion. I do not find that the Registered Mark is descriptive of the goods and services.

127 Third, given my discussion above, I reject the applicant’s argument that the average general consumer today is bombarded with simplified, abstract, and monochromatic brand logos or icons, and that such logos or icons will often depict animals (including birds). ¹⁴ The upshot of this appears to be that the Registered Mark should not enjoy a considerable degree of inherent technical distinctiveness since the general consumer will scrutinise various marks more carefully. However, this, to my mind, is impermissible evidence about the effect of the Registered Mark.

128 As for non-technical distinctiveness from a visual perspective, I agree with the applicant that there should not be a high level of distinctiveness in this regard. Indeed, I disagree with the respondent that the non-technical distinctiveness of its mark is at a high level as it provides for a fanciful and inventive depiction of the bird consisting of a series of overlapping circles whereby every part of the bird is shaped by arcs of different sizes. ¹⁵ In my view, not much weight can be given to the respondent’s arguments. After all, the respondent has also asserted in evidence that its registered bird mark is meant to be a “simplified Twitter bird” with “simple geometry”. ¹⁶

129 Accordingly, taken in the round, I find that the Registered Mark has a normal level of distinctiveness from a visual perspective in all aspects of the concept, and will correspondingly enjoy only a normal threshold before a competing sign will be considered to be dissimilar to it (see Caesarstone (CA) at [30]). As such, I find that distinctiveness, in both its technical and non-technical senses, is at best a neutral factor in the present case.

130 Having considered the distinctiveness (in both its technical and non-technical aspects) of the Registered Mark from a visual perspective, the next question I need to decide is whether the PAR erred in adopting the respondent’s guidelines to assess animal marks (see [123] above). The applicant argues that the PAR’s use of these guidelines finds little basis in law, and that it is hard to see why there should be a special carve-out for animal marks to subject them to a special set of considerations. ¹⁷

131 Against this, the respondent contends that its guidelines were based on detailing principles from a survey of the various animal mark cases and are not meant to be a special carve-out. ¹⁸ Instead, its proposed manner of analysis is merely a specific manifestation of the test laid down in Hai Tong, which allows the court to compare the animal marks more systematically.

132 Having considered the parties’ arguments, I reject the applicant’s objection against the use of the respondent’s guidelines. Like the PAR, I found that these guidelines provide me with a helpful and objective basis to assess the competing marks. By themselves, the guidelines are result-neutral and do not slant in favour of a particular type of mark until they are applied to assess two competing marks. I also do not think that the use of these guidelines will lead to a slippery slope whereby special carve-outs are had for different types of marks. In my view, that is the very essence of applying general principles to particular circumstances, which will inevitably lead then to a curated set of factors for each circumstance. I see nothing wrong with that. I therefore adopt the respondent’s framework of assessing the competing marks, focusing on “the general shape, movement, features and composition of the animals depicted in the competing marks”. ¹⁹

133 The applicant focuses its case on what it says are the dissimilarities between the general shape and composition of the competing marks. ²⁰ Its arguments can be divided into those that are premised on a general examination of the marks, and those that are based on case law.

134 I turn first to address the applicant’s arguments founded on a general examination of the competing marks. In this regard, I agree with the respondent that the competing marks are visually similar.

135 First, as to the general shape and composition of the competing marks, the applicant points to the distinctive character of the Application Mark being in the slender and angular three-dimensional depiction of a hummingbird, with its long, pointed beak and its thin, V-shaped body and sharply pointed wings depicting the applicant’s initials (“V V”). I reject these supposed differences between the marks. In the first place, it is not clear to me that the Application Mark is angular or has a V-shaped body. Also, it is not at all clear to me how the Application Mark could be said to have more sharply pointed wings, or a long beak as compared to the Registered Mark. Instead, I agree with the respondent that the general shape of both parties’ marks is of a two-dimensional, side profile of a small bird with curvilinear features, swept back wings opening out and up behind the bird’s head, a pointed tail curved outward, as well as a pointed beak.

136 Second, the applicant also argues that the competing marks are of markedly different orientations. ²¹ The applicant submits that this is clear if a line is drawn across the depicted bird’s body as shown below:

I reject the applicant’s submissions in this regard. It is unclear to me what counts as a line drawn “across” the depicted bird’s body. To elaborate, the line across the Application Mark is from the back of the bird’s head to where its feet might be. In contrast, the line across the Registered Mark is from where the eyes of the bird would be to its tail. The point I make is that the orientation can change depending on how one draws the line.

137 Third and more broadly, the applicant’s focus on pedantic stylistic dissimilarities is based on a side-by-side comparison of the marks, which is precisely what should not be done in the marks-similarity assessment (see [27] above). I accept the respondent’s point that this is not how the average consumer views trade marks. ²² In real life, consumers simply do not analyse and remember the various minute details of the marks. As such, while it is true that the differences listed by the applicant may be evident from a detailed side-by-side comparison, these differences are, when viewed holistically, ultimately trivial. They are not likely to feature in the average consumer’s imperfection recollection.

