Attempted Trademark Workaround to 47 USC 230 Immunity Fails Badly—Ascentive v. PissedConsumer [Catch-Up Post]
By Eric Goldman
[This is one of the top dozen or so most important Internet law opinions of 2011, but unfortunately it came out just as I was going into my exam-grading exile and I had to put blogging it on hold. Even over a month later, it's still worth your careful review.]
In my Regulation of Reputational Information paper, I explain how vendors are misusing intellectual property to control consumer perceptions of their businesses. One example is Medical Justice, which tried to use copyright law to work around 47 USC 230 and suppress unwanted reviews. Fortunately, Medical Justice has abandoned that effort.
Other vendors try to use trademark law to work around 47 USC 230. By definition, consumers must reference a vendor’s brand in order to review it, and trademark’s doctrinal plasticity means that such references arguably support a prima facie trademark claim. (I explain that issue more in my Online Word of Mouth paper). As a result, we’ve seen a number of vendors dabble with trademark claims against consumer reviews. For two examples, see Lifestyle Lift v. RealSelf and Eppley v. Iacovelli. (For more on the noteworthy litigiousness of doctors against consumer reviews, see this post).
In this case, the plaintiffs used trademark law to make a no-holds-barred assault on the 47 USC 230 immunity’s applicability to consumer reviews. Their arguments go nowhere. I hope this emphatic ruling will discourage other plaintiffs from trying to use trademark law to work around 230.
Likelihood of Consumer Confusion
The court tried to do a straight-laced multi-factor LOCC analysis, but as I’ve noted before, the LOCC factors don’t make sense when comparing apples and oranges like a vendor and a review site of the vendor. On the bad faith factor, the court says:
While it may be true that PissedConsumer has engaged in sharp-elbowed and perhaps unethical SEO tactics meant to make its webpages appear more relevant to search engines such as Google or Yahoo! than they actually are, that fact has no bearing on the inquiry here—whether PissedConsumer has attempted to sow confusion as to the source, origin, or affiliation of its products and services with those of plaintiffs.
The court instead observes: “Indeed, it is clear that PissedConsumer is not using plaintiffs’ marks as source identifiers at all.” Well, that’s only partially true–PissedConsumer is using the plaintiffs’ marks as referents for the plaintiffs. (See Deregulating Relevancy for more on the implications of that). In a footnote, the court said there wasn’t a dispute that PissedConsumer was using the marks in commerce, but the court failed to reconcile these seemingly inconsistent statements.
To bolster their unmeritorious trademark claim, the plaintiffs argued that several specific technological features used by PissedConsumer supported trademark infringement. The court rejects the plaintiffs’ arguments on each feature:
* using the plaintiff’s trademark as a third level domain name, i.e., ascentive.pissedconsumer.com. The court said that the pissedconsumer.com domain name makes it clear to consumers that the site is critical of, and therefore not affiliated with, the mark owner.
* using the plaintiff’s trademark in the consumer reviews. The court says there’s no consumer confusion here either:
after a brief inspection of the content of PissedConsumer’s website, the user would realize that they were visiting a third-party gripe site for “pissed” consumers.
* metatags. The court rejects initial interest confusion. First, there can’t be competitive diversion because PissedConsumer isn’t selling anything to consumers. Second, no one searching for the plaintiffs would be “diverted” to the defendants’ website. (A point I make in gory detail in my Deregulating Relevancy article). Third, initial interest confusion imposes minimal (if any) harm on consumers because they can hit the back button. Finally, the court recognizes that technology has evolved since the 1999 Brookfield ruling such that metatags don’t matter (citing, among other things, Google’s 2009 blog post to that effect—thanks, Matt Cutts, for doing that!)
* black hat SEO. The opinion talks in some detail about linking archive posts from Twitter with the hope that Google will treat the posts as fresh content. The court says:
While it may be—and likely is—the case that PissedConsumer’s SEO practices are intended to make its webpages seem more relevant to search engines than they actually are and these methods may indeed violate the search engines’ terms of services, the remedy for this conduct is not trademark law but instead with the search engines themselves.
