Ops and Robbers? Amazon's trade mark travails in the US

Whatever your feelings about the doctrine of initial interest confusion [Mr Justice Arnold was in favour here and here; "no, no, no" said the Court of Appeal for England and Wales], it's a fascinating doctrine that is of great potential value to trade mark-owning litigants in the United States, where it is still alive and kicking.  In this guest note, Katfriend and occasional contributor Kevin Winters guides readers through a recent decision in which an action brought against Amazon, originally dismissed with a summary judgment, has been effectively brought back to life on appeal, in part on account of initial interest-based arguments. Kevin explains as follows:

Did Amazon’s search results infringe MTM’s trade mark?  Yes, said the Court of Appeals for the Ninth Circuit in Multi Time Machine, Inc., v Amazon.com, Inc.; Amazon Services, LLC (No.2:11-cv-09076-DDP-MAN, here) when it reversed the District Court’s summary judgment order in Amazon’s favour. 

The background 

Multi Time Machine (MTM) -- which makes and sells watches under a variety of brands including MTM, MTM Special Ops and MTM Military OPS -- holds the US registered trade mark for MTM SPECIAL OPS for timepieces.  It sells its timepieces direct to customers, using various distributors and retailers, the target market being men between 22-55 years of age who are attracted to “…rugged, military-style outdoor products.”  Seeking to preserve its image as an exclusive watch brand, MTM did not sell its watches to Amazon for resale.  

Amazon claims to offer “Earth’s Biggest Selection of Products”, including those of MTM’s competitors.  Amazon users who search for “MTM Special Ops” reach a screen displaying the phrase “MTM Special Ops” in the search field, accompanied by “MTM Special Ops” with quotation marks immediately below it and again after the words “Related Searches”.  Below these three restatements of MTM’s trade mark the screen provides search results, including watches provided by MTM’s competitors which are listed by name.  Customers must then proceed to the “product detail” page by clicking on one of the search results, where they will see the product’s brand name and title, accompanied by the brand name.  The top of the product detail page will still display the customer’s initial inquiry “MTM Special Ops” in the search field.  There is no indication on either page that Amazon does not carry MTM’s products.  Amazon’s competitors however, Buy.com and Overstock.com, make it clear that no search results match the “MTM Special Ops” search; nor do they direct the user to a page with both MTM’s trade mark “MTM Special Ops” repeatedly at the top of the page and competitor watches below.  Their pages display the search query playback, stating explicitly that no search results for “MTM Special Ops” were found, and list competing products.   

Amazon lists MTM’s competitors’ products in its search results partly because of its search algorithm, which responds to customers’ behaviour using a Behaviour Based Search technology (BBS).  The BBS uses data about what customers view and purchase after searching for certain products.  Amazon does not programme the terms; the technology only responds to customer behaviour.  Where enough customers search for a certain keyword, and then view or purchase another product, even if both products are not obviously related, future customers who search for one product may receive results including competing products.  The results a customer receives include matches based on a search of terms on Amazon’s pages.  The fact that a particular result shows up because of BBS or a traditional search of matching terms is not evident from the matches, and the relevant products (based on search terms) and ‘recommended products’ (based on BBS) are mixed together in the search results.   

MTM sued Amazon for trade mark infringement, seeking injunctive relief and damages.  The District Court granted summary judgment to Amazon on its motion. MTM appealed. 

The decision

The Court of Appeals, whose opinion was given by Judge Bea, allowed the appeal against the summary judgment, finding that a jury could find that Amazon had created a likelihood of confusion between MTM’s products and those of its competitors. The Court first recounted the relevant provision of the Lanham Act: 

“…a defendant infringes a trademark when the defendant uses the mark in commerce in a manner likely to cause confusion as to a good’s source.  Trademark infringement also occurs when the trademark’s use in commerce is likely to cause confusion as to the affiliation, association, or approval of the trademark holder with the trademark user…Put another way, a defendant who creates likelihood of confusion by using another’s mark has infringed the mark.” 

The Court then recited the legal requirements for a claim for trade mark infringement on the basis of initial interest confusion:

“Initial interest confusion occurs not where a customer is confused about the source of a product at the time of purchase, but earlier in the shopping process, if “customer confusion…creates initial interest in a competitor’s product.”  Even if that confusion is dispelled before an actual sale occurs, initial interest confusion still constitutes trademark infringement because it “impermissibly capitalizes on the goodwill associated with a mark and is therefore actionable trademark infringement.”

MTM’s appeal rested on two grounds: 

1. The likelihood of confusion

The court first had regard for the labelling of the watches on Amazon’s website given the observation in Network Automation v Advanced System Concepts (available here) that, when dealing with internet commerce, while a likelihood of confusion could be shown where “consumers saw banner advertisements that were ‘confusingly labelled or not labelled at all’…clear labelling ‘might eliminate the likelihood of confusion.’” 

