Proving a Likelihood of Confusion Remains an Uphill Battle for Trademark Owners in Keyword Advertising Cases

How confused are you when you search terms and see paid advertising generated as the result of the keywords you inserted? An increasing number of cases show a likelihood of confusion is a difficult element to meet by plaintiff trademark owners.

People use search engines to weed through millions of products that are available on the Internet. Search engines have thus become an important marketing tool to direct web traffic to businesses and a major source of revenue for search engine providers, such as Google.  Search engine providers sell search terms or keywords to advertisers, allowing advertisement links to appear on the results page when the purchased keywords are searched. Google generated over 50 billions in revenue for its advertising program in 2013.

In 2004, Google updated its Adwords advertising policy permitting advertisers to purchase trademarks as search terms with or without the trademark owners’ approval. Thus, if a soft drink company buys a keyword “coca cola,” anytime a user searches the term “coca cola,” Google would display both the paid advertising listings that would show the soft drink advertiser’s link and the free organic search results. This controversial policy resulted in more than 20 lawsuits filed against Google.

Trademark Infringement under the Lanham Act

Under the Lanham Act, 15 USC 1114(a)(1), a trademark owner who claims an infringement must establish two elements: (1) the defendant used the registered mark in commerce; and (2) the use is likely to cause consumer confusion.

Use in Commerce

Typically, the first prong of the test must be met in order for the court to consider the second prong. The majority of federal circuit courts have held that the keyword advertising constitutes a “use in commerce” under the Lanham Act.

Prior to 2009, district courts in the Second Circuit ruled that a use of trademarked keywords was not a “use” under the Lanham Act because the use was “strictly internal” and was not visible to the public. See e.g. S & L Vitamins, Inc. v. Australian Gold, Inc. and Merck & Co., Inc. v. Mediplan Health Consulting, Inc.

In 2009, the Second Circuit decided Rescuecom Corp. v. Google Inc., finding that the district courts “over-read” its prior decision in 1-800 Contacts, Inc. v. WhenU. Com, Inc.. The Second Circuit held that the search engine keyword advertising program, i.e. Google Adwords, constitutes a use in commerce under the Lanham Act because unlike the defendant’s program in 1-800, Google made the trademarked keywords available for purchase and displayed them in the search results. It further held that use of the trademark in an internal software program need not be immune from infringement.

With the Second Circuit joining the majority, most defendants can no longer make the argument that a search engine online advertising program is not a use of commerce. Following this decision, cases against Google and purchasers of trademarked keyword have increased dramatically.

Likelihood of Confusion

Showing likelihood of confusion, however, remains a challenge for the trademark owners.

Courts determine likelihood of confusion by weighing several factors. The determination of the likelihood of confusion is dynamically evolving and courts have arrived at different interpretations of what factors to apply and how much weight should be given to each factor. The federal courts use these factors as a guide and apply them on a case-by-case basis with some variation. The common factors considered by courts are: (1) strength of the mark; (2) degree of similarity of the goods or services; (3) degree of similarity of the marks; (4) evidence of actual confusion; (5) use of the Internet as marketing channels; (6) degree of care exercised by buyers; (7) the alleged infringer’s intent in selecting the mark; and (8) likelihood of product lines expansion. See e.g. AMF Inc. v. Sleekcraft Boats, Gibson Guitar Corp. v. Paul Reed Smith Guitars, LP, Polaroid Corp. v. Polarad Electronics Corp., and Interpace Corp. v. Lapp, Inc..

Adding to the mud, many federal circuits in early cases adopted the initial interest confusion doctrine – holding that “initial interest confusion” is enough to meet the likelihood of confusion element of the test. Initial interest confusion occurs when an alleged infringer uses another’s trademark to create initial customer interest in its own competing goods or services. This doctrine could apply even if, before purchasing the item, the customer is no longer confused because the customer realizes the source of the goods or services and understands that they are not of the trademark holders. The federal circuits are split on whether the unauthorized act of diverting the potential customers from the trademark owner’s website to the competitor’s itself is enough to satisfy the initial interest confusion doctrine. Many courts suggested that initial interest confusion is present if the trademark is clearly mentioned in the competitor’s advertisement.

Showing the presence of initial interest confusion in keyword advertising cases is difficult because the trademark is typically referenced in the internal search engine program and does not usually itself appear in the advertisement. In determining the likelihood of initial interest confusion, courts have looked at factors that are almost identical to the factors of those of the likelihood of confusion – only fewer. The common main factors for the initial interest confusion are: (1) marks similarity, (2) products similarity, (3) alleged infringer’s intent, and (4) evidence of actual confusion. The initial interest confusion theory is arguably a “subset” of likelihood of confusion that lowers the standard of proof for showing actual confusion.

In 2011, the Ninth Circuit updated its view in the matter in Network Automation, Inc. v. Advanced System Concepts, Inc., making it more difficult for trademark holders to show likelihood of confusion. The Ninth Circuit did not expressly reject the initial interest confusion doctrine, but instead “merge[d]” it into the eight traditional Sleekcraft factors for the likelihood of confusion. The Ninth Circuit noted the four most important factors out the eight factors are: “(1) the strength of the mark; (2) the evidence of actual confusion; (3) the type of goods and degree of care likely to be exercised by the purchaser; and (4) the labeling and appearance of the advertisements and the surrounding context on the screen displaying the results page.”

Following the Ninth Circuit’s decision in Network Automation, many of the subsequent district courts decisions found the Ninth Circuit’s approach persuasive. See e.g. Morningware, Inc. v. Hearthware Home Prods. and Collegesource, Inc. v. Academyone, Inc., 2012 WL 5269213 (E.D. Pa. Oct. 25, 2012).

However, the initial interest confusion has not completely vanished yet. In the 2013 appellate decision in 1-800 Contacts, the Tenth Circuit applied the initial interest confusion doctrine. It denied the plaintiff’s claim of trademark infringement, stressing that the consumers were able to differentiate between the search results and the paid advertising links.

Despite the different tests the courts are using to determine likelihood of confusion, the author is only aware of one case where a trademark owner was successful in proving liability against a competitor for trademark infringement as a result of using a trademarked keyword through Google Adwords. In 2011, just a few months prior to the Ninth Circuit’s decision in Network Automation, the Central District of California court in Binder v. Disability Group, Inc. held in favor of the plaintiff, finding that there was a strong likelihood of confusion. It also found damages in lost profits against the competitor.

After Network Automation, trademark owners do not seem to have a better luck in the likelihood of confusion issue. In 2012, there was one case that initially brought optimism for trademark owners. The Fourth Circuit in Rosetta Stone Ltd. v. Google, Inc. vacated the district court’s summary judgment for the defendant, holding that genuine issues of material fact were present when determining the defendant’s intent, evidence of actual confusion, and the consuming public’s sophistication, without reaching the likelihood of confusion issue. Unfortunately, the parties reached a settlement before the district court was able to decide on the issues on remand.

Internationally, there has been a growing acceptance of keyword advertising services. In 2013, the Australian High Court ruled in favor of Google, holding that Google’s display of the sponsored links generated by its Adwords program did not mislead consumers. Additionally, since 2010, the European Court of Justice had permitted Google’s keyword advertising services in its decisions. While the door is not completely shut for trademark owners in keyword advertising cases, proving likelihood of confusion appears to have an increasingly slim likelihood of success.