Indirect confusion: What is it good for?

Published Date
Jun 9, 2023
Related people
While both direct and indirect confusion occur in the real world, the use of this distinction in the legal test of a likelihood of confusion is unnecessary. Further, the requirement to explain a finding of indirect confusion (or a lack thereof) currently hinders the consistent and efficient application of trade mark law in the UK.

The L.A. Sugar categories of indirect confusion should be abandoned as superfluous and overly uncertain. Instead, the likelihood of confusion test should only require a finding of (non-specific) confusion using the ample law that exists outside of the L.A. Sugar framework. 

Likelihood of confusion and the role of the average consumer

Under s.5(2) of the Trade Marks Act 1994 (the TMA), a mark shall not be registered where (because it is identical or similar to an earlier registered trade mark and is to be registered for identical or similar goods/services) “there exists a likelihood of confusion on the part of the public, which includes the likelihood of association with the earlier trade mark”. The same likelihood of confusion test applies for trade mark infringement under s.10(2) of the TMA.

The law on applying the likelihood of confusion test is well established. The matter must be judged through the eyes of a legal fiction, “the average consumer” of the goods/services at issue. We allow our fictional average consumer certain attributes that enables them to guide us in striking a balance between the strong monopolies granted by trade marks and the conflicting rights of third parties. For example, the average consumer:

is deemed to be reasonably well informed, circumspect and observant, but rarely has the chance to make direct comparisons between marks and must instead rely upon the imperfect picture of them kept in the mind;1

  • takes into account the overall impressions created by the marks, bearing in mind their distinctive and dominant components;2 and
  • is attributed a certain level of attention, which reflects how closely consumers look at the branding when purchasing the goods/services at issue.3

Among other situations, a likelihood of confusion may be found where an earlier mark used by a third party in a new composite sign retains an independent distinctive role in that composite sign. In such a case, the overall impression produced by the composite sign may lead the public to believe that the goods/services at issue derive from companies which are linked economically.4

Direct and indirect confusion At present, a likelihood of confusion under ss.5(2) and 10(2) of the TMA is assessed by reference to two species of confusion: “direct confusion” and “indirect confusion”. Direct confusion occurs when a consumer mistakes one mark for another. Indirect confusion occurs when a consumer realises that two marks are not the same, but puts the similarity that exists between the marks (and their goods/services) down to a mistaken belief that the responsible undertakings are the same or somehow related.

The concept of indirect confusion was addressed in detail by Mr Iain Purvis KC sitting as the Appointed Person in L.A. Sugar Ltd v By Back Beat Inc, BL-O/375/10. At [17], Mr Purvis set out a non-exhaustive list of scenarios where he considered one may expect the average consumer to be indirectly confused: “…

(a) where the common element is so strikingly distinctive (either inherently or through use) that the average consumer would assume that no-one else but the brand owner would be using it in a trade mark at all. This may apply even where the other elements of the later mark are quite distinctive in their own right (“26 RED TESCO” would no doubt be such a case).
(b) where the later mark simply adds a non-distinctive element to the earlier mark, of the kind which one would expect to find in a sub-brand or brand extension (terms such as “LITE”, “EXPRESS”, “WORLDWIDE”, “MINI” etc.).
(c) where the earlier mark comprises a number of elements, and a change of one element appears entirely logical and consistent with a brand extension (“FAT FACE” to “BRAT FACE” for example).”

Subsequently, the L.A. Sugar categories have been frequently cited in the UK as a framework for finding indirect confusion, including in the Court of Appeal as recently as 2021.5

By contrast, under European Union (EU) trade mark law, the practical existence of what we call “indirect confusion” is acknowledged but there is no separate legal sub-test for it.6 Instead, a likelihood of (non-specific) confusion is assessed through an analysis of the similarities between the overall impression of the marks at issue bearing in mind their distinctive and dominant elements, the level of attention attributed to the average consumer and the similarities between the goods/services in question.

The L.A. Sugar categories

The L.A. Sugar categories of indirect confusion are unhelpful in three ways: they are imprecisely articulated, they are superfluous and they are uncertain.

