The Road Not Taken: Initial Interest Confusion, Consumer Search Costs, and the Challenge of the Internet

• Publication year: 2004
• Citation: Vol. 28 No. 01

SEATTLE UNIVERSITY LAW REVIEWVolume 28, No. 1FALL 2004

The Road Not Taken: Initial Interest Confusion, Consumer Search Costs, and the Challenge of the Internet

Michael Grynberg(fn*)

Yesterday I went shopping for soy milk. I usually buy the "Silk" brand. A carton of plain Silk is light-red with white lettering. Beneath the lettering is a picture of soy milk pouring and splashing into a bowl of cereal. At the store's soy milk display, a light-red carton with white lettering over a picture of soy milk pouring and splashing (but no cereal) caught my eye. Closer inspection revealed that it was not Silk (which was placed directly above in the store's display), but a store brand with its own prominent trademark. I compared the ingredients (identical), bought the cheaper Silk substitute, and saved a dollar. Though the outcomes of trade dress cases are often hard to predict, the store brand's copying of the Silk carton was likely legal.

Suppose instead that I do my grocery shopping on the Internet. Not knowing where to buy soy milk (and not knowing any brand other than Silk), I might have typed "Silk" and "soy milk" into a search engine. My result list would likely include the Silk web page.(fn1) But what if the store brand had a web site that incorporated the word "Silk" on its page in order to be included in my results list? Though no more confusing to me as a consumer than the mimicked trade dress, this act would give Silk a likely cause of action for "initial interest confusion." This article critiques trademark claims based on initial interest confusion and proposes a new framework, focused on consumer search costs, for analyzing these claims.

Likelihood of confusion is the touchstone of trademark infringement. But confusion is often transitory. Although a consumer may make an initial mistake about a product's source, affiliation, or sponsorship, she might still choose to buy the product after the true facts are known. Nonetheless, some courts find trademark injury where there is initial interest confusion. They hold that the initial confusion gives an unfair "foot in the door" to infringers.(fn2)

The popularization of the Internet as a marketing tool has increased judicial interest in initial interest confusion. Courts aggressively police defendants who use plaintiff trademarks in domain names or metatags, which are hidden text on a web page designed to draw search engines to a site. Web surfers never see the trademark used as a metatag (which is invisible when reading the page), but may visit the site because it is returned by the search engine in response to a query that includes the trademark as a search term. Thus, a search for "Coca-Cola" could return the "Pepsi" web site if that site includes "Coca-Cola" in its metatags. Several circuits take the view that such inclusion infringes Coke's trademark despite the fact that no one would mistake the Pepsi site for Coca-Cola's. Nor does it matter that the cost to the consumer of any diversion is no more than an extra click of a mouse. These holdings potentially deprive consumers of an easy means to learn about web sites in the same category as their search.(fn3)

The application of initial interest confusion to the Internet has generated extensive criticism.(fn4) Most critics complain that courts have failed to understand the nature of the Internet. But the problem is as much a problem of doctrine as of technology. Courts applying initial interest confusion to the Internet are fully consistent with the logic of the cases that established the doctrine.(fn5) That logic, unfortunately, is vague as to why and when initial interest confusion should be enjoined. It is therefore not surprising that the resulting body of law produces questionable results. Simply put, if applying initial interest confusion to the Internet was a mistake, it was an accident waiting to happen.

The rationale for policing initial interest confusion is not immediately obvious. If there is no confused consumer at the point of sale, where is the harm? The usual answer points to the trademark owner's goodwill (the positive associations consumers have with the trademark owner's product): A junior user should not free ride on the senior user's goodwill as a means of getting his foot in the door with prospective customers.(fn6)

But this explanation begs the question. The law does not prohibit free riding across the board.(fn7) What makes this brand of free riding wrong? Is it always wrong? Courts have not provided coherent answers to these questions.

The root of the problem is the doctrine's focus on trademark owner goodwill and not on consumer interests. Just because a mark's goodwill may attract a potential customer to further examine a product does not guarantee a sale to the trademark owner. If an initially confused consumer (who is not confused at the time of purchase) mistakenly looks elsewhere at first but then chooses the alternative product, notwithstanding any "bait and switch" tactics, the trademark owner's goodwill could not have been that strong. Indeed, the consumer may have been drawn to the trademark because it represented a category of product and not a particular product.(fn8) If so, the "confusion" may have expanded the consumer's awareness of alternatives, thus promoting competition and enabling the purchase of a preferable product. In such cases, policing initial interest confusion potentially harms consumers by cutting them off from a source of information about the broader market.

But permitting initial interest confusion may also harm consumers. The class of initially confused consumers includes those who are specifically seeking a particular brand to the exclusion of others. They must expend extra effort to determine which product is which, and to find their preferred choice. For these consumers, initial interest confusion impedes the trademark's function of reducing consumer search costs. This perspective suggests that a balancing is possible: Courts should police initial interest confusion only when it produces greater harms than benefits.

Focusing on goodwill does not provide any basis for balancing. Instead, courts should consider the costs of dispelling initial confusion to determine whether initial confusion causes a trademark injury. Under this consumer-based search cost analysis, misappropriated "goodwill" is not a factor except insofar as it represents the diversion of a consumer's search "budget" from an intended recipient. Phrased another way, there is point-of-sale confusion of sorts, but the "product" is further exploration of a potential purchase, which is paid for with the consumer's time and opportunity costs.(fn9) Consistent with this view, several courts have refused to apply initial interest confusion on the grounds that fleeting or de minimus initial confusion does not work a trademark injury,(fn10) although some other courts have taken a more aggressive approach."(fn11)

Whatever balance the courts have struck in initial interest confusion cases has been upset by cases involving Internet domain names and metatags. When the "confusion" results in a web surfer being directed to the wrong web site, correction costs are negligible. Accordingly, one would expect judicial reluctance to apply initial interest confusion. The opposite is true. Claims of appropriated goodwill and misplaced analogies to the "real" world persuaded several circuit courts to apply a robust interpretation of initial interest confusion to cyberspace.(fn12) Their holdings ignore, and at times reject, any consideration of search costs. The perverse, if predictable, effect has been to invite broader claims of initial interest confusion in cases involving the brick and mortar world. Having welcomed broad claims in cyberspace, courts find reversing course difficult, especially without a clear doctrinal basis for doing so.(fn13) Initial interest confusion cases involving the Internet may therefore have ramifications extending far beyond the Internet context.

This article critiques the development and application of initial interest confusion and argues for a doctrine based on consumer search costs rather than a trademark owner's goodwill. Part I traces the origin of initial interest confusion and presents a theory, based on minimizing search costs, of when the concept should be applied. It then examines the application of initial interest confusion in light of the courts' uncertainty as to the purpose of the doctrine. Part II describes the doctrinal difficulties caused by the uncritical adoption of initial interest confusion to cases involving the Internet. These problems can be resolved by focusing on search costs and the costs of correcting consumer confusion.

I. Initial Interest Confusion in the Brick and Mortar World

A. The Goals of Trademark Law and the Lanham Act

Trademark law has multiple purposes.(fn14) Many of them can be grouped under the general goal of consumer protection, though this goal subsumes benefits to merchants as well as consumers.(fn15) By identifying products by source or brand, trademarks lower search costs and protect consumer expectations.(fn16) Lowering search costs benefits producers by enabling them to charge higher prices and compete on the basis of product quality.(fn17)

Trademark law also protects the goodwill symbolized by a trademark-the positive associations a customer has with a specific...