The central question in trademark law — whether in clearance, examination, opposition, or infringement proceedings — is whether two marks are similar enough to create a likelihood of confusion among consumers. Two identical marks for identical goods will almost always conflict. Two completely different marks for unrelated goods will almost never conflict. Everything in between requires analysis.
Understanding how similarity is assessed helps you make better decisions at every stage of the trademark lifecycle: when choosing a new brand name, when evaluating clearance search results, when deciding whether to oppose a third party's application, and when assessing whether someone else's use of a mark infringes your rights.
This article explains the framework that trademark offices and courts use to evaluate similarity, and how each dimension works in practice.
European trademark law assesses likelihood of confusion through a global appreciation — an overall evaluation that considers all relevant factors together rather than mechanically comparing each element in isolation. The leading case law from the Court of Justice of the European Union (CJEU) establishes that the assessment must take into account the visual, phonetic, and conceptual similarity of the marks, the identity or similarity of the goods and services, the distinctive character of the earlier mark, and the relevant public and their level of attention.
These factors are interdependent. A higher degree of similarity between the goods can compensate for a lower degree of similarity between the marks, and vice versa. A mark with a strong distinctive character — either inherent or acquired through use — enjoys a broader scope of protection than a weak mark. The assessment is always conducted from the perspective of the average consumer of the goods or services in question, who is deemed to be reasonably well-informed, reasonably observant, and reasonably circumspect.
Phonetic similarity evaluates how the marks sound when spoken aloud. This dimension matters because consumers often encounter trademarks through word of mouth — asking for a product in a shop, discussing a brand with colleagues, hearing an advertisement.
The assessment considers the number of syllables, the sequence of vowels and consonants, the stressed syllables, and the overall rhythm and cadence of each mark. Marks that look different on paper may sound very similar when pronounced — for example, CALEDON and KALYDON, or RESPICUR and RESPICORT.
In multilingual markets — and particularly in EU-wide assessments — phonetic similarity may vary depending on the language. A mark that sounds distinctive in English may be phonetically similar to an existing mark when pronounced in French, German, or Dutch. The EUIPO and national offices consider pronunciation across the relevant languages of the territory where the marks coexist.
Phonetic analysis is not about exact matches. It is about whether the overall sound impression is close enough that a consumer hearing one mark might confuse it with the other. The beginning of a word tends to carry more weight than the ending, because consumers typically pay more attention to the first part of a mark — though this is a tendency, not a rule.
Visual similarity evaluates how the marks look when seen. For word marks, this involves comparing the length of the words, the letters they share, their sequence, and the overall visual impression. For figurative marks (logos), it involves comparing shapes, colours, layout, and graphic elements.
In word mark comparisons, marks that share the same beginning and differ only in their ending are generally considered more similar than marks that share the same ending but differ at the beginning. The length of the mark also matters — a single-letter difference is more significant in a four-letter mark than in a twelve-letter mark.
For figurative marks, the comparison considers the overall visual impression rather than a detailed element-by-element breakdown. Consumers perceive logos as a whole and do not typically analyse individual design elements in isolation. A logo featuring a stylised eagle will be compared to another eagle logo based on the overall impression, not on whether the feathers point the same direction.
When a mark combines word and figurative elements, the word element is often considered more distinctive because consumers tend to refer to figurative marks by their verbal element — they will say the word rather than describe the image. However, this varies depending on the nature of the mark and the goods in question.
Conceptual similarity evaluates whether the marks convey the same or related meanings. Two marks with completely different sounds and appearances can still conflict if they share the same conceptual content.
Classic examples include cases where one mark is a word and the other is a corresponding image (the word APPLE and a picture of an apple), where marks are translations of the same word in different languages (SOL and SUN, both meaning the same thing), or where marks evoke the same idea through different words (DREAMLAND and SLEEPWORLD both evoke the concept of a sleep-related paradise).
Conceptual similarity can also work in the opposite direction — a clear conceptual difference can sometimes neutralise phonetic or visual similarities. If two marks sound similar but convey obviously different meanings, the conceptual distinction may reduce the overall likelihood of confusion. The CJEU has recognised this neutralisation effect, though its application depends on the specific circumstances.