138 I now consider the applicant’s arguments founded on case law, which are elaborations of its submissions on a general examination of the competing marks. Before I go through the cases, I should emphasise that these are, at best, guides as to the correct resolution of an instant case. This is because no two cases are completely alike, and each case is ultimately resolved on its own precise facts. Thus, I do not regard it as necessarily decisive for either party to raise a case with ostensibly similar marks that the court and tribunal found to be similar or dissimilar (as the case may be). What ultimately matters are the facts of the instant case, and a judgment made on those facts.

139 First, the applicant cites the UKIPO’s decision in Trade Mark Inter Partes Decision O368/21 (“Worldremit”), which is an opposition based on s 5(2)(b) of the English Trade Marks Act 1994. The case involved the following marks:

140 The Hearing Officer concluded that the competing marks were visually dissimilar despite the competing marks being made up of a polygon and a “W” element. The Hearing Officer concluded that, in relation to the opponent’s first mark, the only point of visual similarity between the marks is the presence of an ordinary polygon shape with the remaining elements of the marks being different. Thus, although the applicant’s mark uses an octagon whereas the opponent’s marks use a hexagon, this was insufficient to warrant a finding that the marks are visually similar. Further, there is no visual similarity between the respective depictions of “W” as they are stylistically very different.

141 The applicant relies on the principle of law emanating from Worldremit that “just because two marks may share representations of the same thing, it does not automatically result in a finding of visual similarity between them” (at [49]). I do not think anyone would disagree with this principle of law. Indeed, the respondent does not go so far as to say that just because the applicant’s mark is a bird, therefore it should be regarded as visually similar to the Registered Mark, which is also a bird. As I understand it, the respondent’s argument is more nuanced but does not rise to the level of an overfocus on the pedantic details.

142 Second, the applicant cites S Tous, S L v Ng Wee Ping [2010] SGIPOS 6 (“Ng Wee Ping”). In that case, the PAR held that the application mark was visually dissimilar to the opponent’s mark notwithstanding that both “consist of the device of a bear, and specifically, a teddy bear” (at [26]). The PAR further observed that the opponent’s earlier mark had facial features (eyes and nose) and chubby, lower limbs which give the overall impression of a “cute, bloated-bodied creature” as compared to the application mark, which appears “flatter and two-dimensional” (at [26]–[28]). The applicant therefore submits that Ng Wee Ping fortifies its argument that the competing marks in the present case give rise to different overall impressions due to their distinctive characters, the fact that they depict different types of birds (ie, a hummingbird versus a mountain bluebird), and the fact that they are in different orientations.

143 I do not think that Ng Wee Ping assists the applicant. I agree that the competing marks in that case differ in their shapes, movements, and compositions. But this does not mean that we should reach the same outcome in the present case. The point in Ng Wee Ping is that if the competing marks leave an average consumer with markedly different impressions, only then would a finding of visual dissimilarity be warranted. But, it cannot stand for a more precise proposition than that. Moreover, I do not think that an average consumer would be able to identify the precise species of birds in the competing marks. Given that I have already found the competing marks to be visually similar from a general examination of them, Ng Wee Ping does not assist the applicant.

144 Third, the applicant cites Morinaga & Co., Ltd. v Crown Confectionery, Co., Ltd [2008] SGIPOS 12 (“Morinaga & Co”) and points to the fact that the two marks were found to be different because the Registered Mark only depicts a two-dimensional side profile of a bird, while the Application Mark employs a three-dimensional effect arising from gradations in shading. ²³ In that case, a finding was made that the application mark was visually dissimilar to as the learned PAR found (at [147]) that the former is “two-dimensional” whereas the latter “almost stands out of the page with its three-dimensional shadow effect”.

145 I also do not think that Morinaga & Co assists the applicant. I cannot see how the Application Mark is three-dimensional. It appears to lay on a flat surface and the features do not pop out of the page, unlike what was cited in Morinaga & Co. Furthermore, to find otherwise would be encroaching onto nit-picking territory as these seem to be pedantic differences which the average consumer will overlook.

146 In contrast to the cases cited by the applicant, I find the cases cited by the respondent to be helpful in the present case. It suffices to just refer to one of those cases. In Arctic Cat Inc v European Union Intellectual Property Office (Case T-113/16) (“Arctic Cat”), the General Court of the European Union found that the application

mark was similar to the prior registrations for . The court had found that the overall impression of the marks was dominated by their general shape, movement, features and composition of the animal devices. The court had characterised this as “the black silhouette of a member of the cat family represented in profile, conveying an impression of movement, characterised by the position of the fore- and hind legs extending from the central part of the body and not supporting it”. The court regarded this as being immediately noticeable and easily remembered by the average consumer (at [37]).

147 In my view, cases such as Arctic Cat support my finding that the competing marks are visually similar in terms of their general shapes, when one adopts a higher level of abstraction. There is similarity between the overall impression of the competing marks’ general shapes in the present case. More broadly, I do not agree with the applicant’s attempt to introduce precise differences in the elements to the marks-similarity assessment. This is not only inconsistent with the case law, but it also does not reflect the common-sense depiction of how real-life consumers form their impressions of two marks.

