Hindware v. Google and the US divergence: Invisible keywords, visible consequences in keyword advertising

Type "Hindware" into Google today and you will get Hindware. That was not always true. For years, a search for the brand pulled up its competitors first, because those competitors had paid Google to sit there. In May, the Delhi High Court called that trademark infringement, made Google pay Rs 30 lakh, and ordered it never to sell the word again. Eleven weeks later, a US appeals court looked at exactly the same practice and said there was nothing wrong with it at all. Same conduct, same invisible word, opposite answers.

The Indian ruling came on 22 May 2026 in Hindware Ltd. v. Grohe India Pvt. Ltd. & Ors., where Justice Mini Pushkarna held that Google infringes a trademark by selling it to competitors as an advertising keyword, even though the word itself never appears on screen. The American ruling came on August 4 in Deltona Transformer Corporation v. The NOCO Company, where the Eleventh Circuit held that keyword bidding cannot infringe for precisely that reason, because the customer never sees it. Google's appeal against the Delhi judgment is pending before a Division Bench, which has refused a stay, and Google has already argued that the ruling puts India out of step with the rest of the world. It now has a fresh American judgment to point to.

The Indian dispute began in 2013, when Hindware discovered that Cera Sanitaryware and its website developer had bought HINDWARE as a keyword through Google AdWords. Grohe was found to have done the same the following year. Every advertiser eventually settled, which left Google India and Google LLC as the only defendants still contesting the case. That is how a routine competitor dispute turned into a ruling about platforms.

Justice Pushkarna held that a mark need not appear in an advertisement to be used in advertising. HINDWARE is a coined word with no dictionary meaning, so anyone typing it is looking for Hindware and nothing else. By suggesting the word through its Keyword Planner, auctioning it and earning revenue on every click, Google was monetising commercial value it did not own. That attracted Section 29(8) of the Trade Marks Act, and it also cost Google its safe harbour under Section 79 of the Information Technology Act, because a platform that actively suggests and sells is not a passive host.

The American case involved battery chargers. Deltona sells them under the registered mark "Battery Tender." Its competitor NOCO bid on that mark as an Amazon keyword, used the words in its own ad copy, put them into its product descriptions, and told customers the phrase had become generic. A jury found against NOCO on everything, and on appeal it lost almost all of it, including a disgorgement of over USD 12 million. The single thing it won was the keyword point, and it won that completely.

Both courts reason from the same starting fact. The keyword is invisible. The customer sees a competitor's advertisement, never the word that triggered it.

Delhi treats that invisibility as irrelevant. Traffic meant for one trader is diverted to another, the diversion is paid for, and the mark is what makes it work. The absence of the word from the screen does not undo any of that.

The Eleventh Circuit treats invisibility as decisive. If the customer cannot see the mark, the customer cannot be confused by it, and confusion is the whole of American trademark law in this area. A shopper who sees a rival's sponsored listing may be distracted, but distraction is not deception. Judge Newsom used a shop-floor analogy: a customer who walks in looking for Coke and finds Pepsi beside it on the shelf has been offered a choice, not misled. On that reasoning the Court joined the Second, Fifth, Ninth and Tenth Circuits, which had reached the same conclusion over the past fifteen years. The American position is settled, not eccentric.

Where it drew the line is more revealing than the holding itself. NOCO's product descriptions were held to infringe even though the Court accepted that shoppers may never scroll far enough to read them, because a description sits on the page rather than behind it, and because a keyword-triggered advertisement carries a "Sponsored" tag warning the shopper that this is not the brand searched for. Liability therefore attached to text nobody reads, on the footing that it borrows another trader's goodwill to pull in traffic. Free-riding is doing real work in the American analysis too. It has simply been let in through a side door after being turned away at the front.

The deeper disagreement is not about what customers see. It is about what trademark law protects.

The American answer is confusion and nothing else. Lost traffic, diverted custom and freeriding on goodwill are irrelevant unless someone ends up mistaken about who is selling what. The Delhi High Court reached liability without finding that a single consumer was confused, resting instead on unfair advantage and departure from honest commercial practices. That is the ground Google is pressing on appeal.

Google's public position is that the judgment makes India a global outlier. That is only half accurate. A free-riding rationale, decoupled from confusion, is precisely what the Court of Justice of the European Union developed in the Interflora litigation and applies under the reputation limb of European trademark law. India has not invented an idiosyncratic rule. It has chosen the European rationale over the American one, and that choice deserves to be debated on its merits rather than dismissed as an aberration.

There is a cost to it, and Indian commentary has largely ignored it. In the United States, agreements restricting rivals from bidding on trademarks drew a competition challenge from the Federal Trade Commission. Broad keyword protection hands established brands a tool to keep challengers out of the one channel where a smaller competitor can still reach a customer who has already decided what they want.

There is a temptation to hold up the Eleventh Circuit's judgment as an answer to Delhi. It is not, and the point is worth making before an opponent makes it for you.

Amazon was never a party in the American case. NOCO was an advertiser, and the entire judgment is about what an advertiser may do. No question of intermediary liability arose and no safe harbour defence was argued. On the question that actually decided Hindware, the American case says nothing.

The genuine comparator is European. In Google France, the Court of Justice held that the platform does not itself use the mark, and that its exposure turns on whether its role was neutral and technical. That is the holding Justice Pushkarna declined to follow, on the basis that a tool suggesting third-party marks, an auction pricing them, and a revenue model paying out on every click amount to active commercial participation rather than hosting. Whether that survives the appeal is the most consequential question now open in Indian trademark law

The Division Bench issued notice on 10 July, refused a stay and made the matter the lead case in a batch of connected disputes. The injunction and the findings hold the field, and a US circuit judgment binds nobody in Delhi. It is persuasive material for an argument Google was already making, not a development that changes the odds.

For brand owners, the win is narrower than the coverage suggested. HINDWARE succeeded because it is coined, which is what let the Court say that anyone searching it wanted the brand. A descriptive mark will find that reasoning much harder to borrow.

For advertisers, exposure is no longer confined to what appears in the advertisement, though the American case is a reminder that visible copy remains the greater risk in any event. NOCO lost on its ad text, its product listings and its sales conversations, and won only on the invisible part. For platforms, the open question is whether an auction that suggests third-party marks can still be called neutral, and that question is about to get harder. If the wrong identified in Hindware is monetising a mark you do not own rather than confusing anyone, then a generative search system that answers a brand query by surfacing a competitor sits inside the same logic, whether or not anything on the screen is labelled sponsored. The reasoning Google is resisting in Delhi may matter far more for what replaces search than for search itself.

About the author: Nilanshu Shekhar is a Partner at KAnalysis.

Disclaimer: The opinions expressed in this article are those of the author(s). The opinions presented do not necessarily reflect the views of Bar & Bench.

If you would like your Deals, Columns, Press Releases to be published on Bar & Bench, please fill in the form available here.