The Court of Appeal disagreed with the District Court's finding that there was no evidence that Amazon users were likely to be confused as to the source that made the competing goods: a jury could infer that users who are confused by the search results are confused why MTM products had not been listed.  This was exacerbated by the fact that Amazon did not attempt to prevent any confusion by making it clear to people searching for “MTM Special Ops” products that it did not sell them.  

Moreover, said the Court, a jury could infer that users who are confused by the search results would wonder if MTM had been acquired by a competitor, or was affiliated or associated with it -- which was particularly true for a brand like MTM, since luxury goods brands are often produced by lower-priced, better known brands.  Amazon’s use of MTM's mark would thus cause initial interest confusion, attracting potential customers’ attention to buy the infringing goods because of the MTMs reputation.  In any event, noted the Court, while MTM's competitor’s products on the Amazon search results page were clearly labelled, the clarity of Amazon’s search results page remained open to dispute. 

The Court then considered the eight Sleekcraft factors in attempting to ascertain the likelihood of confusion, but restricted itself to  the five which it deemed relevant to the case: 

1. Strength of the marks

The two relevant types of trade mark strength here were commercial and conceptual, and the Court considered the conceptual strength of MTM's mark as no evidence was given as to its commercial strength.  Said the Court: “A mark’s conceptual strength depends largely on the obviousness of its connection to the good…to which it refers.  The less obvious the connection, the stronger the mark and vice versa.”  

The phrase “MTM Special Ops” was described as requiring “a mental leap from the mark to the product” because the phrase did not expressly refer to watches. The evocation of military forces suggested goods that were as likely to be protective gear or weapons as watches: a jury could find the mark (i) suggestive and strong as a concept because it did not explicitly refer to watches, or (ii) descriptive owing to the watches being made in a military style.  Either way, this was a question of fact to be determined by a jury.  

2. Similarity of the goods

MTM and Amazon were both found to sell specialised, military-style watches. Given that MTM watches were not displayed anywhere on Amazon’s website, it was an open question whether customers would believe that the options on Amazon’s page were clearly marked as having no association with, or approval by MTM, and whether they would be confused.  A jury could find that this weighed in favour of finding a likelihood of confusion.  Further, Amazon did not label search results showing MTMs competitors’ products as BBS results rather than as keyword searches and the similarity of the watches meant that Amazon customers who looked for “MTM Special Ops” and then explored watches made by competitors, even without any confusion, would still have been subject to “confusion, not mere diversion.”  While the similarity of the watches weighed in MTM's favour, it was for a jury to determine precisely how much. 

3. Defendant’s intent

MTM submitted evidence that Amazon vendors and customers had complained to Amazon because they could not understand why they received some non-responsive search results when they searched for products that Amazon did not carry; Amazon employees had failed to address these complaints by explaining to customers how the BBS technology worked.  Said the Court, from the evidence that “Earth's most customer-centric company” did nothing to address the complaints, a jury might infer that Amazon intended to confuse customers.  The Court did not comment on how important intent was, reserving this duty for a jury. 

4. Evidence of actual confusion

MTM had offered no colourable evidence of actual confusion, though its president testified that he was aware of instances of it.  This was not enough to give rise to a genuine issue of fact.  Amazon on the other hand provided the court with evidence that attempted to show that no customers were confused, because those who searched for one product were more likely to buy it than customers who searched for another product.  The Court agreed with Amazon here, but added that a jury could view Amazon’s evidence as flawed, given that a user researching watches may suffer initial confusion on the availability of MTM watches online and so not purchase a competitors watch on the same day.  Further, some users did search for “MTM Special Ops” and buy competitors' watches on the same day, which a jury may find to be probative of some confusion.  

5. Degree of care

The court acknowledged that where expensive goods are concerned, purchasers are expected to exercise greater care, while confusion may still be possible.   MTMs watches were priced between several hundred to two thousand dollars.  The court agreed with the District court that customers could be presumed to use a high degree of care in buying that kind of watch.  However the court pointed out that other factors gave rise to issues of fact, and that a jury may find that evidence from Amazon that “same day sales” are high for competitors’ watches could be indicative that some buyers are not as careful as others.  Therefore a jury could accord little weight with the price of watches compared with other factors.  Once again however, the court stated that this factor’s importance was for a jury to decide.  