148 As to the general movement of both birds, both parties accept that both are similarly that of one in flight. However, the applicant says that while both birds are in flight, the movements implied by the position of their bodies, wings and position of their heads are different. To substantiate this point, the applicant relies on the drawing of a line through the body of the birds, which it terms a “body line” (see also [136] above). ²⁴

149 According to the applicant, the body line of the bird in the Registered Mark is at a diagonal, which suggests that the bird has just taken flight. However, the body line of the bird that is depicted in the Application Mark is perpendicular to the ground. The applicant suggests that this depicts “hovering flight”, which leaves the consumer with the impression that the bird is remaining in the same position. ²⁵

150 The applicant also attempts to distinguish the cases of PUMA SE v Sinda Poland Corporation Sp. z o.o. (Case R 1304/2016-1), Arctic Cat, and Puma SE v Slazengers Limited (Opposition No B 2 531 351) (“Slazengers”). By drawing the body line through the respective felines in the cases, the applicant says that the body lines are roughly mirrored images of each other: ²⁶

The applicant therefore argues that this analysis may explain why the respective competing marks were found to be visually similar as the overall impression was that of a leaping feline. This is not the case here as between the impressions of movements of the birds in the competing marks.

151 I reject the applicant’s argument for several reasons. In the first place, as I alluded to above (at [136]), it is unclear to me what counts as a line drawn “across” the depicted bird’s body. For instance, in relation to the left-most mark in the Slazengers row above, the feline’s front legs are pointed down. One could equally draw the body line slanting downwards, to give the impression of the feline jumping down. The point is that there is no one definite way to draw the body line. Indeed, the applicant candidly accepts in oral submissions that the body line for the applicant’s mark could also be drawn with a diagonal line from the eyes of the bird to its tail, thus reinforcing my point that the exercise in line-drawing might be quite an arbitrary one and unhelpful to the inquiry at hand.

152 Second, the applicant’s rather involved analysis of drawing body lines and measuring the angle of these lines exposes the inherent flaw of this approach. It is much too involved. It simply does not align with how an average consumer with imperfect recollection would assess the competing marks. Rather than think that the birds have “just taken flight” or are “in stationary hovering flight”, ²⁷ the average consumer is simply going to perceive two birds which are in flight. I accept, of course, that a bird in flight may well be too broad a protection, but here, it must be recalled that the marks-similarity inquiry does not just involve the movement depicted in the marks.

153 As for the general features of the competing marks, the applicant has highlighted ten differences between them, in that the Application Mark has: (a) a longer beak; (b) a closed beak; (c) an eye; (d) a neck; (e) not a rounded head; (f) a smaller head; (g)  not a rounded belly; (h) a thinner V-shaped body; (i) more sharply pointed wings; and (j) different wings. I should point out that the applicant has raised the point about the birds having different wings only in the present appeal. I reproduce the competing marks again for ease of exposition:

154 More specifically, the applicant says that the Application Mark depicts the bird as having three wings, whereas the Registered Mark only depicts a single wing but with three distinct feathers. The applicant also says that the wings of the Application Mark follow a concave line, whereas the wing of the Registered Mark follows a convex line, which the applicant has helpfully added to the marks in the images above.

155 I reject the applicant’s arguments about the supposed differences in the general features of the competing marks. In the first place, I cannot understand how a bird, unless it is of the mutated variety, can have three wings. I think that the applicant is truly splitting feathers with this argument, which goes against the general flight of the correct approach to be taken in assessing the similarity between marks. Again, I agree with the respondent’s point that this is not meant to be a “detailed spot-the-difference” comparison, as these differences identified by the applicant are ultimately trivial and unlikely to feature in the average consumer’s overall impression of the marks. ²⁸ Thus, while one can always point out minute differences between the two signs, these broad coincidences between the competing marks’ general shapes, movement, features, and compositions are sufficient to establish visual similarity.

156 For all these reasons, I conclude that the competing marks have a fair degree of visual similarity.

157 The PAR held that aural similarity is not relevant in the present case for four reasons. These reasons are: (a) device-only marks by definition contain no aural component; (b) almost all device marks will be used in conjunction with a word mark; (c) where a device mark appears on its own, it is usually because the device mark is well-known to the public; and (d) it is hard to imagine a situation where a device mark would be referred to using a verbal description (see the GD at [88]–[93]).

158 Indeed, in Polo/Lauren Co, the court had also expressed reservations and stated that (at [22]): “[t]o find aural similarity where no aural component exists seems to allow for visual or conceptual similarity to be accounted for within the assessment of aural similarity”. Given that the competing marks in question here are both device-only marks with no aural component, there are no verbal or aural elements which may be subject to such an analysis.

159 Since the parties have not questioned the PAR’s holding that aural similarity is not relevant in the present case (and referring to previous observations in Polo/Lauren Co), I will go directly to consider conceptual similarity and make no finding on aural similarity.

160 As the PAR noted (see the GD at [94]), conceptual similarity is directed at the ideas that lie behind or inform the marks or sign in question (see Hai Tong at [70]). These ideas must be evident in the look and feel of the mark, and not in something that is known only to the creator of the mark (see the High Court decision of The Audience Motivation Company Asia Pte Ltd v AMC Live Group China (S) Pte Ltd [2015] 3 SLR 321 at [43]).

