The EUIPO Opposition Division upheld Chanel’s opposition to a three-dimensional perfume-bottle application in a decision dated 9 October, rejecting the application in its entirety. As reported by The Fashion Law, the contested mark was filed in April 2024 by the Romanian company Chaos Perfumes and depicted a white-and-gold bottle carrying the words CHAOS ANGELIC TOUCH EAU DE LIGHT; Chanel opposed that September, relying on its earlier French national registration for the shape of the No. 5 bottle and arguing that the later mark would ride on that shape’s reputation.
The applicant’s defence was the standard one against shape marks, and it is worth reading closely because most brands’ files are vulnerable to exactly it: consumers treat perfume bottles as packaging rather than as badges of origin; Chanel’s evidence proved the fame of the fragrance and the name, not of the container; and what was once an innovative silhouette has become conventional perfume packaging. The Opposition Division rejected it on the evidence. Chanel showed genuine trademark use and reputation for the shape in France through advertising, press coverage, exhibitions and commercial records, and — the detail that did the heavy lifting — a 2016 survey of 2,438 women in France in which 72% spontaneously associated an unbranded image of the bottle with Chanel. The office also noted the bottle’s appearances in art, including works by Andy Warhol.
What it then did with that evidence is the part practitioners should diarise. The Division accepted that the bottle shape has low inherent distinctiveness, because it is, after all, a container for perfume. It accepted real differences between the two marks — Chaos Perfumes’ wording, its colours, its labels — and held the marks visually and conceptually similar only to a low degree. The common ground was narrow: a compact rectangular body, rounded corners, a short central neck and a broad, flattened stopper. Reputation plus identity of the goods carried the rest. French consumers would make a mental link between the two, and that link would hand the applicant an unfair commercial advantage by transferring part of the recognition and image of Chanel’s bottle to a rival product, making it easier to recognise and remember. No likelihood of confusion was needed, and none was found necessary.
What it changes
Your clearance search was looking in the wrong register. The right that killed this EU-wide application was a national French registration, not an EUTM. An applicant who searched the EUTM register for identical or similar bottle shapes would have seen nothing and filed with confidence — and lost two and a half years later. For anyone launching packaging into the EU, a shape-mark knock-out has to run across the national registers of the major markets as well as the EUTM register, and it has to search by figurative/shape classification rather than by word, because the obstacle here contains no words at all. That is a different exercise from a word-mark availability check, and it is the cheap half of this decision’s lesson; see our trademark search work for the shape of it.
“We changed the name, the colours and the labels” is no longer a defence. The dupes-market design brief — keep the silhouette, change everything that carries a trademark — is precisely what failed. Low similarity did not save the application; it was weighed and then outranked by reputation and goods identity. If you are advising on a lookalike launch, the governing question is not “would a consumer be confused?” but “will a consumer make a link, and will we benefit from it?” Those have different answers, and only the second one matters on this ground. Note too that the rejection was total: there was no partial survival for the classes further from the opponent’s core goods.
Reputation in the shape, proved separately from the word marks, is the asset — and almost nobody is building the file for it. Chanel won because it could show the container doing origin work on its own. Most brands cannot: their shape appears in every advertisement underneath the logo, so the evidence establishes the name’s fame and nothing about the silhouette. Two practical consequences. First, if a shape or packaging mark matters to you, commission the unbranded-stimulus survey now, while the look is current — Chanel’s decisive survey was ten years old at decision date, which is the point: this evidence appreciates with age and cannot be manufactured once a dispute starts. Second, audit how the shape is actually used. A registered shape mark that only ever appears with the word mark on it is exposed to non-use revocation after five years, and a registration you cannot defend on use is a registration you cannot oppose with. Renewal dates, five-year use windows and the two-month window to appeal an Opposition Division decision to the Boards of Appeal all belong in a docket, not a diary — our trademark docketing practice exists for the gap between “we own it” and “we can enforce it”.
Low inherent distinctiveness is no longer a reason to abandon a shape filing. Shape marks are hard to register and easy to write off as decorative, and many portfolios quietly stopped filing them. This decision is a reminder that the ceiling on their value is set by the evidence of reputation you accumulate, not by the weakness of the mark as filed. A shape registration that scraped through on acquired distinctiveness can still, decades later, block an EU-wide application on a finding of low similarity. That changes the cost-benefit on maintaining a thin national shape registration you were about to let lapse.
One timing point for both sides. The application was filed in April 2024, opposed in September 2024 and decided in October 2026 — roughly two years of pendency before a first-instance answer, with an appeal still open to the losing party. Neither a brand owner planning enforcement nor a challenger planning a launch can treat an EUIPO opposition as a quick clearance. Build the evidence file before you need it, and assume the commercial decision will have to be made long before the office makes its own.