Each of the categories is imprecise because paragraph 17 fails to take into account goods/services—this is fatal to a sound analysis of likelihood of confusion. Even assuming that Mr Purvis was considering the goods/services at issue to be at least similar (given this is required by ss.5(2) and 10(2) of the TMA), this only gets us so far. For example, the idea that FAT FACE and BRAT FACE are close enough to cause indirect confusion might be acceptable where the goods are identical (clothing) because BRAT FACE might be a “logical brand extension” of FAT FACE in the context of a new sub-brand of children’s clothing. However, it cannot be assumed that merely similar goods would cause such confusion.

The first and second categories add nothing to UK trade mark law. As to the first category, Mr Purvis explains that indirect confusion occurs “where the common element is so distinctive that the average consumer would assume no one else but the brand owner would be using it in a trade mark”. By way of example, he suggests 26 RED TESCO would “no doubt” fall within this category. Yet, there is no need for this approach in the case of 26 RED TESCO: the average consumer already takes into account the overall impression created by a mark, bearing in mind its distinctive and dominant components. In this instance, “TESCO” is clearly the distinctive element of the mark, both inherently and as a result of acquired distinctiveness. “TESCO” is also likely to be the dominant element of the mark because it is visually outstanding, being larger than the other two elements (albeit an argument might be made that some amount of dominance is lost by its position at the end of the mark). Therefore, the overall impression of the mark would lead to a likelihood of confusion, provided the goods/services are at least similar.

Mr Purvis’s first category remains unnecessary where the earlier mark is not the dominant element. Medion sufficiently caters for non-dominant elements that retain their independent distinctive role within composite marks such that they continue to act as an indication of origin. For example, if the mark in issue was THOMSON TESCO, “THOMSON” would be the dominant element, not “TESCO”, as it is the largest element and it is positioned at the start of the mark. However, the earlier mark “TESCO” retains its independent distinctive role within the composite mark such that the overall impression of the mark would lead the public to believe that the goods/services at issue derive from companies that are linked economically. This might be phrased alternatively as “the common element is so distinctive that the average consumer would assume no one else but the brand owner would be using it in a trade mark”.

It is unclear whether Mr Purvis thought that TESCO would “no doubt” fall within the first category on the basis of its enhanced level of acquired distinctiveness (in addition to its inherent distinctiveness). To the extent that this is the case, such a claim would more sensibly be brought under the dilution of reputation grounds set out at ss.5(3) and 10(3) of the TMA, and not under likelihood of confusion. As to the second category, it is trite law that words such as “WORLDWIDE”, “LITE” and “EXPRESS” are entirely descriptive and/or non-distinctive (depending on the goods/services). As such, these elements are deemed to play no more than a minimal role in the overall impression of a mark on the average consumer. Therefore, any marks falling into this category are adequately dealt with outside Mr Purvis’s framework.

The third L.A. Sugar category adds significant uncertainty to UK trade mark law. Indeed, case law shows that the “logical brand extension” test appears to rest on the experiences and preconceptions of the particular decision maker far more than any other feature of UK trade mark law. For example, in VIRGIN v VIRGINIC Case O/739/18, the hearing officer found that the average consumer would not, when faced with the mark VIRGINIC, assume that it was “a logical brand extension or an evolution of VIRGIN products”. On appeal by Virgin to the High Court, Arnold J (as he then was) agreed with counsel for Virgin that the average consumer would perceive VIRGINIC as “a newly-minted adjective meaning ‘of or pertaining to VIRGIN’”. It followed that the average consumer would be likely to think that VIRGINIC was a brand extension of VIRGIN.7

Further, the third L.A. Sugar category is also redundant. A logical brand extension is one that retains the core element of the brand (or a significant portion of it) as the dominant and/or distinctive element. Any brand extension failing to achieve this would risk not being associated with the relevant core brand by consumers—and would therefore be illogical. We see the dominant and/or distinctive common elements of “AT FACE” in Mr Purvis’s example and in “VIRGIN” above. Common elements in the likelihood of confusion grounds are appropriately dealt with by the weight given to them by the average consumer in their overall impression of the marks at issue as set out above. Again, for marks with a reputation, additional recourse lies in the dilution of reputation grounds, and it is unclear why Virgin did not rely on such grounds.