161 The PAR found that the competing marks are conceptually similar (if not identical). This is because they each convey the concept of a “bird in flight” (see the GD at [101]). On appeal, the applicant made several conceptual arguments against the PAR’s decision.

(a) First, the applicant asserts that the PAR’s analysis was conducted at too high a level of abstraction. ²⁹

(b) Second, and relatedly, the applicant argues that the PAR’s finding that the competing marks are conceptually similar also comes close to claiming a right in this concept. This is because the concept of birds flying is hardly striking as it is a natural and common action for birds. As such, the applicant argues that it is more appropriate for the analysis to be of a less general nature. ³⁰

(c) Third, the applicant submits that, contrary to the respondent’s arguments below, conceptual similarity should not be given primacy and offset visual dissimilarity. ³¹ However, given that I have found that the competing marks are visually similar, it does not assist the applicant even if I were to agree that conceptual similarity cannot offset visual dissimilarity.

162 On appeal, the respondent supports the PAR’s primary finding that the competing marks are conceptually similar. It did not press its arguments below about the offsetting effect of conceptual similarity. However, the respondent does, understandably, reiterate its point that the PAR had conducted his analysis at the correct level. ³²

163 Considering the parties’ arguments, I will deal with the following conceptual issues before turning to the present facts. The first is the correct level of analysis to take in assessing conceptual similarity, and the second is the weight to be attributed to conceptual similarity when the competing devices depict highly recognisable and distinctive concepts.

164 I start with the correct level of analysis. The applicant argues that the PAR erred in applying an analysis that is too abstract in finding that the competing marks are conceptually similar because both depicted birds in flight. The applicant cites the 2015 EU decision of GSMA Ltd. v ZIH Corp. (Case R 1978/2014-5) (“GSMA”), where the marks in question were the application mark and the earlier mark. In rejecting the opponent’s argument that the competing marks are conceptually identical because both amount to “a mobile device emitting a signal”, the Board observed that “this captures commonality at too high a level of abstraction”. The Board further held that (at [34]):

165 The applicant therefore submits that to treat the competing marks in the present case as being conceptually identical based on them being a “bird in flight” is to “conduct the analysis at too high a level of abstraction”. This is because this is what birds naturally do, viz taking flight, and it is not an inventive graphic. In a related vein, the applicant also contends that the PAR’s finding that the competing marks are conceptually similar “comes close to claiming a right in this concept”. ³³

166 The respondent in turn says that defining the competing marks’ concepts with reference to their animal family, rather than their exact species, is entirely consistent with trade mark jurisprudence. ³⁴ Indeed, in Arctic Cat, the EU General Court rejected the applicant’s arguments that the competing marks and were conceptually dissimilar because the former is a “black panther” while the latter is a “fantasy animal”. The court rejected this analysis because the average consumer “is likely to identify a member of the cat family in each of the signs in the present case, without however being able to specify what species of cat it is” (see Arctic Cat at [45]).

167 Having considered the parties’ arguments, I agree with the respondent. In my view, the PAR did not conduct the analysis at a level of abstraction that was too high. He is perfectly entitled to consider the competing marks based on them depicting birds in flight. This is because, first, as the respondent submits, this is consistent with trade mark jurisprudence. In this regard, I accept that the cases cited by the respondent support an approach that defines the competing marks’ concepts with reference to their animal family rather than their exact species.

168 Second, I do not agree with the applicant that conducting the analysis in this manner would come close to claiming a right in this concept. This is because the marks-similarity inquiry does not comprise only of an assessment of conceptual similarity. As such, while I recognise the concern expressed by the Board in GSMA, I do not think that the concern should be overextended. Indeed, on the facts of GSMA, it is important that the Board had found that the competing marks there are relatively weak signs, in that “there is a significant degree of description in each of these marks in that they could be taken to indicate a particular functionality in the products they label, that is, the device they appear on are in some sense enabled to send and receive wireless signals” (see GSMA at [31]). It was in that context that the Board was concerned that the opponent’s argument in that case, to analyse the conceptual similarity as premised to both marks amounting to “a mobile device emitting a signal”, could come close to claiming a right in that concept. I do not think such a risk is present where the marks, such as in the present case, are arbitrary and meaningless in relation to the goods or services of interest.

169 I turn now to the weight attributable to conceptual similarity in the marks-similarity inquiry. In the proceedings below, the respondent had cited several cases from New Zealand, viz, Glenn Elliott v Heineken Asia Pacific Pte. Ltd [2014] NZIPOTM 18, Red Bull GmBH v Carabao Tawandang Company Limited [2005] NZIPOTM 25 (“Red Bull IPONZ”) and Carabao Tawandang Company Limited v Red Bull GmBH (CIV2005-485-1975) (“Red Bull HCNZ”) in support of the proposition that marks determined to be visually dissimilar may nonetheless be found to be similar overall due to their conceptual identity. The PAR did not analyse these cases in detail but was content to make the observations as I alluded to above.

