The law on initial interest confusion: Interestingly confusing
Oh joy, 9th Cir. embraces initial interest confusion again http://t.co/3Zaqg3DuQd Bad ruling over Amazon’s internal search. I’ll blog soon
— Eric Goldman (@ericgoldman) July 6, 2015
Ah, my old friend initial interest confusion. How far back we go! I don’t remember anyone screaming about the way I was when I wrote this piece in 2003, though this article preceded mine by a couple of years. Everyone in the academy who thinks about how trademarks work on the Interne “has to despise” IIC (the law profs call it IIC); the trademark plaintiff’s bar, on the other hand, is addicted to it. There are those who seek a less black-and-white approach, too — which, to be fair, includes INTA per that last link — but it seems that most courts that accept IIC use it without the surgical level of focus that INTA suggests is appropriate.
But no one has been eagerly awaiting its demise more than Eric Goldman, who has pronounced it all-but-dead too many times for his own taste — and, again, mine, since the ups and downs have involved a lot of my own cases. Memories, memories. Some people can just delete them, toss them away like an old rag, but not me. They burn forever!
Sometimes they burn brightly, such as when the Eastern District of New York rejected the doctrine, though without ringing declaration of death, in Ascentive v. Opinion Corp. and Devere v. Opinion Corp. Other times the burn was harsh, as when the Third Circuit relied on it, in part, to keep the complaint of Amerigas against the same defendant just alive enough.
It wasn’t even so long ago that Eric, who has a special obsession with the use of IIC to find a basis for liability in the unauthorized use of metatags by competitors (or, sadly, non-competitors) of trademark holders, wrote, regarding a recent opinion:
I’ve been chronicling the futility of competitive keyword advertising lawsuits for a while. I still believe trademark owners haven’t won such a case in court since 2011. Indeed, most of the recent opinions have been brief and breezy defense wins. That makes the latest ruling refreshing–it’s a thorough and comprehensive opinion analyzing the legitimacy of competitive keyword advertising. Yet, the result is the same as all of the others: the court rejects the trademark owner’s request. . . .
The court then turns to the initial interest confusion doctrine–a doctrine still hanging around in court even though it’s been functionally dead for years. (The court cites the 2011 CJ Products v. Snuggly Plushez case, one of the last plaintiff-favorable initial interest confusion rulings). The court interprets Second Circuit law to require “intentional deception” as a prerequisite for initial interest confusion, and the Association can’t show AFA engaged in “intentional deception.” The court concludes:
[the Association] failed to establish any actual confusion stemming from the AFA’s internet search engine advertising practices, or to show the intentional deception needed to prevail under the initial interest doctrine.
Thus, effectively due solely to the weakness of the Association’s actual confusion showing, the court says the trademark owner didn’t show a substantial likelihood of success. I interpret this result as a sign the judge simply doesn’t believe consumers were harmed by the competitive keyword advertising because there was no real actual confusion. As I explained in this recent paper, this fundamental lack of actual consumer confusion caused by competitive keyword advertising is driving defendants’ success in keyword advertising cases. How long ago is not long ago? Last week. Well, okay, this week — on July 8th. And then…
What happened? Well, as Eric explains in a bracketed introduction to that last (July 8th) post “I wrote this post over the weekend, before the atrocious MTM v. Amazon case. I think virtually all of this post remains current despite that ruling. I’ll blog the MTM case separately.” I’m certainly looking forward to it. Meanwhile, though, there’s me.
Currently, Amazon users who search for “MTM Special Ops,” a type of watch made by Multi Time Machine, will see a list of watches made by competitors and sold by Amazon. Above the competing products, consumers see the phrase “MTM Special Ops” since it remains in the Amazon search bar into which they type the phrase. Even if users ultimately knew they weren’t buying from MTM, the list of results may cause “initial interest confusion.” That would allow MTM to bring a trademark lawsuit, the two-judge majority wrote. . . .
“A jury could infer that users who are confused by the search result will wonder whether a competitor has acquired MTM or is otherwise affiliated with or approved by MTM,” wrote US Circuit Judge Carlos Bea on behalf of the majority.
In dissent, US Circuit Judge Barry Silverman compared Amazon’s offering up of competing products to a diner that, when a customer asks for a Coke, explains they serve only Pepsi. “No reasonably prudent consumer accustomed to shopping online would likely be confused,” Silverman wrote.