The assessment of conceptual similarity requires an understanding of what the marks mean to the relevant public. A mark that is meaningless in one language may have a clear meaning in another. A mark that is conceptually neutral in one product category may have strong conceptual associations in another.
Trademark rights are not absolute — they are tied to specific goods and services. Two identical marks can coexist if they cover completely unrelated goods and services, because consumers are unlikely to be confused about the commercial origin of the products.
The assessment of goods and services similarity considers the nature and purpose of the goods or services, the distribution channels, the relevant consumers, whether the goods are complementary or in competition with each other, and the usual origin of such goods (whether consumers expect them to come from the same undertaking).
Goods and services are classified under the Nice Classification system, which divides all goods into 34 classes and all services into 11 classes. However, the Nice Classification is an administrative tool, not a legal test. Goods in the same class are not automatically similar, and goods in different classes are not automatically dissimilar. Clothing (Class 25) and fashion accessories (Class 18) are in different classes but may be considered similar because they share distribution channels and are often produced by the same undertakings. Pharmaceutical preparations (Class 5) and cosmetics (Class 3) are in different classes but can overlap for products like medicated skincare.
The level of goods and services similarity required for a finding of likelihood of confusion decreases as the similarity between the marks increases. If the marks are virtually identical, even a moderate degree of goods similarity may suffice. If the marks are only moderately similar, a high degree of goods similarity is typically needed.
The distinctive character of the earlier mark — the mark against which the new mark is being compared — affects the scope of protection. A mark with strong distinctive character, either because it is inherently distinctive (an invented word like XEROX) or because it has acquired distinctiveness through extensive use and reputation, enjoys a broader scope of protection. A wider range of marks will be considered confusingly similar to it.
Conversely, a mark with weak distinctive character — because it is descriptive, common in the trade, or otherwise lacks strong distinguishing power — has a narrower scope of protection. Small differences may be sufficient to avoid confusion.
This is why clearance assessments must consider not only the similarity between marks but also the strength of the earlier marks identified in the search. A high-similarity match against a weak earlier mark may pose less risk than a moderate-similarity match against a famous brand.
In practice, the similarity analysis does not produce a single score or a binary yes/no answer. It produces a nuanced risk assessment that weighs all factors together.
A mark may be phonetically similar but visually and conceptually different. It may cover identical goods but have low mark-to-mark similarity. The earlier mark may be well known in some jurisdictions but unknown in others. The relevant consumers may be highly attentive professionals (reducing the risk of confusion) or casual general consumers (increasing it).
The skill in trademark clearance and prosecution lies in weighing these factors correctly — understanding that the assessment is holistic, that interdependencies between factors matter, and that the outcome depends on the specific factual matrix rather than a formula.
When evaluating clearance search results, the most useful approach is to assess each potentially conflicting mark across all four dimensions (phonetic, visual, conceptual, and goods overlap), consider the strength of the earlier mark, factor in the relevant consumer’s level of attention, and reach an overall conclusion about the level of risk — not just whether the marks are similar in any single dimension.
For brand selection. Understanding how similarity is assessed helps you choose marks that are defensible. Invented words with no meaning in any language give you the strongest position. Marks that are descriptive or suggestive of the goods may be harder to register and harder to enforce.
For clearance. Do not dismiss a search result just because the marks look different on paper. If they sound similar, or if they convey the same concept, or if the goods overlap significantly, there may be a real conflict. Conversely, do not overreact to a mark that shares a few letters if the overall impression — across all dimensions — is clearly different.
For opposition. When deciding whether to oppose a published mark, the similarity analysis determines your likelihood of success. A strong case has alignment across multiple dimensions — phonetic, visual, conceptual, and goods overlap. A case that relies on only one dimension is harder to win.
For enforcement. The same framework applies when you discover a potentially infringing use in the market. Before taking action, assess the similarity objectively across all dimensions. Aggressive enforcement against marks that are only marginally similar can be costly and counterproductive.
If you need a trademark similarity assessment or clearance analysis, get in touch or schedule a meeting with our team.
%402x.svg){kind=small}
%402x.svg){kind=icon}
%402x.svg){kind=icon}