170 In the present appeal, the applicant argues that where the competing marks are device marks, visual similarity will usually be the most important factor. In support of this argument, the applicant cited the High Court decision of MediaCorp at [32] as standing for this proposition. As such, the applicant says that conceptual similarity should have no or at best marginal relevance, and an assessment of conceptual similarity should lead to a neutral conclusion. ³⁵

171 The respondent refutes this by arguing that conceptual similarity is equally as important and is not subsidiary to visual similarity. The proposition espoused in MediaCorp was limited to the facts of that case involving an incomplete triangle or A-frame mark which is not as memorable and digestible as a plain and simple animal mark. The respondent says that where bird marks or most easily recognisable animal marks are concerned, these concepts are highly recognisable and distinctive in the minds of the consumers. Agreeing with the PAR’s views (see the GD at [104]), the respondent argues that such marks already embody a clear and recognisable concept. Consumers would inevitably remember the concept of a small bird in flight when recalling the marks’ visual representations, and that shared concept would dominate the overall impression of the marks and neutralise any specific visual differences. ³⁶

172 In my view, there is no immutable rule that conceptual similarity should be considered as a more important or less important factor in relation to visual and aural similarities. All similarities must be assessed separately, and a holistic judgment is to be made in the round at the end. To argue that visual similarity will usually be the most important factor, and thus will relegate conceptual similarity to merely marginal relevance, is not helpful in the overall analysis. In fact, not taking the view that some aspects of similarity are more important than others would stay truer to the principles laid down in Staywell at [20] (the relative weight of each aspect of similarity having regard to the goods should not be considered at the marks-similarity stage):

173 Accordingly, from the discussion above, I conclude that: (a) the PAR adopted the correct level of analysis in assessing the conceptual similarity of the competing marks as involving birds in flight; and (b) there is no immutable rule that conceptual similarity should be considered as a more important or less important factor in relation to visual and aural similarities. All three similarities must be assessed separately, and a judgment to be made in the round. Given my conclusions on the law, I find that the PAR is correct in concluding that the competing marks are conceptually similar, because they both convey the concept of a bird in flight.

174 However, for completeness, I also consider the applicant’s submission that a more specific degree of analysis should be undertaken in relation to conceptual similarity. In this regard, the applicant submits that the conceptual dissimilarity of the marks is that: (a) the Application Mark’s slender and angular depiction of a hummingbird evokes its swiftness and agility, which is more sleek, whereas the Registered Mark’s overall roundedness evokes a sense of playfulness and whimsy, which is more casual; (b) the Application Mark gives the impression of a bird in stationary hovering flight, evoking the idea of agility, whereas the Registered Mark gives the impression of a bird taking flight, which is the ultimate representation of freedom and wide-open possibility. ³⁷

175 The respondent, in oral submissions, challenged the characterisation of its bird mark as being whimsy and cited the source of the article referred to by the applicant, which stated in the relevant part: ³⁸

From the portion cited above, the respondent argues that the word “whimsy” is an inaccurate description of their mark by itself, since the quote read in context demonstrated that it was describing the evolution of the mark to reach the “right mix” of both whimsy and maturity. The other contention the respondent has was that the applicant’s description of their mark as evoking “swift[ness] and agil[ity]” was based on a singular Facebook post made in 2020, two years after the relevant date of assessment which is impermissible evidence in opposition. In any event, that Facebook post was created by the applicant itself, and does not reflect how the public would understand the applicant’s mark as representing swiftness and agility.

176 I agree with the respondent’s submissions. In my view, the adjectives raised by the applicant to contrast the competing marks are not necessarily mutually exclusive. A bird can be described as swift and playful, or agile and whimsical, or any other of such permutations. What stands in better opposition to the concept of swiftness is perhaps a bird which is inactive and stationary (perhaps depicting the concept of a bird laying down to rest on a branch or a nesting bird). Accordingly, the conceptual dissimilarity between the marks is not as great as the applicant makes them out to be.

177 In conclusion, I have found that the competing marks are visually and conceptually similar to a fair degree. I agree with the PAR’s view that aural similarity is to be disregarded when comparing two device marks and this was also agreed to by the parties in this case – thus, I make no finding in relation to that aspect of similarity.

178 As I have noted above, since the applicant is not disputing the PAR’s finding that the services of the competing marks are similar, I will not deal with this point in any detail. I will only note my agreement with the PAR that, at the very least, some of the services are evidently very similar (see the GD at [109]).

179 I come then to the next step of the Staywell test, which is whether there is a likelihood of confusion. The general principles in relation to the likelihood of confusion are well-established. The Court of Appeal in Staywell held that there are three issues to consider: (a) how similar the marks are; (b) how similar the goods/services are; and (c) given the similarities, how likely the relevant public will be confused (at [55]). To this end, the Court of Appeal prescribed the following non-exhaustive list of extraneous factors for courts and tribunals to consider (at [96]):

(a) Factors relating to the impact of marks-similarity on consumer perception, such as (1) the degree of similarity of the marks themselves; (2) the reputation of the marks; (3) the impression given by the marks; and (4) the possibility of imperfect recollection of the marks.

