MTM sells “military style” watches under the marks MTM Special Ops and MTM Military Ops through its website and authorized distributors. It doesn’t sell or allow its distributors to sell watches on Amazon. Amazon’s search function nonetheless provides results for a search on “mtm watch” or similar queries, offering other product listings—here, watches made by MTM competitors.
When a user searches on Amazon, the user’s search term will appear twice on the search results page: once in the search query box and directly below, in what Amazon calls the “breadcrumb.” The breadcrumb “displays the original query in quotation marks to provide a trail for the consumer to follow back to the original search.” Below that, Amazon provides “Related Searches,” here including “mtm special ops watch.” A gray bar separates these three instances from the product listings. For the “mtm special ops” search, one of the “Sponsored Links” results included a link for “Tactical Watches By MTM,” with the description “MTM Tactical Watches Worn By Military, Police, Sportsmen,” and another link to MTM's website.
You can’t buy from the search results page, only from a product detail page, which includes “a large image of the product, a hyperlink identifying the brand of the product, and a title identifying the product in larger font.” Thus, the Luminox search results for a “mtm special ops” search include large brand identifiers. On that product detail page, your original search query—here “mtm special ops”—still appears in the search query box, and there are also links suggested uner “Customers Viewing This Page May Be Interested in These Sponsored Links,” which in this case displayed sponsored hyperlinks labeled “MTM Watches,” with a link to Yahoo, and “Military Watches Sale,” which went to some sort of shopping site.
The court found this situation to squarely present the question posed by Judge Berzon in her Playboy concurrence:
I walk into Macy's and ask for the Calvin Klein section and am directed upstairs to the second floor. Once I get to the second floor, on my way to the Calvin Klein section, I notice a more prominently displayed line of Charter Club clothes, Macy's own brand, designed to appeal to the same people attracted by the style of Calvin Klein's latest line of clothes. Let's say I get diverted from my goal of reaching the Calvin Klein section, the Charter Club stuff looks good enough to me, and I purchase some Charter Club shirts instead. Has Charter Club or Macy's infringed Calvin Klein's trademark, simply by having another product more prominently displayed before one reaches the Klein line? Certainly not....
... If I went to Macy's website and did a search for a Calvin Klein shirt, would Macy's violate Calvin Klein's trademark if it responded (as does Amazon.com, for example) with the requested shirt and pictures of other shirts I might like to consider as well? I very much doubt it.
MTM’s argument was that, to avoid confusing consumers, Amazon was “obliged to inform the consumer that Amazon does not carry any products with that brand before offering products from other brands.” Amazon responded that, so long as it labeled its results clearly to indicate their brands, consumers get what they want without infringement. Amazon wins.
The court began with Network Automation, Inc. v. Advanced Systems Concepts, Inc., 638 F.3d 1137 (9th Cir. 2011). Because it found no likelihood of confusion, it didn’t reach the question of whether Amazon’s internal use of a term as part of its behavior-based search technologies constituted “use in commerce,” though given the breadth of that term the court indicated that it would likely have found use in commerce.
Though likely confusion is fact-intensive, and rarely appropriate for summary judgment, it was so here. MTM argued that Amazon was engaged in passing off, like serving customers Pepsi when they asked for Coke. In Coca-Cola Co. v. Overland, Inc., 692 F.2d 1250, 1252 (9th Cir. 1982), the court found that menus and posted signs indicating that Pepsi-Cola was the only beverage served in the restaurant were not “sufficiently conspicuous” to avoid liability. But that’s a question about confusion, and Sleekcraft still provides the basic test for confusion. Anyway, this case was different:
It is akin to the consumer asking for a Coca–Cola and receiving a tray with unopened, labeled, authentic cans of Pepsi–Cola, RC Cola, Blue Sky Cola, Dr. Pepper, and Sprecher Root Beer, and a copy of Coca Kola: The Baddest Chick, by Nisa Santiago. This is a substitution, but given the context it is not infringing because it is not likely to confuse.
In the internet context, the confusion factors should be applied flexibly. Network Automation said that the most important factors in a keyword search case were (1) strength of the mark, (2) actual confusion, (3) type of goods/degree of care, and (4) labeling and appearance of the ads and the surrounding context on the screen displaying the results page. This was a good starting point.
