Can a seller use another brand’s name to sell a compatible accessory in the EU? Yes, within a limit the Court of Justice drew in 2005. A trader may use a trade mark it does not own where that is necessary to tell the buyer what the product fits, and only if the use stays honest: no impression of a commercial connection with the brand, no free-riding on its reputation, no discrediting it, and no presenting the product as an imitation. Whether “Acme charger” and “charger for Acme” sit on the same side of that line is the question this article is about, because a growing class of e-commerce targets earns its margin in the space between them.
The model is familiar and, in itself, legitimate. A seller builds a catalogue of accessories for other companies’ products: cables, cases, filters, blades, capsules, batteries, replacement heads. That is the aftermarket, and EU law protects it on purpose. The grey zone is in the presentation. Titles lead with the original manufacturer’s name. The seller’s own brand is absent, or sits in the brand field as a string of capital letters nobody searched for. Sponsored ads bid on the manufacturer’s brand term and show a title that reads like the manufacturer’s own accessory. Then the reviews arrive: “not original”, “thought this was genuine”, “stopped working after a month”. The buyer paid for an origin they did not get, and the brand whose name did the selling now has a one-star product sitting in its own search results.
The Gillette Case
The facts were small and the ruling was not. LA-Laboratories, a Finnish company, sold razor blades under its own mark, Parason Flexor. On the packaging it put a sticker: “All Parason Flexor and Gillette Sensor handles are compatible with this blade.” Gillette sued for infringement of its Finnish marks GILLETTE and SENSOR. The Helsinki District Court found for Gillette. The Court of Appeal reversed. The Finnish Supreme Court referred the question to Luxembourg, and on 17 March 2005 the Court of Justice handed down Gillette v LA-Laboratories, Case C-228/03.
Four points from the judgment still govern every compatible-accessory listing in the EU:
- Necessity. Using someone else’s mark to indicate the intended purpose of a product is lawful where that use is in practice the only means of providing the public with comprehensible and complete information about that purpose. If the buyer cannot know what the blade fits without the word Gillette, the word Gillette is permitted.
- Honest practices. The use must follow honest practices in industrial or commercial matters, and the Court listed what fails that test: use that gives the impression of a commercial connection between the third party and the mark owner; use that affects the value of the mark by taking unfair advantage of its distinctive character or repute; use that discredits or denigrates the mark; and use that presents the product as an imitation or replica of the product bearing the mark.
- Own competing product is no bar. That LA-Laboratories also sold its own handle did not disqualify it from referring to Gillette’s.
- Purpose is not a quality claim. Naming a mark to indicate what a product fits does not by itself present that product as being of the same quality as the original, so the reference alone is not misleading. The national court must assess the overall presentation to decide whether the honest-practices conditions hold.
One feature of the case is easy to miss and matters for the rest of this piece: it was fought by Gillette. No Finnish consumer authority brought it. That is the first clue to who actually polices this line.
Where the Rule Sits Today
The exception the Court interpreted is now Article 14(1)(c) of the EU Trade Mark Regulation and the identical provision of the Trade Marks Directive, which national laws such as Sweden’s Varumärkeslag implement. The 2015 recast widened the wording: a mark owner cannot prohibit use of the mark “for the purpose of identifying or referring to goods or services as those of the proprietor”, with intended purpose of accessories and spare parts named as the leading example. Article 14(2) keeps the honest-practices condition, and the Gillette list is still how courts read it.
The Court revisited the boundary in January 2024 in Audi v GQ, Case C-334/22. A seller of aftermarket radiator grilles had shaped the mounting point to the outline of Audi’s four rings. The Court held that the exception covers saying a part is for an Audi; it does not cover building the mark into the part. The referential use is the sentence, not the logo.
For the sponsored title, a second line of cases applies. In Google France (2010), Portakabin (2010) and Interflora (2011) the Court held that bidding on another company’s brand as a keyword is not infringement in itself. The ad infringes where it does not enable a normally informed and reasonably attentive internet user, or enables that user only with difficulty, to ascertain whether the goods come from the mark owner, from an economically connected company, or from a third party. That test does not read the keyword. It reads the ad.
“For Acme” or Just “Acme”: Where the Line Sits
No EU text prescribes a word. The law tests the impression the overall presentation creates, not the preposition. But three things follow directly from the case law, and marketplaces have turned them into rules.
The mark must be doing descriptive work. “Charging cable for Acme Model X” tells the buyer what fits. “Acme charging cable” names an origin. The Unfair Commercial Practices Directive lists commercial origin among the main characteristics a trader may not mislead about, and the trade mark cases treat the second phrasing as exactly the use that stops the buyer ascertaining origin.