(b) Factors relating to the impact of the goods or services-similarity on consumer perception, such as (1) the normal way in or the circumstances under which consumers would purchase the goods or services of that type; (2) whether the goods or services in question are expensive or inexpensive items; (3) the nature of the goods or services, and whether they would tend to command a greater or lesser degree of fastidiousness and attention on the part of prospective consumers; and (4) the likely characteristics of the relevant consumers and whether they would or would not tend to apply care or have specialist knowledge in making the purchase.

180 Like their positions taken below, the parties’ main areas of contention relate to (a) the characteristics of the average consumer of the services in question; and (b) the effect of the reputation of the Registered Mark on the likelihood of confusion.

181 As I have found earlier, the marks have a fair degree of similarity and thus would factor into the likelihood of confusion between the competing marks. Also, I have found that the services for which the marks are registered for are similar. Accordingly, these are relevant in considering the likelihood of confusion. The next step is to assess the terminal question of whether there exists a likelihood of confusion on the part of the public arising from (or because of) the similarities of the trade marks and services as assessed in the earlier stages. The court must assess the effect of the objective similarity between the marks on the perception of consumers (see Staywell at [20]). External factors may be considered provided they relate to how the two similarities will affect the consumer’s perception as to trade origin.

182 Given the similarities, I turn to the likelihood of confusion. This requires me to first identify the relevant public. There are two groups which are potentially relevant consisting of the general public and specialist consumers. While specialist consumers are likely to make purchase decisions carefully, the general public is unlikely to pay much attention and care in the purchasing process, especially where the services in question are unsophisticated and inexpensive (see the GD at [118], summarising the respondent’s arguments).

183  In the hearing below, the PAR noted that both parties agreed that the average consumer would include general members of the public, but the PAR did not explicitly accept the respondent’s submissions that it would also include specialist consumers (see the GD at [120]). Nevertheless, in this appeal, the applicant appears to have accepted in the oral submissions that the relevant public comprises both specialist consumers and the general purchasing public, and proceeded to make submissions on that basis.

184 The applicant’s argument is that, even if the relevant public consisted of specialist consumers and the general public, there would be no likelihood of confusion. First, as to specialist consumers, the applicant argues that they are unlikely to be confused because they make purchase decisions carefully. ³⁹

185 Second, as to the general public, the parties’ actual and notional goods and services are or will be predominantly offered to them through online channels such as the Google Play Store and the Apple App Store. ⁴⁰ The applicant says that such online channels generally display the name of the mobile application and its developer/publisher prominently below the name of the mobile application:

This would therefore inform the consumer of the trade origin of the goods and services before he or she purchases them. ⁴¹

186 Accordingly, the applicant argues that the average general consumer in Singapore – who has been exposed to widespread education and global events and who is one that is unlikely to be easily deceived or hoodwinked (see McDonald’s Corp at [64]) – would, even if initially drawn to the Application Mark because of its similarity to the Registered Mark, make the necessary inquiries as to the source of the applicant’s goods or services. Thus, any initial interest confusion as to the origin of the goods or services would be dispelled before the eventual purchase. ⁴²

187 The respondent argues in turn that the general public is unlikely to pay much attention and care in the purchasing process, especially where the goods or services in question are unsophisticated and inexpensive. The competing specifications cover a wide range of IT and computer software-related goods or services. Accordingly, they cover a broad price range, with simpler and less expensive products or services being more likely to be purchased by individuals who have less specialised needs and knowledge, and who would pay considerably less attention in the purchasing and selection process. ⁴³

188 I agree with the respondent. In my judgment, the applicant’s argument does not cohere with how the general public makes such purchases. I find support for my conclusion in the High Court’s reasoning in Digi International. In that case, the court found that the relevant public of the computer and IT-related goods or services in Classes 9, 38 and 42 comprised not just sophisticated consumers, but also the general public (at [174]). Crucially, the court found that those goods or services would, notionally, involve “high-end and complex technology at one end of the spectrum, and low-end and simple technology on the other” and that the simpler and less expensive products would be “more likely to be purchased by individuals who would generally have less specialised needs and knowledge and who would pay less attention in this distinction when selecting the products for purchase” (at [185]).

189 Accordingly, I find that a significant proportion of the relevant public would likely be confused by the Application Mark in relation to the Registered Mark.

190 The concept of a trade mark’s reputation has yet to be clearly defined in Singapore for the purposes of the confusion inquiry, but it generally appears to refer to the level of fame the trade mark has garnered amongst the public (see Vaerhn and Yew at 643). Quite apart from the effect on the relevant public, the applicant urges me to accept the PAR’s decision that the reputation in the Registered Mark would tend to reduce the likelihood of confusion between the competing marks. ⁴⁴ This is because, if consumers are more familiar with a mark, their recollection of that mark would be better than imperfect and they will be more astute as to the differences in another mark (see the GD at [127]). The applicant submits that there is no exaggeration to say that the Registered Mark has completely infiltrated the minds of the relevant public such as the average consumer. This in turn means that general consumers would be more familiar with the details of the Registered Mark, and they will be more discerning such that they “will likely be able to detect differences at a glance”. The applicant reiterated this argument during the oral hearing and relied on McDonald’s Corp (at [64]) and Discovery Communications, LLC v A-STAR-Education Discovery Camps Pte. Ltd. [2020] SGIPOS 4 (at [111]–[112]).

