A stylized image of a search tab with the underlying code and metadata behind it

The central requirement for trademark infringement is a likelihood of confusion (“LOC”). However, confusion need not occur at the point of sale. Under the initial interest confusion doctrine, a consumer may be drawn to a competitor’s product because of its similarity to another’s trademark, even if the consumer recognizes the product’s true source before making a purchase. This is often applied in metatag trademark cases. Metatags are descriptive words contained in the HTML code that describe websites contents, which help generate search engine results through keyword advertising and search ranking. The issue arises when a party hides the trademark of another in its own website’s metatags. 

In Brookfield Communications Inc. v. West Coast Entertainment Corp, 174 F.3d 1036, 1057-58 (9th Cir. 1999), the court found that the defendant’s use of the plaintiff’s trademark ‘MovieBuff’ in its metatags amounted to “free-riding” on the plaintiff’s goodwill. It observed that even if consumers would not be mistaken as to the origin of the services when purchasing, they may decide to remain on the defendant’s site and find it adequate to shop there. This was determined to be a mala fide use of the plaintiff’s trademark. 

In “Initial Interest Confusion in Metatags: The Move from Confusion to Diversion,” Niki Woods criticizes the Brookfield decision and subsequent developments in trademark law. She argues that courts have misapplied the initial interest confusion doctrine by treating the use of another’s trademark in metatags as nearly automatic evidence of infringement. In doing so, courts have increasingly treated consumer diversion as inherently harmful, rather than requiring proof that consumers are actually confused.

First, Wood opposes Brookfield’s extrapolation of a brick-and-mortar billboard analogy to internet usage. In Brookfield, the court discussed a hypothetical scenario: when consumers on a freeway see a billboard advertising a store, take the exit, and then later find the store was actually owned by a competitor, they are more likely to continue shopping there than getting back on the freeway and looking for the real store. This, Woods argues, grossly overstates the costs of being diverted on the internet, where a consumer merely has to click “back.” Secondly, Woods suggests there may be legitimate justifications for such usage. Lastly, she argues that the emphasis on preserving goodwill and preventing free riding is overweighted. Trademark protection serves only to protect consumer welfare, not give owners a property right against the world. Instead, fear of liability will result in rivals publishing less information, leading to less choice and higher search costs for consumers. 

Woods’s assertions, however, are theoretically flawed for the following reasons. 

Returning to search results costs more than just “one click.” While Woods’s argument may hold true for individual consumers, the ease of returning does not undo the diversion of consumer attention. Such use can be harmful because it takes advantage of search rankings and misleads consumers, who may not realize they have been redirected.

It is impossible to apply the standard LOC test to metatag cases. The LOC test relies on what the consumers saw, in terms of similarity of marks, ad labelling, etc. Metatags and keyword searches are invisible, so the LOC factors cannot adequately address the harm. 

Legitimate justifications may sometimes be pretextual, making a rebuttable presumption model more appropriate. Woods argues that competitors may legitimately use another’s trademark to market their own products, such as through comparative advertising or descriptive uses. While this argument holds for clearly labeled comparative advertising, hidden use of a competitor’s trademark provides consumers with no useful information and suggests bad-faith intent. Defendants could rebut this presumption by demonstrating a legitimate purpose, such as nominative or descriptive use.

Protecting trademark goodwill is essential to consumer welfare. Free riding on another’s reputation can harm consumers because trademarks convey information about a product’s quality and source. This, in turn, encourages trademark owners to invest in quality by allowing them to benefit from the reputation they have built. A competitor that simply bids on its rival’s name or buries it in metatags appropriates the value of that signal—such as  consumer attention—without paying for the quality that built it. Consumers searching for a particular brand now must search through results they never sought, which is precisely the search cost trademark law exists to prevent. At scale, the mark stops being a reliable source identifier. This erodes the signal and runs afoul of trademark law objectives.

Today, Woods’s proposition finds judicial acceptance. In Lerner & Rowe PC v. Brown Engstrand & Shely LLC, 119 F.4th 711 (9th Cir. 2024), the court did not presume diversion to constitute infringement, and instead demanded evidence of actual confusion. It observed that reasonably prudent consumers on Google are used to scrolling past ads to find organic results. This tracks with Woods’s argument that consumers are more technologically savvy than courts give them credit for. The court also noted that intent to compete is not intent to deceive, thus supporting Woods’s idea that the mere use of a rival’s mark behind the scenes may sometimes not be infringing. 

However, this is a false dichotomy. “Competing” on factors such price, quality, and comparative advertising is legitimate. “Intercepting” by grabbing the attention a rival’s name generated is dishonest. It involves fully exploiting another’s goodwill without paying for the cost in building the reputation behind it. Treating both as an “intent to compete” blurs this important differentiating factor. 

Trademark law must protect the mark that signals a consumer towards a directed search. When a competitor hijacks search traffic to divert consumer attention, courts should treat the bad faith as strong evidence of trademark infringement, not dismiss it as mere “competition.” Where a competitor deliberately diverts a consumer by unauthorized use of the plaintiff’s mark, the diversion itself is the confusion, amounting to trademark infringement. This is exactly the “bait and switch” confusion the initial interest doctrine seeks to avoid, and the traditional LOC test fails to address such infringements in isolation.