The seller’s own identity must be visible. In Gillette the blade carried the seller’s own mark and the Gillette name appeared in a compatibility sentence. A title that carries the manufacturer’s name, no seller brand, and “Generic” in the brand field is a listing made mostly of someone else’s trade mark. That is the “commercial connection” limb failing by omission.
Prominence and quality words are origin claims. “Original”, “genuine”, “OEM”, “official” are statements about who made the product, not about what it fits. Logos are out, per Audi. A title or image that gives the manufacturer’s mark more weight than the seller’s own cuts against honest practices even when a connector word is present.
Amazon’s published compatibility guidance encodes this as a format: the seller’s brand, then the product, then “for”, “compatible with”, “fits” or “intended for”, then the other brand and its product. The other brand’s name may appear; its logo may not. Titles that do not follow the format may be removed as potentially infringing. Google’s Ads trademark policy allows the mark in ad text where the landing page is primarily dedicated to selling components, replacement parts or compatible products corresponding to the mark, and requires the page to be clear about whether the advertiser is a reseller or an informational site. The platform rules are stricter than the courts and act in days rather than years. In practice they are the rules that bite.
Does Bad Quality Hurt the Brand in a Legal Sense?
A seller can fail the Gillette test with a perfect product, so quality is not the test. But poor quality under a borrowed name engages two of the four limbs, and both are the brand’s to raise.
The third limb is discrediting or denigration. A product that fails, and is read by the market as the manufacturer’s product, does to the mark precisely what the honest-practices condition exists to prevent. For marks with a reputation, the Court gave this its own definition in L’Oréal v Bellure (2009): detriment to repute arises where the third party’s goods are perceived in a way that reduces the mark’s power of attraction, in particular where those goods have a characteristic or quality liable to have a negative impact on the mark’s image. A cable sold as “Acme cable” that melts is that paragraph in product form.
The second limb, unfair advantage, needs no proof of harm at all. L’Oréal held that riding on the coat-tails of a mark with a reputation, to benefit from its power of attraction without paying for the marketing effort behind it, is itself the infringement. A title that converts because the buyer trusts Acme is the coat-tail.
The Misleading and Comparative Advertising Directive says the same thing from the advertising side. Any advertising that identifies a competitor’s goods, which a compatibility claim does, is permitted only if it does not take unfair advantage of the reputation of a trade mark, does not discredit it, and does not present the goods as imitations or replicas of goods bearing a protected mark.
So the answer is yes: selling low-quality accessories under a presentation that lets buyers take them for the original is a legally recognised harm to the brand. The next question is who gets to do something about it.
Compliance Risk, or the Brand’s Problem?
The exposure sits in three layers, and the enforcement posture differs sharply between them.
Trade mark law is private. No authority enforces a trade mark. The owner sues, or the owner complains to the platform, or nothing happens. The Gillette line is drawn by courts but patrolled by brands, and only by brands that choose to spend on it.
Consumer law is public, with the brand as a second plaintiff. The Unfair Commercial Practices Directive reaches the same conduct from the consumer’s side. Annex I, point 13, blacklists promoting a product similar to a particular manufacturer’s “in such a manner as deliberately to mislead the consumer into believing that the product is made by that same manufacturer when it is not”. Article 6(2)(a) catches any marketing that creates confusion with a competitor’s products or trade marks, and Article 6(1)(b) covers misleading claims about commercial origin. National authorities enforce, and for widespread infringements member states must provide fines whose maximum is at least 4% of turnover in the markets concerned. Article 11 also gives competitors standing. Two caveats keep this a tail risk rather than a headline one: point 13 requires deliberate deception, and consumer authorities’ sweep priorities have sat elsewhere, on reviews, pricing and green claims. The public route gets live fast if the accessory is unsafe, because then the seller is the manufacturer or importer of record under GPSR in its own right, with no brand to hide behind.
National marketing law merges the two. Sweden’s Marketing Act (Marknadsföringslagen 2008:486) implements both directives in one statute. It prohibits misleading imitations of another trader’s known and distinctive product, and its general clause has long been applied to renommésnyltning, free-riding on another’s reputation. Both the Consumer Ombudsman and an affected competitor can bring the case before the Patent and Market Court, with prohibition orders, a market disruption fee and damages available. Germany’s unfair competition act gives competitors similar standing. In these systems the brand does not have to wait for a regulator; it is a plaintiff with the regulator’s remedies.
The honest summary: this is primarily the impacted brand’s problem to raise, with public enforcement as a secondary route that widens if consumers are demonstrably deceived or hurt. That gives the risk a different shape from the rest of this series. There is no transposition date. There is a counterparty with a legal budget and a Brand Registry login, and enforcement arrives as a takedown, not a fine. The Digital Services Act obliges marketplaces to process infringement notices expeditiously and to suspend repeat infringers, a mechanism mapped in The Marketplace Enforces First. A brand owner with a registered mark and a test purchase can clear a seller’s catalogue from a marketplace in days. The seller then appeals, re-titles, and discovers what its listings convert at under their own name.
