Tuesday, March 02, 2010

Subway ruling on failure to disclose/230

Doctor’s Associates, Inc. v. QIP Holder LLC, 2010 WL 669870 (D. Conn.)

Sit back, because this is going to take a while.

Plaintiff (Subway) sued defendants (Quiznos and iFilm) for Lanham Act false advertising and state law commercial disparagement and unfair trade practices, arising out of an ad campaign including two national TV commercials and internet ads centering on Quiznos’ “double meat” line of sandwiches. (Oh, Buffy the Vampire Slayer, how I miss you.)

Quiznos aired a TV ad in September 2006 comparing its Prime Rib Cheesesteak sandwich to the Subway Cheesesteak, focusing on the message that the Quiznos version had twice the meat. The sandwiches were side-by-side on a tray and actual consumers commented: “one man, while looking at the sandwiches, states ‘meat, no meat.’ In another shot, two men are looking at the Subway sandwich and one says he can’t see any meat and after a search says ‘Oh, there it is I see it,’ while the other man says ‘it’s hard to find’ and ‘oh, there's a little.’ Another man, while comparing the two sandwiches, refers to the meat on the Quiznos sandwich as ‘busting out of the sub.’ The Commercial also contained frames consisting solely of text, which read: ‘Quiznos New Prime Rib Cheesesteak v. Subway Cheesesteak ... Only Quiznos has real Prime Rib. And more than 2x the meat.’”

Quiznos then partnered with iFilm to create the “Quiznos v. Subway TV Ad Challenge” at meatnomeat.com. It solicited people to create a video demonstrating “why you think Quiznos is better.” Quiznos posted four sample videos and contestant videos that complied with the rules.

In January 2007, Quiznos aired a second TV ad comparing its double meat Ultimate Italian to Subway’s Italian BMT.

Further on the Cheesesteak: According to Quiznos, the specifications for the Quiznos Prime Rib sandwich called for 5.0 ounces of meat, whereas the specifications for the Subway Cheesesteak sandwich called for less than 2.5 ounces of meat. To check the claim, Quiznos commissioned an independent expert, Restaurant Marketing Group (RMG), which concluded that the amount of meat in an average small Quiznos Prime Rib sandwich was at least twice that of the meat in the standard 6-inch Subway Cheesesteak sandwich. Quiznos also conducted an audit of its stores in the fall of 2006, and alleged that over ninety-four percent of franchisees were making sandwiches that contained at least 4.0 [sic] ounces of meat. Subway argued that Quiznos knew it was making false claims, based on an August 2006 field operations survey showing that 27.65% of the 651 sandwiches tested contained less than 5.0 oz. of meat, with 10.29% of the 651 containing less than 4.0 oz. of meat. A second survey in September 2006 at 39 Quiznos stores found that 74% of the stores made sandwiches containing less than 4.5 oz. of meat. Of 690 stores surveyed, 21.6% served Prime Rib sandwiches with less than 4.5 ounces, though this was still approximately one ounce more than the Subway Cheesesteak with double meat. Though the networks requested substantiation, Quiznos didn’t disclose these results.

Subway argued that it makes all sandwiches to order, including a double meat portion for an extra dollar. The Cheesesteak was available with either a 2.5 ounce portion of a steak, pepper and onion mixture—between 1.67 and 1.77 ounces of meat—or a 5 ounce/3.34-3.54 ounce portion. But the Quiznos ad used a single portion as a comparison, failing to disclose the more comparable sandwich. Moreover, the ad didn’t disclose the price; the Subway was $3.59 or $4.59 with double meat, while the Quiznos sandwich sold for up to $6.79. (Quiznos, however, argued that only 50% of Subway consumers even knew that double meat was an option, and at most 10% of Subway consumers ordered double meat.) Finally, Subway argued that it had discontinued the Cheesesteak at the time the ad aired, replacing it with a Steak & Cheese sandwich with 2.5 ounces of meat (single). While the ad was airing, the Cheesesteak was still being offered in a majority of Subway stores to get rid of remaining inventory.

After the surveys, Quiznos audited 4,370 stores and found that 44.14% of the Prime Rib sandwiches tested contained less than 5 ounces of meat, with 5.86% containing less than 4 ounces; ½ ounce deviation was allowed, so 4.5 ounces was acceptable. Quiznos ultimately terminated about 300 franchisees for failing to comply with the meat specifications for the Prime Rib. Two Colorado franchisees sued, alleging there was no real compliance standard in place an that enforcement was a sham driven by concerns over the Prime Rib ad. The court agreed with the franchisees, finding that the terminations were a charade “driven not by Quiznos’ genuine concern about whether its franchisees were making sandwiches to spec, but rather by its overriding public relations desire to be able to proceed with its national advertising campaign targeting Subway. But the public relations monster had to serve two masters-the action Quiznos took once it ferreted out non-complying franchisees had to look serious (otherwise what would Subway say?), but it couldn’t actually be serious, unless Quiznos was willing to lose a potentially huge number of non-complying franchisees, which it was not.”