Amen to getting trademark law out of the way and letting search engines fix the gaming! This is another point I made ad naseum in my Deregulating Relevancy article.
* serving ads (through Chitika) showing the plaintiffs’ trademarks, presumably automatically triggered by keywords on PissedConsumer’s pages. The court says that, at most, PissedConsumer as the publisher is contributorily liable to any infringement committed by the ad network (Chitika), but the plaintiffs didn’t allege contributory infringement. The court seemed to treat Chitika as the direct infringer instead of the advertisers, but in fact I think Chitika should be evaluated under contributory infringement as well, with the advertiser being the direct infringer (if there is one).
Although the court gets to the right place, its doctrinal jujitsu shows what happens when trademark law is stretched to places it doesn’t belong. We’ve lost too many of the limiting principles in trademark law that should help make a case like this an easy one for judges. Among other things, a more robust use in commerce doctrine would have ended much of this case early, and the very lengthy opinion oddly doesn’t mention the seemingly applicable doctrine of nominative use at all.
47 USC 230
Having dispatched the plaintiffs’ trademark assault, the court mops up all of the remaining state law claims using 47 USC 230. The court says “a website such as PissedConsumer constitutes an ‘interactive computer service,’” which makes PissedConsumer’s officers “providers” of an ICS. This is an unusual reading of the statute, but it’s all good.
The court rejects the plaintiffs’ Roommates.com attack on 230, saying “determining what makes a party responsible for the ‘development’ of content under § 230(f)(3) is unclear, and the CDA does not define the term.” Thus, the court says it’s appropriate to examine the totality of the circumstances; plus, “one is responsible for the ‘development’ of information when he engages in an act beyond the normal functions of a publisher (such as deciding to publish, withdraw or modify third-party content) that changes the meaning and purpose of the content.” The Roommates.com attack fails here because the plaintiffs provided no evidence that PissedConsumer actively created the content; their unsupported general assertions weren’t enough. The court rejected the application of the old (and quite outmoded, IMO) Badbusinessbureau opinion, saying PissedConsumer’s “actions are not unlike the targeted solicitation of editorial material engaged in by a narrow genre of publishers.” (Huh?) Inviting consumers to post reviews and SEOing the pages didn’t change the analysis. Accord Asia Economic Institute v. Ripoff Report.
Separately (and not relying on 230), the court tosses the RICO claim because the plaintiffs didn’t show that PissedConsumer engaged in commercial bribery or extortion.
On these bases, the court rejects the plaintiffs’ request for a preliminary injunction. However, the case is ongoing, and the plaintiffs still get discovery.
Although not a party to the suit, the real party-at-interest in this case is Google, because both Ascentive and PissedConsumer depend on Google traffic as virtually their entire marketing plan. In Ascentive’s case, it said that 99% of its sales are made online, and a majority of that came from Google searches. Indeed, Ascentive had previously sued Google for trademark infringement before abandoning that claim. Meanwhile, PissedConsumer’s business is to get favorably indexed in Google for businesses’ names and then sell them services that take the edge off any negative user content that gets indexed. As a result, both litigants are competing against each other for favorable placement in Google search results. In my Online Word of Mouth paper, I discuss how brand owners face unusual and effectively unprecedented competition on their own brands for scarce consumer attention—in this case, the scarce resource of top search engine placement—and how that dynamic leads to weird trademark lawsuits like this one.
The legal ruling may be good for PissedConsumer, but this opinion isn’t exactly a clean bill of health for its business model. Indeed, “the Court finds some aspects of PissedConsumer’s business practices troubling and perhaps unethical.” I continue to believe that all consumer review businesses that seek to get paid by the vendors they review have a major structural conflict-of-interest—especially when the review site’s sales pitch to the vendor is reputation management. I ultimately think Google will need to restructure its algorithm to reflect the inherent untrustworthiness produced by these conflicts of interest.
Paul Levy’s comments on the ruling.