Further, the non-exhaustive nature of the L.A. Sugar indirect confusion categories has led to surprising results. In short, L.A. Sugar has encouraged decision makers to look for any mental process on the part of the average consumer that is likely to lead them to believe that the goods/services at issue come from the same or linked undertakings. To give a recent example, in APPLE v X-APPLE, Case O/517/22, the hearing officer held that:

“…the average consumer would view the presence of the ‘X’ [in X-APPLE] to be indicative of a collaboration between two brands. In my view, this points to the marks coming from the same or economically connected undertakings. While this is not an example of indirect confusion as set out by Mr Iain Purvis Q.C. in the case of L.A. Sugar, I bear in mind that the examples given in that case are not exhaustive.”

This is confusing, and probably incorrect, given there is no second brand with which to “collaborate” within the mark, but it serves to highlight further the uncertainty created by L.A. Sugar. Again, readers will be able to see how this further category is also appropriately dealt with by likelihood of confusion case law that exists outside of L.A. Sugar and, if applicable, the dilution of reputation grounds.


There is no valid justification for requiring the decision maker to decide whether the average consumer would be directly or indirectly confused merely because these two species of confusion occur in the real world.

First, there is no requirement at law to do so. Under the TMA, it is only necessary to find a likelihood of confusion—and this includes the likelihood of association. Introducing a requirement for the decision maker to decide whether the average consumer is directly or indirectly confused is unnecessary—it need only be determined that the average consumer is confused or not.

Second, given the average consumer is a mere legal device for striking a balance between competing rights in trade mark disputes, it should only be imbued with qualities that benefit this purpose. It follows that the average consumer should not be attributed a given characteristic merely because it reflects reality.

Third, the existing characteristics attributed to the average consumer already account for both direct and indirect confusion. It is not the case that by removing the direct-indirect distinction that cases of indirect confusion will no longer be assessed, only that they will be categorised as cases of non-specific confusion, as is the case in the EU.

Fourth, the L.A. Sugar categories are superfluous and bristling with uncertainty. Decision makers squeeze cases into the categories (or create new categories) of indirect confusion, which encourages incorrect findings. This not only further evidences the lack of a need for a separate species of indirect confusion, but also militates against the present distinction between direct and indirect confusion.

Therefore, it is clear some housekeeping is in order. The procedural inefficiency introduced by the distinction between direct and indirect confusion, with no tangible benefit from the same, necessitates its removal from the likelihood of confusion analysis. Further, the L.A. Sugar indirect confusion categories should be abandoned as gratuitous, if not damaging, guidance for assessing a likelihood of confusion. Making these changes would result in no change to the law in practice, but would leave us with a more efficient and more certain legal framework for assessing the likelihood of confusion under UK trade mark law.

1. Lloyd Schuhfabrik Mayer & Co GmbH v Klijsen Handel BV (C-342/97) EU:C:1999:323; [1999]
2. C.M.L.R. 1343 at [26]–[27]. 2 Sabel BV v Puma AG (C-251/95) EU:C:1997:528; [1998] 1 C.M.L.R. 445 at [23].
3. Lloyd (C-342/97) EU:C:1999:323; [1999] 2 C.M.L.R. 1343 at [26].
4. Medion AG v Thomson Multimedia Sales Germany & Austria GmbH (C-120/04) EU:C:2005:594; [2006] E.T.M.R. 13 at [30]–[31]. 344 European Intellectual Property Review (2023) 45 E.I.P.R.
5. Sazerac Brands LLC v Liverpool Gin Distillery Ltd [2021] EWCA Civ 1207; [2021] E.C.C. 25 at [11].
6. See, for example, Canon Kabushiki Kaisha v Metro-Goldwyn-Mayer (C-39/97) EU:C:1998:442; [1999] 1 C.M.L.R. 77 at [29]–[30].
7. Virgin Enterprises Limited v Virginic LLC [2019] EWHC 672 (Ch) at [21]–[22].

This material was first published by Thomson Reuters, trading as Sweet & Maxwell, 5 Canada Square, Canary Wharf, London, E14 5AQ, in  E.I.P.R. 2023, 45(6), 344-346 and is reproduced by agreement with the publishers. For further details, please see the publishers’ website.

Content Disclaimer
This content was originally published by Allen & Overy before the A&O Shearman merger