Where This Sits in the Waterfall
The exposure lands in GM3, with a GM2 echo.
In GM3 the mechanism is the one this series calls a borrowed signal. The title and the brand-term bid deliver traffic at a cost the seller would never achieve on its own name, and convert at a rate that is partly the manufacturer’s reputation at work. Two normalisations expose it. First, rewrite every title to the form the brand owner could not object to, and estimate what click-through and conversion do. Second, remove the traffic that arrives on the manufacturer’s brand terms, since an enforcement sweep removes it anyway. What remains is the GM3 the target actually owns. A seller that converts at a category multiple on titles its own brand never appears in is the same pattern as Too Good to Be Earned, with a trade mark doing the work the dark pattern does there.
In GM2 the echo is returns and refunds driven by origin confusion. A buyer who expected the original and received a compatible returns at a rate the catalogue average hides, and the review language quantifies it before the returns data does, in the way The Promise on the Product Page uses timestamped reviews for delivery.
Above both sits concentration. If most of the target’s revenue runs through listings whose titles fail the marketplace format, the catalogue is one complaint campaign away from a revenue event, with the ASIN review history, the sales rank and the sponsored placements going with it. That is not a margin compression. It is a cliff with the trigger held by someone else.
The mirror case matters for brand acquirers. A target that is the manufacturer, with its own accessory line, carries the cost of this pattern from the other side: attach rate eroded by a cloud of compatibles on its own search terms, a ratings environment polluted by products it never made, and a policing budget for test purchases, counsel and platform programmes that sits in G&A and grows with the brand’s success. That budget is a fixed cost the EBITDA bridge should carry, and the accessory line’s GM1 and GM3 should be read against the compatible cloud rather than against the brand’s own history.
Why This Slips Through Due Diligence
Legal diligence audits the marks the target owns. The IP schedule lists registrations and oppositions. Nothing in it lists the third-party marks the catalogue depends on to be found, which is the asset that is actually at risk. Platform account health is read as a status, green or not, rather than as a history, and the takedown log lives in the seller console, not the data room. Commercial diligence reads the ratings and the rank, both of which are high, because the manufacturer’s name converts. And “compatible with” sounds like compliance when it appears in the CIM, so nobody asks whether every title says it, in the right order, with the seller’s own brand in front.
The target’s own team may describe the practice as keyword optimisation, because that is what the listing tools call it.
What Outside-In Analysis Can Detect Before the Data Room
The whole pattern is visible on the surface the buyer sees, and a title census is the part of a screen that maps a seller’s dependency on names it does not own:
- Title structure across the catalogue: the share of listings carrying a third-party mark; whether a connector word is present and where; whether the seller’s own brand appears and leads; the brand field reading “Generic” or a registration-only string; origin words such as “original”, “genuine” or “OEM”
- Ad copy on the manufacturer’s terms: sponsored titles and search ad text served against the other brand’s name, read against the Interflora test of whether the ad lets the buyer tell who made the product
- Imagery: product photos and packaging showing the manufacturer’s logo, or the mark reproduced on the product itself, which is Audi v GQ territory rather than Gillette territory
- Review language: “not original”, “fake”, “thought it was Acme”, “doesn’t fit”, with the rate compared against compatible sellers whose titles are compliant, so that the confusion is measured relative to the format rather than to zero
- Takedown scars: listings that vanished and returned under a new title, review histories with a discontinuity, suppressed listings and merged variations
- For brand targets, the compatible cloud: how many third-party accessories surface on the brand’s own terms, their average rating against the brand’s own accessory line, and how many present as the brand’s own
None of this is proof of infringement. Each is a measure of how much of the catalogue’s conversion is on loan, and together they scope a request that rarely gets made: the full IP complaint and takedown history from every marketplace, the title and ad-copy policy, and any correspondence with the brands whose names carry the catalogue.
The Pre-LOI Question Every PE Fund Should Ask
Gillette settled that an accessory seller may use the brand’s name to say what fits. It did not settle that the name may do the selling. The question for an acquirer is: would the target’s revenue and GM3 survive every title being rewritten to the form the brand owner could not object to, and every ad on the brand owner’s name being switched off?
A target that survives that rewrite owns its margin and sells into a legally protected aftermarket. A target that does not is renting its conversion from a trade mark whose owner can end the tenancy on notice, and the acquirer would be buying the lease without having read it.
This analysis is part of Tronvik’s GP3 Waterfall methodology, focused on GM3 marketing-efficiency mapping. Nothing in this article constitutes legal advice. To initiate an outside-in listing-title and brand-dependency screen on a specific acquisition target, contact info@tronvik.com.