191 I agree with the applicant that, on the facts of the present case at least, the reputation of the Registered Mark would reduce the likelihood that the average consumer would confuse the Application Mark with the Registered Mark. Indeed, in Staywell (at [96(a)]), the Court of Appeal noted that a “strong reputation does not necessarily equate to a higher likelihood of confusion, and could in fact have the contrary effect” (as was the case in McDonald’s Corp). The average consumer would be more discerning of the details that comprise a reputable mark. That said, I do not foreclose the possibility of a different relationship between reputation and confusion in a different case with different facts. Ultimately, the effect of the reputation of the registered trade mark on the confusion inquiry is equivocal and highly dependent on the facts (Vaerhn and Yew at 644). However, this is not the end of the inquiry as that reputation could also lead to a perception of an economic link between the competing marks.

192 The PAR had concluded that the reputation of the Registered Mark increases the likelihood of confusion that consumers may perceive an economic link between the competing marks. The PAR considered that the reputation of the Registered Mark increases the likelihood of confusion in two ways: (a) that the Application Mark is a new iteration of the Registered Mark; and/or (b) that the Application Mark is a modified mark that the respondent is using for new closely-related digital services which are extensions of the respondent’s existing range of services (see the GD at [129]). I cite the relevant portions of the PAR’s decision for ease of reference:

193 There are therefore two substantive issues I need to consider here, namely, whether the reputation of the Registered Mark increases the likelihood of confusion arising from the relevant public’s perception of the Application Mark as: (a) a new iteration of the Registered Mark; or (b) a modified mark of the Registered Mark.

194 However, as a preliminary issue, the applicant submits that the PAR should not have considered the actual use of the Registered Mark, such as the evolution of the Registered Mark. This is because it is not a factor relating to the impact of marks-similarity on consumer perception (citing Staywell at [96(b)]. That there have been several iterations of the Respondent’s logo is a fact external to the only two matters crucial to the confusion inquiry – namely, the impact of the similarity between the marks, and the goods/services. It is only relevant to consider the notional fair uses of the Registered Mark. The applicant thus points out that, as the PAR himself suggested (see the GD at [144]), it would be more appropriate to take such external factors into account in the test of misrepresentation under passing off. ⁴⁵

195 I disagree with the applicant on this point. In my view, the applicant has misconstrued the PAR’s actual reasoning. While the PAR did refer (see the GD at [132]) to the previous bird marks, this does not mean that he relied on those bird marks in his reasoning in relation to whether there would be indirect confusion. All the PAR did was to use it as a helpful “reality check” as traders tend to diversify their trademarks as part of their business and to confirm that putative uses are not fanciful. The PAR double-checked his “answers” against reality, after making a finding of indirect confusion, but that did not mean that he took that into account in making his finding.

196 It is therefore clear that the PAR knew that considering the evolutionary history is impermissible in the confusion inquiry within s 8(2)(b) of the Act. This is evident from the PAR’s decision (at [144]) where he discussed the differences from the analysis in the passing off inquiry where one could legitimately have regard to all the circumstances (citing Hai Tong at [110]), and where “it is permissible to consider all the [respondent’s] prior bird devices, including those which have not been registered as trade marks in Singapore”. This quoted line demonstrates that the PAR was cognisant that for the s 8(2)(b) analysis, he could not actually consider the earlier bird marks when considering whether a likelihood of confusion exists.

197 I turn then to the likelihood of confusion arising from the Application Mark being perceived as a new iteration of the Registered Mark. In this regard, the applicant alleges that the PAR had erred in his observation that it was common for brand logos to be refreshed periodically to keep up with the times (see the GD at [133]). The evidence suggests otherwise that the Registered Mark has been used since 2012 for almost a decade and running. This final iteration of the logo appears to be what the respondent has settled on for the foreseeable future having used it for a much longer period than all previous iterations. There was thus no basis for the PAR to conclude that a consumer would think that the Application Mark is a “refreshed” version of the Registered Mark. ⁴⁶

198 I disagree with the applicant. Indeed, the fact that the Registered Mark has been used since 2012 means that the public may perceive a similar mark to be a new iteration. This is likely all the more so because of the similarities between the marks.

199 Lastly, the applicant also argues that there was no basis for the PAR to conclude that the Application Mark would be perceived by the public as a modified mark that the respondent is using for closely-related digital services which are extensions of the existing range of services. ⁴⁷ In Staywell, the Court of Appeal concluded that the application mark and the opponent’s mark, ST. REGIS, would be perceived to be economically linked as it was “common for large hotel chains to operate differently branded hotels carrying different logos, united only by use of a common denominator in their names”, with these common denominators serving “as an assurance of source and therefore quality” (at [102]). The applicant argues that the situation here is different, as the respondent has failed to show that it was common for providers of its kind of services to operate various software applications carrying different logos, or that the respondent itself has already used another kind of bird mark in relation to a closely related digital service. ⁴⁸

200 The respondent disagrees with the applicant on this point in oral submissions. If the applicant is allowed to use the mark, especially within the Class 42 specification, consumers might think it to be a modified bird mark on a separate product line and perceive some indirect economic link. This is accentuated by the fact that for the services provided under Class 42 specifications, those services relate to things including the provision of information on computer technology and programming via a website, which are provided at low or no cost to consumers. This means that the consumer will likely give an average level of attention during the procuring process, making it ripe for indirect confusion.