Among other things, the similarity of goods (note that this isn’t the similarity of the parties’ goods, but we’ve all agreed to ignore that usual requirement) was less important if ads are clearly labeled or consumers exercise a high degree of care. A consumer may well type in a term identical to MTM’s mark, but that too only matters if the consumer ends up confused. Though both Amazon and MTM sell watches, “this is misleading only if the consumer is confused, not if the consumer simply has clearly marked options.” The fact that both Amazon and MTM sell watches on the internet is “too commonplace” to matter in the likelihood of confusion analysis. (There are some interesting assumptions about consumer behavior encoded here—we could’ve decided that convergent marketing channels just increased the risk of confusion generally and thus always weighed against a defendant, but instead we’ve expected consumers to adapt to the internet as a marketing channel—paging Fred Yen.) Intent also needed to be considered in the context of the clarity of labeling.
Looking at the strength of the marks at issue, the court didn’t find them especially strong. “MTM Special Ops” and “Military Ops” referred to watches designed for members of the armed forces involved in special military operations. MTM advertised “Special Ops Watches worn by Special Ops and Special Forces worldwide,” though to the PTO MTM’s attorney said, “The term Special Ops is nothing more tha[n] a suggestive reference to military type watches.” (MTM claimed it had two other marks at issue—“American Watch” and “Pro Ops.” American Watch was used for a company that sells promotional watches that can be inscribed with company names; no watches were branded American Watch. While those two, and Military Ops, were “arguably distinctive enough” to be protected, they were also phrases that could be used to search for other products unconnected with MTM, and weaker than “MTM Special Ops”; thus the court’s analysis applied equally to them.)
The marks were not conceptually strong—at best suggestive and more likely descriptive. The addition of “MTM” might make them protectable, but didn’t neutralize their descriptive connection to the product. Thus, this factor favored Amazon. (This discussion makes no sense in the context of product search. Weak or strong, the question is whether it’s clear you’re being offered an alternative. Compare a search for Xerox that offers Canon products as alternatives.)
There wasn’t admissible evidence about MTM’s market share, rendering evidence of its sales and ad expenditures of little significance; MTM didn’t show evidence of brand recognition. This factor was neutral because, though it was MTM’s burden to show likely confusion, it didn’t need to prevail on each Sleekcraft factor and neither party presented evidence of commercial strength. Overall, given the conceptual weakness and absence of other evidence, strength favored Amazon.
Amazon argued that there was evidence of a lack of confusion, coming from its retained data about how often a consumer's search for “mtm special ops” or “mtm special ops watch” results in the consumer placing a product in a shopping cart or in a purchase. Amazon compared the same data for queries for “luminox” or “luminox watch.” Since consumers were 21 times more likely to buy a product after searching for Luminox than after searching for MTM Special Ops, Amazon argued that there was no confusion: if they were confused, one would expect a substantial conversion of MTM searchers into Luminox buyers. MTM argued that Amazon’s data registered only sales/selections made on the same day of the search, whereas consumers might put a product into a cart but buy it later. Amazon rejoined, persuasively, that there was no reason to think that such behavior differed for Luminox v. MTM searchers. The court agreed that this was evidence of no actual confusion.
In addition, queries for “mtm” were much more common than for “mtm special ops,” and the average price of units sold based on a “mtm” search was dramatically lower than the price of MTM’s competitors’ watches. Thus, this wasn’t evidence of confusion. (The court didn’t say so, but I think what this means is that there are other kinds of “MTM” products—ammo cans turned up high in my Amazon search—that aren’t watches and that, for all that appears, aren’t sold by plaintiff.)
MTM’s president testified that there was actual confusion, but couldn’t present specific instances or records; his testimony was too vague to count.
MTM’s watches start in the several-hundred-dollar range and go up; its competitors’ products are likewise relatively expensive, triggering a high degree of care. MTM argued that watches in general could be pretty cheap, but not so with the least expensive watch resulting from a “mtm special ops” search, which was $145 (the first five results are more expensive than that). “[T]he relatively high price of the goods in question, combined with the increased degree of care used in Internet purchases, mean that consumers are presumed to use a high degree of care in such purchases.”
As for the key question of labeling and context, this too favored Amazon. MTM offered an expert report concluding that the search results were “ambiguous, misleading, and confusing.” MTM’s expert conducted no study. At most, he could say that consumers might be confused about how the site functions, but not about the source of the products. “A consumer could, for instance, puzzle over why the search query ‘mtm special ops produced a results page listing ten watches but none of them with the MTM brand without also being confused as to the source of the watches presented on the results page.” While not ruling out the possibility that a survey could show confusion in a case like this, the court found that there was no such evidence. The court also noted that confusion might result if the products resulting from the search had substantially similar marks, which they didn’t here.
No confusion; no liability.