Search engines have become essential tools for consumers as they look for products and make purchases online. On July 6, 2015, the US Court of Appeals for the Ninth Circuit reversed a grant of summary judgment in a trademark lawsuit brought against Amazon.com (Amazon), potentially signaling greater rights for trademark owners in the search context. Multi Time Machine, Inc. v. Amazon.com, Inc. et al.1
The Multi Time Machine Case
Multi Time Machine, Inc. (MTM) manufactures upscale, military-style wristwatches and owns a registered trademark for MTM SPECIAL OPS. MTM sued online retailer Amazon based on the way in which Amazon displayed search results when a consumer looked for the MTM SPECIAL OPS mark on the Amazon website. Because Amazon did not sell MTM's watches, its "Behavior-Based Search" algorithm (BBS) instead populated the page with similar products, including military-style watches manufactured by MTM competitors like Luminox and Invicta. On this results page, Amazon displayed the MTM SPECIAL OPS mark three times: (1) in the search bar; (2) in a description stating "9 results for 'mtm special ops'"; and (3) in a "related searches" list.
In reversing a grant of summary judgment in favor of Amazon, the Ninth Circuit tailored its likelihood of confusion analysis applicable in trademark infringement cases to the search context. Here, the Court employed a modification of the Ninth Circuit's traditional likelihood of confusion factors articulated inAMF Inc. v. Sleekcraft Boats, 599 F.2d 341, 348-349 (9th Cir. 1979).
The Court stated that three factors in the traditional Sleekcraft analysis are not apt in the context of internet search-based trademark actions: similarity of marks; marketing channels; and likelihood of expansion.2
Of the five remaining factors, the Court found that three favored MTM and were sufficient to present the issue of likelihood of confusion to the jury: strength of the mark; similarity of goods; and defendant's intent. The intent factor weighed against Amazon: Consumers had complained to Amazon regarding non-responsive search results, but Amazon had taken no steps to alleviate the consumers' confusion.
The Court also held that, under its precedents in Network Automation, Inc. v. Advanced Sys. Concepts, Inc., 638 F.3d 1137, 1145 (9th Cir. 2011) and Playboy Enterprises, Inc. v. Netscape Commc'ns Corp., 354 F.3d 1020, 1024 (9th Cir. 2004), an additional factor should be added to the traditional test: "labeling." The Ninth Circuit found that this factor weighed in favor of a finding of "initial interest confusion", i.e., confusion that occurs "not where a customer is confused about the source of a product at the time of purchase, but earlier in the shopping process."3 This was because: (1) unlike other Internet retailers, Amazon did not explain in its search results that it did not stock MTM products; and (2) the search results page repeatedly displayed MTM's mark.4
Thus, the Ninth Circuit reversed and remanded the grant of summary judgment in Amazon's favor.
What to Look for on Remand in Multi Time Machine
In remanding, the Ninth Circuit wrote "[w]e are by no means certain that MTM will be able to prove likelihood of confusion … but we are confident the matter can be determined only by resolving genuine issues of material fact."5 Those questions of fact may well focus on the two factors that the circuit court identified as favoring Amazon: (1) MTM's failure to introduce any evidence of actual confusion; and (2) the high degree of care presumably exercised by consumers of expensive watches.6
Perhaps most likely, however, is that the fight will continue over the clarity of Amazon's labeling of its search results. The Ninth Circuit's decision certainly does not prevent Amazon from introducing additional evidence that consumers searching for MTM SPECIAL OPS are not likely to be confused upon finding competitors' products.
Implications of Multi Time Machine
Multi Time Machineis far from the first trademark dispute based on search results, and it may shed light on the scope of trademark protection in analogous cases, such as those involving keyword advertising.
A common subject of search-based trademark disputes involves the practice of keyword advertising. Companies can purchase the right to display their advertisements when a consumer's search incorporates certain words or phrases, i.e., keywords, in the search bar. Numerous companies have purchased keywords based on their competitors' marks, so that when a consumer types the competitor's mark, a sponsored ad for the purchaser will appear alongside the native search results. Trademark owners have filed lawsuits against both the search companies and the companies purchasing the keyword advertisements for trademark infringement, with mixed results.
Trademark owners initially had success challenging keyword purchasers.7 In these cases, trademark owners successfully argued that keyword advertising led to a form of initial interest confusion even where the sponsored advertisement did not include the trademark owner's mark and was not otherwise confusing.
Victories for trademark owners in this area decreased, however, after the Ninth Circuit's 2011 decision inNetwork Automation. There, the Ninth Circuit held that a competitor who purchased keyword advertisements based on another's marks could not be held liable for trademark infringement based on his mere diversion of potential consumers; the trademark owner had to show consumers were actually likely to be confused by the sponsored search result.8 The Ninth Circuit also held that such a finding of likelihood of confusion was less probable where the search engine provider separated and labeled keyword advertisements as "sponsored links."9
Throughout the more recent cases, it has remained unclear what standard courts should use to examine the likelihood that sponsored search results will confuse consumers. Multi Time Machine is the latest in a series of decisions by the Ninth Circuit guiding lower courts on the question of consumer confusion in the keyword context. Here, as described above, the Ninth Circuit rejected three of the traditional likelihood of confusion factors as inapposite. The appellate court instead reaffirmed its focus, announced inNetwork Automation, on the "labeling" of search results; i.e. whether the search results appear in such a way as to reduce the likelihood of confusion, as an important likelihood of confusion factor to be assessed in search engine cases.
Perhaps as a result of this lack of clarity, and the recent difficulty trademark owners have faced in the courts, at least one recent report suggested that there are no currently pending keyword lawsuits.10 It remains to be seen whether the MTM decision - specifically a positive result for MTM on remand - will cause other trademark owners to reconsider the probability of success in keyword cases.
Multi Time Machineis potentially important for another reason. It may expand the scope of search-related activity that could create trademark liability. Unlike the keyword cases discussed above, this case involved the result of Amazon's proprietary BBS algorithm. And, unlike keyword cases, in which a rival company purchases advertisements tied to the plaintiff's marks, the only use of MTM's mark was by the consumer searching for MTM's product. In other words, Multi Time Machine denied summary judgment for Amazon concerning what might be considered, in the search engine context, native results - not search results intentionally generated by Amazon. Many online retailers use search engines to guide consumers to the products they sell; Multi Time Machine raises the specter of trademark liability for such retailers where their results do not clearly indicate to the consumer whether the retailer is selling the trademarked item the consumer has searched for or only comparable products. Only time will tell whether this potential increased scope of liability - at least in the Ninth Circuit - leads to more challenges in this context.