“A jury could infer that users . . . will wonder“?! The court explains, or tries to explain, what’s so bad about users wondering (link added by me):
In any event, even as to expensive goods — for instance, pianos sold under a mark very similar to the famous Steinway and Sons brand’s mark —- the issue is not that a buyer might buy a piano manufactured by someone other than Steinway thinking that it was a Steinway. The issue is that the defendant’s use of the mark would cause initial interest confusion by attracting potential customers’ attention to buy the infringing goods because of the trademark holder’s hardwon reputation. Brookfield, 174 F.3d at 1063 (citing Grotrian, Helfferich, Schulz, Th. Steinweg Nachf. v. Steinway & Sons, 523 F.2d 1331, 1341–42 (2d. Cir. 1975)).
A jury could infer that the labeling of the search results, and Amazon’s failure to notify customers that it does not have results that match MTM’s mark, give rise to initial interest confusion. If so, a jury might find that Amazon customers searching for MTM products are subject to more than mere diversion, since MTM is not required to show that customers are likely to be confused at the point of sale. Playboy, 354 F.3d at 1025.
Again: The issue in finding a LIKELIHOOD OF CONFUSION is not, the court says, “that a buyer might buy” the wrong thing. In other words, we’re not worried about the likelihood of damage to the trademark holder because a confused person buys the competitor’s goods thinking they are the trademark holder’s goods. Rather — read carefully —
The issue is that the defendant’s use of the mark would cause initial interest confusion by . . .
Note that the words after “by” have been ellipsed by me, because . . . contrary to the court’s suggestion, they don’t matter. Not only has the court accepted the premise of IIC that tort damage, as traditionally understood, i.e., a loss of revenue — or even the likelihood of tort damage — is not necessary once you invoke “initial interest.” It has switched out the result (liability for causing X) for the process (liability simply for doing Y). Or, read a different way, the “issue” is that there could be liability for initial interest confusion because it might be initial confusion.
Yes, a tautology. Justified by what? The possibility of “attracting potential customers’ attention to buy the infringing goods because of the trademark holder’s hardwon reputation.”
Really? “Attracting attention” is a tort?
Someone’s “reputation” — that’s not good will, is it? — is something, says the court, you can’t use to “attract attention” to your own offerings. Where is that concept found in the Lanham Act, exactly?
Let’s look at the dissent. It’s clearly the work of, well, one of those younger judges… like someone my age (in fact Judge Silverman is two years older than I am). Here goes:
Live! From New York! It’s Saturday Night! . . . and the scene is the Olympia Restaurant, Chicago, January, 1978. Dan Aykroyd is manning the grill, Bill Murray is working prep, and John Belushi is up front taking orders. A customer, Jane Curtin, walks in and orders two cheeseburgers. Belushi yells to the grill: “Cheezborger, cheezborger.” Curtin then orders a Coke. Without looking up, Belushi replies: “No Coke. Pepsi.”
Pause it right there.
Would anyone seriously contend that the diner violated Coke’s trademark by responding to the customer’s order that it doesn’t carry Coke, only Pepsi?
Ok, that was the fun part. Now Judge Silverman applies the Cheezborger Principle, to wit:
I would affirm. “The core element of trademark infringement” is whether the defendant’s conduct “is likely to confuse customers about the source of the products.” E. & J. Gallo Winery v. Gallo Cattle Co., 967 F.2d 1280, 1290 (9th Cir. 1992). Because Amazon’s search results page clearly labels the name and manufacturer of each product offered for sale and even includes photographs of the items, no reasonably prudent consumer accustomed to shopping online would likely be confused as to the source of the products. Thus, summary judgment of MTM’s trademark claims was proper. . . .
[I]n evaluating claims of trademark infringement in cases involving Internet search engines, we have found particularly important an additional factor that is outside of the Sleekcraft test: “the labeling and appearance of the advertisements and the surrounding context on the screen displaying the results page.” Network Automation, 638 F.3d at 1154. . . .
Here, the confusion is not caused by the design of the competitor’s mark, but by the design of the web page that is displaying the competing marks and offering the competing products for sale. Sleekcraft aside, the ultimate test for determining likelihood of confusion is whether a “reasonably prudent consumer” in the marketplace is likely to be confused as to the origin of the goods. Dreamwerks, 142 F.3d at 1129. Our case can be resolved simply by an evaluation of the web page at issue and the relevant consumer. . . . In other words, the case will turn on the answers to the following two questions: (1) Who is the relevant reasonable consumer?; and (2) What would he reasonably believe based on what he saw on the screen? . . .
