Glossary
French: recherche de similarité
The similarity search (recherche de similarité, also recherche de similitude) consists of searching for all marks similar to a proposed mark in order to determine whether a sign is available. It is the full-scope search a US practitioner would call a comprehensive search, as opposed to a knock-out screen.
The search applies the criteria that trademark law itself uses to decide whether one sign is similar to another: similarity is assessed on the conceptual, phonetic and visual levels — the same three-way analysis used in oppositions and infringement cases (see similarity of signs).
In practice, search engines generate lists of marks sharing structural features with the proposed sign:
The similarity search differs from the more limited identical search (recherche à l’identique), which only retrieves marks that are identical or contain the same string of letters. Because most real-world conflicts — oppositions and infringement claims — involve similar rather than identical marks, only a similarity search gives a reliable picture of the risk before a French or EU filing.
An identical search alone is fast and cheap to run, and it does catch the most obvious problem — someone already owns exactly the mark you want. But it is a knock-out screen, not a clearance opinion: it says nothing about the far larger universe of marks that are close enough in sound, appearance or meaning to support an opposition or an infringement claim. Because French and EUIPO opposition deadlines run for a fixed, non-extendable window after publication, a mark that clears an identical search but not a genuine similarity search can still surface a costly conflict the moment it publishes — at which point withdrawal or a coexistence negotiation, rather than a clean launch, becomes the only option.
A similarity search does not return a yes/no verdict; it returns a list of potentially conflicting marks that a trademark attorney must then screen manually, applying the same similarity of signs and similarity of goods and services factors a court or the INPI would apply in an actual dispute. Most hits turn out to be false positives on closer review — different goods, a weak or descriptive shared element, or only a superficial resemblance. The value of the exercise lies precisely in that triage.
US practitioners will recognize the similarity search as the trademark-clearance equivalent of a full or comprehensive search — as opposed to a knock-out (or “preliminary”) screen — run before advising a client on the strength of a mark and before filing with the USPTO. The underlying logic is identical on both sides of the Atlantic: a knock-out search only protects against the most obvious conflicts, while a full search is what actually supports a reasoned clearance opinion. One structural difference matters for US clients: because the INPI and EUIPO do not examine new applications against earlier marks on their own initiative, a thorough similarity search is, in practice, the only line of defense against a future opposition — there is no examiner cross-check standing behind the applicant the way TTAB practice sometimes assumes.
A US beverage company wants to launch “AQUABLUE” in France. An identical search shows no exact match and might tempt the company to file immediately. A similarity search, however, surfaces an earlier French mark “AQUA BLEU” registered for related beverages — visually near-identical, phonetically indistinguishable in French, and conceptually the same (the French translation of the English term). That earlier mark would not have appeared as a problem under an identical-only search, yet it is exactly the kind of prior right an opposition would be built on.
A similarity search is the analytical core of any serious trademark clearance search, run before choosing a trademark and before committing to a French or EU filing. See also prior-rights search and availability of a sign.