2. Whether the use of MTM's trade mark was a “use in commerce”

Said the Court, this question had already been answered in the affirmative in Network Automation v Advanced System Concepts (available here).  Amazon argued that the user-generated search term “MTM Special Ops” was not a use in commerce under the Lanham Act, but the Court said that the customer-generated use of a trade mark in the retail search context was a use in commerce: Amazon’s purpose was no less commercial in selling watches than the sale of advertising space in Network Automation.  

The Court of Appeals, while not convinced that MTM would be able to prove a likelihood of initial interest confusion, found there were sufficient issues of material fact to be determined to resolve the matter.  It therefore reversed the decision of the District Court, granting MTM trial on its claims of trade mark infringement. 

The dissenting judgment 

Judge Silverman was not convinced that Amazon’s search results were likely to result in confusion between MTM's and its competitor’s products. He acknowledged the use of the Sleekcraft criteria in analysing the likelihood of confusion but pointed out that they should not be used as a checklist; and in evaluating infringement claims involving internet search engines, the criterion of labelling is particularly important (and also fell outside the Sleekcraft list).   He added:

“Sleekcraft aside, the ultimate test for determining the likelihood of confusion is whether a “reasonably prudent consumer” in the marketplace is likely to be confused as to the origin of the goods”  Our case can be resolved simply by an evaluation of the webpage at issue and the relevant consumer…”

The outcome of the case would thus depend on the answers to two questions: 

1. Who is the relevant reasonable consumer?

Here the goods at issue were expensive, some selling for several hundred dollars, and the relevant customer “is a reasonable prudent consumer accustomed to shopping online.” 

2. What would he reasonably believe, based on what he saw on the screen?

Judge Silverman recounted MTM's assertion that the labelling and appearance of the products on Amazon’s webpage was the most important factor.  This was attributed to the fact that the court had noted in Playboy Enters v Netscape Communs. Corp (available here), that clear labelling can eliminate the likelihood of confusion in cases concerning internet search terms.  In this case, the products were clearly labelled by Amazon and a customer who searches for “mtm special ops” on its website would be presented with several different products -- all of which were clearly labelled with the product’s name and manufacturing details in bright, bold lettering and accompanied by a photograph.  Given Amazon’s actions:

“…it is simply bizarre to suppose that a reasonably prudent consumer accustomed to online shopping would be confused about the source of the goods.”

Judge Silverman also dealt directly with MTM’s argument that initial interest confusion could occur because Amazon listed the search term “mtm special ops” three times at the top of the search page.  He found confusion unlikely: no product on Amazon’s search results page contained the word “MTM”, phrase “Special Ops” or “MTM Special Ops”, and some of the products were not even watches, where a book was displayed.  He added:

“It is hard to fathom how a reasonably prudent consumer accustomed to shopping online would view Amazon’s search results page and conclude that the products offered are MTM products.  Some of the products are not even watches!  And the watches that are offered for sale are clearly labelled as being manufactured by Luminox, Chase-Durer, TAWATEC or Modus – not by MTM.  It is possible that some dolt somewhere might be confused by the search results.  But, “[u]nreasonable, imprudent and inexperienced web-shoppers are not relevant.”

He saw no evidence that anyone had ever been confused as to whether the results displayed on Amazon’s search results were connected with MTM.  To establish a likelihood of confusion, MTM would have to demonstrate that confusion was likely and not simply possible, and there was no merit in MTM’s argument that Amazon should change its search results page: it was in his view unambiguous.  

What next?

It is difficult not to have some sympathy for the view of the dissenting judge, since MTM would be facing a steep mountain to climb in proving a case based on actual confusion -- and the outcome of even a claim based on initial interest confusion may be hard to press home. However, even an apparently weak case will go to trial if it is at least an arguable one, and MTM cannot be denied its right to plead its cause.  

The observations about the Sleekcraft criteria are interesting: has the time come to revise or re-prioritise the Sleekcraft list, which dates back to 1979, and tailor it to internet-based trade mark infringements?

As an aside, Amazon has been involved in some other fascinating trade mark litigation of late.  Its recent spat with Apple (here) was halted when Apple dropped its claim that Amazon had infringed its trade mark through use of the term ‘Amazon Appstore’.  

IPKat blogmeister Jeremy adds: sometimes the difference between a weak case being summarily dismissed and its being allowed to go to trial if the difference between a plaintiff getting nothing at all and its being allowed to obtain some sort of face-saving settlement as a quid pro quo for going no further -- but that is unlikely in a scenario such as this, where Amazon is conscious that only a clear court ruling in its favour will provide the sort of precedental guidelines that will either ward off future similar cases or enable it to modify its business model so as to avoid future liability.

Merpel adds, there's a good bet that, when the current incumbents of the Court of Appeal for England and Wales have all retired and have been replaced by Lord Justice Arnold, initial interest confusion will make its welcome return.