The vast majority of trademark disputes revolve around a single concept: likelihood of confusion. Whether your application has been refused, you have been opposed, or you are about to oppose someone else, this is the issue that will actually be argued. So when exactly are two trademarks considered "similar"? In this article we explain the criteria used in the assessment, which arguments actually work, and which ones carry no weight — with examples throughout.
What Does Likelihood of Confusion Mean?
Likelihood of confusion is the risk that consumers will mistake one trademark for another, or — even without actually confusing them — will believe there is an administrative or economic connection between the two. This second scenario matters just as much as the first: if a consumer thinks "these must be marks from the same group," likelihood of confusion exists.
The key point is that the law does not require confusion to have actually occurred. A likelihood is enough. The fact that no one has actually been misled in practice does not invalidate an opposition.
Two Conditions Must Both Be Met
The assessment runs along two axes, and both must be satisfied together:
- Similarity of the signs: Do the marks themselves resemble each other?
- Similarity of the goods and services: Are they used for identical or related products/services?
There is a balancing relationship between the two: if the signs are highly similar, a lower degree of similarity between the goods/services may be enough; if the goods/services are identical, small differences between the signs may not be considered sufficient.
The Three Dimensions of Sign Similarity
1) Visual similarity
The spelling, letter count, word length, typography and figurative elements of the marks are compared. Identical opening letters carry particular weight, since that is where a consumer's eye lands first.
Example: "KALMERA" and "KALMEDA" differ by a single letter and carry a high degree of visual similarity.
2) Phonetic similarity
This looks at how the marks sound when spoken in Turkish. Marks that are pronounced identically are treated as similar even if their spelling differs.
Example: "CLEVER" and "KLEVER" are spelled differently but sound identical. In a radio advert or a spoken recommendation, a consumer cannot tell them apart.
3) Conceptual (semantic) similarity
This compares the meaning or idea the marks evoke. Marks that convey the same meaning in different languages can be regarded as conceptually similar.
Example: "YEDİ DENİZ" (Turkish for "seven seas") and "SEVEN SEAS" are completely different words but express the same concept.
A high degree of similarity found in just one of these three dimensions can, combined with overlapping goods or services, still be enough to result in a refusal or a successful opposition.
The Principle of Overall Impression
Marks are not broken into parts and compared piece by piece; what matters is the overall impression they leave. Consumers do not place two marks side by side and inspect them — when they see one, they compare it against an imperfect memory of the other. This is known as the "imperfect recollection" principle, and it lowers the threshold for similarity.
That is why the defence "our logos are completely different" is usually insufficient on its own. Consumers remember a brand by its name, not its artwork — they search for it by name and recommend it by name.
The Concept of the Dominant Element
For composite marks, the assessment identifies which element is dominant. The generally accepted rules are:
- A word element is treated as more dominant than a figurative element.
- Descriptive additions ("group", "textiles", "market", "digital") add no distinctiveness; the comparison is carried out on the core word.
- Geographical names are not, on their own, considered distinctive.
- The beginning of a word draws more attention than its ending.
The practical consequence is this: if "ARYA" already exists as a mark, filing "ARYA TEXTILES" does not remove the similarity. The additional word is descriptive, and the dominant element remains "ARYA".
How Is Similarity of Goods and Services Measured?
Being in the same class is neither necessary nor sufficient for similarity. The assessment applies the following criteria:
- Nature and purpose: Do the products meet the same need?
- Manner of use: Are they used together?
- Complementarity: Is one non-functional without the other?
- Competitive relationship: Can one substitute for the other?
- Distribution channel: Are they sold on the same shelves, in the same stores?
- Target audience: Do they address the same consumers?
For example, footwear (class 25) and retail services for footwear (class 35) sit in different classes but are highly related. By contrast, "medical devices" and "aesthetic devices" within the same class may not be found related at all.
Who Is the Average Consumer?
The assessment is carried out from the perspective of the average consumer of the relevant goods or services — someone who is reasonably well informed, observant and circumspect, but who does not scrutinise the product under a magnifying glass.
The assumed level of attention varies by product, and this directly affects the outcome:
| Product / service | Consumer attention | Similarity threshold |
|---|---|---|
| Everyday consumables (food, cleaning products) | Low | Low — even small similarities carry risk |
| Clothing, cosmetics | Medium | Medium |
| Cars, real estate | High | High — greater differences are needed |
| Industrial machinery, medical devices | Very high | Very high — the specialist buyer notices the difference |
This table doubles as a roadmap for defence strategy: in a sector aimed at expert buyers, "the consumer's level of attention is high" is a strong argument.
The Distinctiveness of the Earlier Mark
The scope of protection a mark receives is directly proportional to the strength of its distinctiveness:
- Invented (fanciful) marks: Made-up words with no meaning receive the strongest protection. Even minor similarities can amount to infringement.
- Arbitrary marks: Real words with no connection to the product. Strong protection.
- Suggestive marks: Marks that allude to the product indirectly. Medium strength.
- Quasi-descriptive marks: Marks containing words common in the sector. Their scope of protection is narrow; small differences may be considered sufficient.
This is why choosing a fanciful, invented name both improves your chances of registration and strengthens the protection you get afterward.
Defences That Do Not Work
Arguments that come up often in practice but are usually rejected:
- "Our logo and colours are different." Not enough if the word element is dominant.
- "We're in a different price segment." Trademark protection is not divided by price point.
- "Our quality is better." Quality has no bearing on likelihood of confusion.
- "We're in a different city." Trademark protection applies nationwide.
- "No one has ever actually confused us." Actual confusion is not required — a likelihood is enough.
- "We were using it first." This is a valid argument, but it must be documented — an assertion alone is not enough.
Defences That Actually Work
By contrast, the following arguments genuinely strengthen your case:
- Requesting proof of use: If the opposing mark has been registered for more than five years, you can require its owner to prove use. If they cannot, the opposition is refused. See Trademark Use Requirement.
- Narrowing the scope: Disclaiming the overlapping goods/services to save the registration for the remaining scope.
- Letter of consent: Obtaining permission from the other party. See Trademark Registration with a Letter of Consent.
- Weak distinctiveness: Showing that the opposing mark is a term common in the sector, so its scope of protection is narrow.
- Coexistence: Documenting that both marks have operated side by side in the market for a long time without any actual confusion.
Run Your Own Test Before Filing
When choosing a brand name, ask yourself:
- Does this name sound identical to an existing mark in my sector?
- Have I simply derived it from an existing mark by adding a descriptive word?
- Does its equivalent in another language mean the same thing as a registered mark?
- Do the consumers in my sector decide carefully, or quickly?
- If I say the name over the phone, will the other person write it down correctly — or write down a different mark instead?
For research methods, see our articles How to Do a Trademark Search and e-Devlet Trademark Search.
The last question is especially useful. In what we call the "phone test," you say your mark out loud and see how the other person writes it down. If the spelling comes out differently, you have chosen a name that carries a phonetic-similarity risk. You can run the same test on a search engine: if typing your mark prompts a "did you mean…" suggestion pointing to a different brand, that brand is your most likely source of opposition.
Measure Your Similarity Risk with Webx
Likelihood of confusion is a matter of judgment, not a mechanical calculation — the same two marks can produce different outcomes in different sectors. Making that judgment call before you file protects you from the cost of refusals and oppositions.
Webx's trademark consultants run a variation-based similarity search for the name you have chosen and report the risk level in plain terms. Explore our trademark registration service or request a free risk assessment.