201 In my judgment, even if we consider the notional fair uses of the Registered Mark, I am convinced that there will be indirect confusion. This is because both parties use or intend to use their respective marks in relation to closely proximate services. The applicant intends to use the Application Mark in relation to its prospective “lifestyle app” called the “V V Life App”. The services offered under the V V Life App coincide with those under the respondent’s Twitter platform in three aspects: namely, social media and social networking; e-commerce; and concierge services. Accordingly, I am of the view that there is a fair possibility of indirect confusion arising.

202 For all these reasons, I find that there is a likelihood of confusion between the Application Mark and the Registered Mark.

203 For all the reasons above, I find that the respondent has made out the ground of refusal under s 8(2)(b) of the Act. This would be sufficient for me to dismiss the appeal. But for completeness, I turn now to consider the PAR’s decision in respect of s 8(7)(a) of the Act.

204 Section 8(7)(a) of the Act provides as follows:

205 The applicant’s appeal in respect of s 8(7)(a) is that the PAR had erred in finding that the elements of misrepresentation and damage under passing-off were established. ⁴⁹ For the purposes of this appeal, the applicant does not dispute the PAR’s findings that the element of goodwill under passing-off has been established.

206 In response to the applicant’s case on appeal, the respondent’s case is that, for the same reasons it advances in respect of s 8(2)(b), the use of the Application Mark will likely amount to a misrepresentation that the application services in Class 42 are economically linked or associated with the respondent, thereby resulting in a likelihood of confusion. Accordingly, the respondent’s goodwill is likely to be damaged as well. ⁵⁰

207 I pause to note that while it is common for an opposing party to also raise the ground that the use of the later mark by a trade mark applicant would constitute passing off and contravene s 8(7)(a) of the Act, the success in the objection raised under s 8(2)(b) does not necessarily mean assured success in the objection raised under s 8(7)(a), or vice versa. These are somewhat related but independent inquiries. It may be possible that a case can be made out under one head but not the other (see, for example, Staywell at [117] and [162]).

208 The applicant argues primarily that if their position was accepted that the competing marks were found to be similar, then it follows that no misrepresentation is made out. As a fallback argument, even if the competing marks are found to be similar, as the analysis under s 8(7)(a) of the Act is broader than that in s 8(2)(b), we can reference the respondent’s brand guidelines which will show that the respondent’s mark will only be used in a highly circumscribed way, which further militates against the finding of misrepresentation. ⁵¹

209 The respondent’s brand guidelines are rather strict and prescribe that the “Twitter logo is always either blue or white”, while the Application Mark will only be registered in the yellow-orange mix as applied. Further, the respondent’s brand guidelines do not permit even small deviations from their mark, and some examples of misuse cited include:

The respondent’s intention was for consumers to only be exposed to the Registered Mark without any gradations in colour, shadows, or effects, and only in the orientation in which the mark is registered in. Thus, due to the highly circumscribed manner of usage, there would be no misrepresentation.

210 The respondent contested this by pointing out at the hearing that the applicant has missed the point entirely and made a fatal mistake. These brand guidelines are meant to be addressed to third parties but they do not govern the respondent’s own use of its Registered Mark. This is evident from a closer reading of the preamble of the guidelines provided (amongst other parts): ⁵²

It is quite clear from the overall context of this short snippet cited that these guidelines do not bind the respondent, and it has the autonomy to use the Registered Mark in any manner that it wished to. The respondent also points out that it has indeed made modifications to its own marks and displayed them in various colours and against various backgrounds for different events or causes. This would dispel the applicant’s argument that the manner of usage would militate against the finding of misrepresentation.

211 In my judgment, the threshold question for misrepresentation is whether the Registered Mark is distinctive of the respondent’s goods and services. This necessarily follows from my consideration of whether the competing marks are found to be similar. Thus, since the competing marks are similar and the relevant public is likely to be confused in the light of the fair and notional uses of the Application Mark under s 8(2)(b), this would be sufficient to amount to a misrepresentation under s 8(7)(a).

212 The applicant argues that even if misrepresentation is found, there would be no likelihood of damage as the goods and services offered by the parties are not directly in competition.

213 I disagree. As I have found above (at [201]), the parties operate in the same fields of business activity and their respective products provide similar services. They are thus in direct competition with each other. There is a real risk that the misrepresentation could divert sales and customers away from the respondent.

214 For all the reasons given above, I dismiss the applicant’s appeal. In sum, I agree with the PAR that the respondent’s opposition to the applicant’s registration of the Application Mark succeeded under both ss 8(2)(b) and 8(7)(a) of the Act.

215 Unless the parties can come to an agreement on costs, they are to write in with their very brief submissions of the appropriate costs order within 14 days of this judgment.

216 It remains for me to thank Ms Koh and Mr Thng, as well as their representative teams, for their comprehensive and helpful submissions, in a case that has raised some complex issues.