10th Circuit Recognizes Initial Interest Confusion Doctrine–Australian Gold v. Hatfield
By Eric Goldman
Australian Gold, Inc. v. Hatfield, 2005 WL 3739862 (10th Cir. Feb. 7, 2006)
Following on a regressive initial interest confusion/metatags case from January, last week the 10th Circuit found initial interest confusion based on metatag usage and keyword purchases. I believe this is the first initial interest confusion case from the 10th Circuit, and the 10th Circuit now joins the ranks of many other circuits recognizing the doctrine. Unfortunately, the 10th Circuit is moving in the wrong direction given that several circuits carved back the initial interest confusion doctrine in 2005 (including the 4th Circuit and the 6th Circuit). Could the initial interest confusion doctrine strike back in 2006?
This particular case is a textbook example of how bad facts make bad laws. The product at issue is “indoor tanning lotion”–I don’t know anything about this product, but I infer it’s a skin dye to give the impression of a natural tan. The manufacturer tries to control its channel from unwanted direct-to-consumer Internet distribution. Plaintiffs distribute the product to distributors, who are authorized to redistribute only to tanning or beauty salons. The plaintiffs claim that this distribution channel is necessary because applying the product properly requires training, plus face-to-face distribution to end consumers creates upsell opportunities. (I’m always skeptical when channel control is justified as being in the consumer’s benefit, but let’s assume that is true for now).
The defendants obtained the product using false pretenses; they told intermediate distributors that they operated a tanning salon and they hid their true identity. The defendants then resold the products directly to consumers via websites that also sold competitive products. The defendants’ initial acquisition from intermediate distributors, as well as the Internet direct-to-consumer resale, violated the intermediate distributors’ agreements with the manufacturer.
To procure Internet traffic for their websites, the defendants put the plaintiffs’ trademarks in metatags and bought the trademarks as Overture keywords. This behavior continued even during periods when the defendants were not actually selling the plaintiffs’ products via the websites.
Unquestionably, the defendants engaged in bad behavior, such as lying to get the product and causing the intermediate distributors to breach their agreements. The defendants also don’t look so good when they failed to produce documents responsive to a discovery request, which the plaintiffs discovered by dumpster diving the defendants’ trash. But did the defendants commit trademark infringement by reselling the products over the Internet?
There are two separate trademark issues here: (1) the resale of an illegitimately-acquired but legitimate and authorized product, and (2) the marketing of a website using metatags and keywords when the product isn’t sold there (or even when it is). Unfortunately the court bypasses all of the subtleties of both issues by relying on the initial interest confusion doctrine. The court says that the defendants committed initial interest confusion by using the manufacturer’s trademarks in metatags and keywords to drive consumers to a website where they could buy competitors’ products. The court also rejects that a disclaimer cures initial interest confusion.
Finally, the court rejects the defendants’ first sale defense. The court says that the first sale doctrine does not apply when resellers convey the false impression that they are authorized, and the court says that the metatag usage and keyword purchases do just that.
Ultimately, the court upholds the jury finding in favor of plaintiffs and an aggregate damages award of $5M against the group of defendants (some of these damages are attributable to a tortious interference claim, not just the trademark infringement claim).
Consistent with a bad facts/bad law case, there are several structural problems with this opinion, of which I’ll only touch on two. First, the court’s reliance on the initial interest confusion doctrine is lazy. Let’s put aside the question of whether there was trademark use in commerce when the trademarks were put in the metatags or used as keywords to trigger ads (a question the court simply ignores). To analyze whether consumers experienced (or could have experienced) initial interest confusion, we would need lots of facts—did search engines even index the metatags? What content did searchers see in search results triggered by metatags that were actually indexed? In the case of the keyword advertising, what did the ad copy say? Should there be a different standard when the defendants were actually selling the manufacturer’s goods from the website? The court does not consider any of these issues.
The first sale analysis is also problematic. Here, the reseller sold genuine unmodified goods. This wasn’t a gray-market goods case where the goods were designed for a different audience. So there is no obvious basis to find consumer confusion—the branding on the product accurately communicated the product’s source. The manufacturer may not like channel leakage, but the first sale doctrine applies despite the leakage. See Sebastian International, Inc. v. Longs Drug Stores Corp., 53 F.3d 1073 (9th Cir. 1995). The court acknowledges the Sebastian case but says Sebastian only applies when the reseller purchased the product legitimately. However, the Sebastian facts are inconsistent with this statement—according to that case’s facts, the reseller “presumably purchases Sebastian products from a salon or distributor who sells the products to [the reseller] in violation of its agreement with Sebastian.”
The Australian Gold court also says that resellers can’t imply an authorization (the implied endorsement argument), which is consistent with the Sebastian case. However, the court treats the metatag usage and keyword purchases as per se evidence that consumers would assume an implied authorization. I would like someone to explain how either technique creates an impression from consumers that there is an implied authorization; this court doesn’t (this is the same problem that the Seventh Circuit created in its infamous sua sponte amendment to Promatek).
As a result of the 10th Circuit’s corner-cutting, I think the 10th Circuit will need to revisit its holding in future cases to clean up its doctrinal errors—much like other circuits that lazily adopted the initial interest confusion doctrine to resolve the dispute at hand (the 9th and 7th circuits come most immediately to mind) have had multiple opportunities to revisit/reconsider their errors.