Competitive Keyword Advertising Lawsuit Will Go To A Jury–Edible Arrangements v. Provide Commerce
The parties compete in the “chocolate and fruit-based gift packages” market. Provide bought competitive keywords that used the plaintiff’s trademark, including the keywords “edible arrangements,” “edible arrangements locations,” “edible arrangements coupons,” “edible arrangements promotional code,” “edible arrangements bouquet,” “edible arrangements flowers,” “cheap edible arrangements,” “edible arrangement discount,” and “incredible edible arrangements.” (The court doesn’t discuss broad matching, but Provide could have achieved similar results by broad matching the “edible arrangements” search term. Perhaps Provide got a better ROI by bidding on these highly specific terms compared to broad matching just the trademark?). At some points, Provide’s ad copy included the phrase “edible arrangements,” but it later changed the ad copy to reference “edible fruit arrangements” as depicted in this screenshot from the opinion:
Provide asserted that it always used its house brands in close proximity to the “edible fruit arrangements” phrase, but Edible provided examples where the house brand only appeared in the URL displayed in the ad copy. Provide has subsequently given up using the “edible fruit arrangements” phrase in its keyword ad copy.
Edible provided evidence that the competitive keyword advertising had some success. They cite evidence where Provide allegedly said the ads were “one of Provide’s most successful tools for converting sales,” and that for “Provide’s 2012 Mother’s Day promotions, ‘EDIBLE ARRANGEMENTS’ is the top generator of ‘impressions’ for Shari’s Berries.”
The court tries to apply a standard multi-factor likelihood of consumer confusion test comparing the trademark “Edible Arrangements” with the ad copy references to “edible fruit arrangements,” but many of the factors are unenlightening or nonsensical. Still, with the inferences stacked against the party moving for summary judgment, the court easily denies Provide’s summary judgment motion.
The court then separately analyzes the trademark implications of Provide’s keyword bidding. Here, the court expressly acknowledges that several factors in the standard multi-factor test are “not particularly helpful.” So instead the court creates a new mutant variation of the “Internet trinity” (which the Ninth Circuit repudiated in the Network Automation case) focusing on three factors:
(i) “the strength of the plaintiff’s mark as a unique search term related to a distinct line of products,”
(ii) “the similarity of the marks and whether the defendant’s mark draws a clear distinction as a competing brand,” and
(iii) “what the consumer saw on the screen and reasonably believed, given the context.”
Oh joy, an all-new expedited test for consumer confusion in the keyword advertising context. I’m sure this test will work perfectly in future cases.
The court says this test indicates that Provide should be denied summary judgment:
* strength as search term: “a reasonable trier of fact could conclude that a consumer searching for ‘edible arrangements’ is looking for a distinct product line of aesthetically shaped fruit and not merely for any and all gifts containing boxes of edible fruits and berries.” OK, let’s assume searchers were using the search term “edible arrangements” to find the trademark owner. If so, then either they will not be steered away by competitive alternatives, or apparently their brand loyalty dissolves the instant they are exposed to any other option. See my Deregulating Relevancy article for more on the risks of overassuming what a searcher wants when they use a trademarked search term.
* brand distinction. “the Court finds that the Provide advertisements it has examined make a very poor effort to differentiate either a competing product or seller.” Hello, it’s a keyword ad. There’s not a lot of room for lengthy expositions about the relative brand positioning of competitors in the marketplace in the tightly limited character space.
* Consumer screen depictions. “Provide’s choice not to identify Pro Flowers or Sherri’s Berries as the advertiser in the text of the advertisement or the link, and only in the small print of the URL, contributes to a misleading environment for the consumer. In particular, a user searching for EA’s products might not even know that they had clicked on a link for a competitor’s product until they actually landed on the webpage of one of Provide’s sellers, or even after that point.” Isn’t this the same point as the prior bullet point? So did the court implicitly collapse this bastardized test to 2 factors instead of 3? Warning: this court is saying that putting the advertiser’s brand only in the URL, when the ad copy otherwise uses a similar phrase to the competitor’s trademark, is confusing.
The court summarizes:
A reasonable trier of fact could conclude that Provide’s purchase of “edible arrangements” as a non-consumer facing keyword could result in a likelihood of confusion by directing consumers to a search results page in which it advertised “edible fruit arrangements” in an text advertisement in which the seller is only identified in the small print of the URL. A jury could find that the purpose and effect of Provide’s keyword bidding – in conjunction with its use of EA’s mark in its advertisement on the search results page – was to mislead consumers as to sponsorship or affiliation with EA and to misdirect the web traffic of users searching for EA’s mark.
Finally, “edible arrangements” is a completely descriptive mark, so Provide has a powerful descriptive fair use argument. The court questions why Provide needed to describe its “fruit” as “edible” and whether its packaging of chocolate-covered strawberries constitutes an “arrangement.” Thus, the court says “EA has raised a material issue of fact as to whether Provide’s use of the phrase ‘edible fruit arrangements’ relies upon the ordinary meanings of those words to describe a product containing a box of ‘deliberately placed’ strawberries.”
This is a bad ruling for Provide and a bad ruling for keyword advertisers. Here’s a short list of the things I like least:
* a new bastardized multi-factor consumer confusion test for keyword advertising. NOTHING GOOD WILL COME FROM THIS.
* the court’s failure to understand that consumers cannot simultaneously be brand-loyal when they search a company’s trademark and then instantly brand-disloyal the moment they see competitive alternatives. I had hoped we had sufficiently debunked this inconsistency a decade ago.
* the court’s skepticism that a house brand in a URL sufficiently clarifies the advertiser’s identity to consumers. If you are using a competitor’s trademark in your ad copy, take note.
* the questions about the applicability of descriptive fair use for a trademark so obviously descriptive, especially when the defendant used a modified variation of the term (i.e., the addition of the term “fruit”)
* the court’s insufficient acknowledgement of the pro-competitive implications of keyword advertising. As I explained in my Deregulating Relevancy article, the real question is: did the people searching for “edible arrangements” get what they were looking for? The court cites almost no evidence suggesting that they didn’t; and my article explains numerous reasons why it’s likely that anyone who searched for “edible arrangements” and ended up buying Shari’s Berries got EXACTLY what they wanted. If in fact no consumer was harmed by Provide’s ads, what exactly is the problem that needs to be solved in court?
Fortunately, not all hope is lost. I know of three jury trials over competitive keyword advertising, and the defense won all three. I am not sure if Provide’s modified trademark usage in the ad copy makes this case sufficiently different from those cases, so perhaps those data points are not totally predictive. Still, a lot depends on the skill of Provide’s trial counsel as well as the jury instructions, and it appears the judge will put the thumb on the scale for the plaintiffs when it comes to drafting those.
I contacted both sides’ lawyers for public comment. Plaintiff’s counsel Mark Cantor emailed me: “Edible is very pleased with the result and believes the analysis was relating to keywords was well reasoned.” I didn’t hear back from defense counsel.
Litigation note: This case seemed unusually lawyer-heavy. 8 lawyers from 3 law firms represent Edible Arrangements; 4 lawyers from 2 law firms represent Provide. Next time you shell out for either an Edible Arrangements or a Shari’s Berries, you might reflect on how many private school tuitions are riding on your order.
Case citation: Edible Arrangements, LLC v. Provide Commerce, Inc., 2016 WL 4074121 (D. Conn. July 29, 2016)
Related Posts on Keyword Advertising