The goods in the present case are expensive. It is undisputed that the watches at issue sell for several hundred dollars. Therefore, the relevant consumer in the present case “is a reasonably prudent consumer accustomed to shopping online.” Toyota Motor Sales, U.S.A., Inc. v. Tabari, 610 F.3d 1171, 1176 (9th Cir. 2010). . . .
Pretty much right on. I do have a quibble with Judge Silverman, though. Reading through his opinion, it’s pretty clear that he doesn’t really buy IIC. Yet he never says, “We have got to away with that nonsense.” The result is a passage such as this one:
MTM argues that initial interest confusion might occur because Amazon lists the search term used – here the trademarked [sic] phrase “mtm special ops” – three times at the top of the search page. MTM argues that because Amazon lists the search term “mtm special ops” at the top of the page, a consumer might conclude that the products displayed are types of MTM watches. But, a review of Amazon’s search results page shows that such consumer confusion is highly unlikely. . . .
Further, some of the products are not even watches. . . . It is hard to fathom how a reasonably prudent consumer accustomed to shopping online could view Amazon’s search results page and conclude that the products offered are MTM watches. Some of the products are not even watches! And the watches that are offered for sale are clearly labeled 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 page. But, “[u]nreasonable, imprudent and inexperienced web-shoppers are not relevant.”Tabari, 610 F.3d at 1176 . . .
Heh. “Dolt.” Besides what looks like a proofreading error (the repetition of “some of the products are not even watches”), the dissent seems to make a formal error — essentially making an argument that IIC doesn’t matter because even a dolt can read the labels on the watches while claiming to argue that for that reason, there can’t be any ICC. What the judge doesn’t say is that the reason there can’t be any IIC is because he doesn’t think, as a legal matter, there is such a thing.
And he has to say that, or at least he should because he means that, as his argument demonstrates. He does not even respond to the majority’s reliance on the idiotic formulation of Brookfield which makes “attracting attention” a “reputation” infringement and hence a violation of the Lanham Act — remember?: “[D]efendant’s use of the mark would cause initial interest confusion by attracting potential customers’ attention to buy the infringing goods because of the trademark holder’s hardwon reputation.”
So to some extent even the dissent here is confusing the issue, which is not really whether there were grounds for holding that a jury could find the “existence” of an “act” called IIC but rather whether this thing called IIC — using a trademark holder’s “hardwon reputation” to “attract attention” and induce a consumer to knowingly buy something else — gives rise to the tort of trademark infringement, i.e., whether it is of any legal significance at all.
For now, however, it is. Or isn’t. Depending on your judge, or your panel, or whatever.
HUGE UPDATE: Wow:
Multi Time Machine, Inc. v. Amazon.com, Inc., No. 13-55575
(9th Cir. Oct. 21, 2015)
Reversing itself (with amicus advocacy from, among others, yours truly), the panel now by 2-1 holds that Amazon should have gotten summary judgment for its practice of responding to searches for “MTM Special Ops” with other, clearly labeled watches, without explicitly stating that none of its
results are for that exact product.
Because of the clear labeling of the results, “no reasonably prudent consumer accustomed to shopping online would likely be confused as to the source of the products”:
To whatever extent the Sleekcraft factors apply in a case such as this – a merchant responding to a request for a particular brand it does not sell by offering other brands clearly identified as such – the undisputed evidence shows that confusion on the part of the inquiring buyer is not at all likely. Not only are the other brands clearly labeled and accompanied by photographs, there is no evidence of actual confusion by anyone.
Sleekcraft isn’t a rote checklist, and different factors may be important in different circumstances. In search engine cases, an additional factor is particularly important, per Network Automation: “the labeling and appearance of the advertisements and the surrounding context on the screen displaying the results page.” By contrast, the multifactor test “is not particularly apt,” because it was developed for a different problem, “i.e., for analyzing whether two competing brands’ marks are sufficiently similar to cause consumer confusion.” [Comment: I’ve rarely seen a better admission that IIC is a bad idea. That “different” problem is called “whether there is trademark infringement.”]
But, the majority notes in a footnote, a panel can’t get rid of IIC. So, let’s go to the argument that’s not about real infringement: MTM’s argument isn’t that Luminox and other competitors’ brands have confusingly similar marks, but rather that Amazon’s search page creates IIC, because the search results page displays the search term used – here, “mtm special ops” – followed by a display of numerous watches manufactured by competitors, without explicitly informing the customer that Amazon does not carry MTM watches. The alleged confusion isn’t caused by competitors, but rather by the web page’s design.
Still, setting aside the multifactor confusion test, ultimately the question is whether a reasonably prudent consumer is likely to be confused.