Quiznos argued that the Cheesesteak ad was created using fair and accurate procedures, based on sandwiches orered from local stores by members of the production crew posing as ordinary customers, with independent third-party public notaries monitoring the procurement and filming. The notaries signed affidavits attesting that no one manipulated the sandwiches. But Subway witnesses testified that the Subway sandwich in the ad appeared “flattened or squished.” And the Quiznos stores were warned that someone from corporate would be there on the day the ad was filmed and that the Quiznos were “cherry-picked” from the available local options. Subway also offered a consumer survey showing that the Quiznos sandwich in the ad looked better than what customers got in stores and the Subway sandwich looked worse.

Further on the Ultimate Italian ad: the side-by-side comparison was the same. One man says: “If I ran out of gas in front of a Subway I would walk ten miles to get the Quiznos sandwich.” “In another vignette, two men are looking at the Subway sandwich and one states, ‘I don’t see any meat.’ The commercial also includes a person saying ‘the Quiznos is like stacked with a bunch of meat and the Subway sandwich is like when your kindergarten and your Mom throws some stuff together real quick.’” The text frames said, “Quiznos New Ultimate Italian v. Subway's Italian BMT. The Quiznos has 2x the meat,” with a disclosure: “Based upon average precooked weight, in an independent national sampling of Quiznos small Ultimate Italian v. Subway regular 6-inch Italian BMT (12/06). Sandwich prices differ.”

The sampling referred to found that the amount of meat on an average small Quiznos Ultimate Italian was twice that on an average Subway BMT sandwich; the specifications for the sandwiches called for 5 ounces and 2.25 ounces of meat respectively. Subway argued that this was false and misleading for similar reasons (the availability of double meat at Subway, and the fact that Quiznos offered a single-portion Classic Italian sandwich; the higher price of the Quiznos sandwich). Subway’s survey expert, Joel Steckel, concluded that the ad misled consumers to believe that the Subway BMT was more expensive than it actually is and that the two sandwiches were closer in price than they actually are.

Subway claimed that overall sales dropped approximately 4% after each ad.

Further on the internet contest: In late 2006, the public had the chance to “grab a camera and show us why you think Quiznos is better.” Prizes included a showing of the winning video on VH1, cash, a year’s supply of Quiznos, iPods, and so on. Quiznos created one of the sample videos (“Barbie Breaks, You Get What You Pay For,” which sounds like it raises other interesting issues), and reviewed and approved the other three, created by iFilm. Subway alleged that these were false and misleading too.

Subway also sued over the contestant videos, which were allegedly posted as they were submitted. But Subway claimed that defendants were responsible for the content, because it claimed ownership of all materials submitted under the contest rules. (This is ridiculous given §230: all websites will give themselves the right to use submitted material; that doesn’t affect whether the material was provided by another information content provider.) Subway also pointed to the requirements for entries, which demanded a video that “compares Quiznos to Subway and illustrates why Quiznos is better than Subway,” the similar judging criteria, and “Thought Starters” suggesting ideas such as “Double Meat.” The rules prohibited any “false or misleading statement, or any libelous, slanderous or disparaging statement regarding Quiznos or Subway, or of either companies’ products or services.” (You can see here why Eric Goldman says Roommates.com will often be a defense cite: those guidelines and suggestions don’t inherently require the submission of defamatory/disparaging content, so under Roommates Quiznos should be protected by §230.)

Making the move people who don’t like §230 always make, Subway argued that defendants “undertook the responsibility to review the contest entries, and had the ability to exclude entries containing inappropriate content.” Nonetheless, Subway contended, defendants posted videos with false, misleading, or disparaging statements about Subway. The sample video—not subject to §230—claimed that the Subway sandwich is “all bread” and used a “beauty shot” of a Quiznos sandwich to compare to a fake Subway picture; called the Subway sandwich “Lettuce Starve” and “Hunger Pain Inducing”; etc. Contestant videos were also objectionable: “a Subway sandwich portrayed as a submarine unable to dive because it does not have enough meat; two persons trying to decide where to eat and referring to Quiznos’ double meat as a deciding factor, thereby falsely implying that Subway has no double meat option; and a video depicting a sandwich ‘build off’ that explicitly states that Subway's sandwich has little meat and much less meat than Quiznos’ sandwich.”

First, the parties fought over Subway’s expert, whose survey backed up Subway’s claim that the Ultimate Italian ad was misleading because the sandwiches weren’t comparable in content or price and the prices weren’t disclosed. The mall intercept survey concluded that the ad led consumers to believe that the Subway price was higher than they’d otherwise believe and that the price was more comparable to Quiznos’ price than it actually was.

The test group saw the ad and the control saw a “beauty shot” of a Subway sandwich and a description from Subway’s menu. All respondents were asked to give the price they’d expect to pay, based on the ad and their knowledge of fast food. In the test group, the respondents, on average, opined that the Subway price was $4.91, with the Quiznos at $5.81. In the control group, the respondents averaged $4.48. (I think there’s an obvious medium problem here: the control should have showed a noncomparative TV ad.) The actual average current market price for the Subway was $3.84 ($5.08 for the Quiznos). The difference between control and test was $0.43, and the test group perceived a $0.90 gap between the parties’ sandwiches when the actual gap was $1.24 (though we have no control for the perceived gap; to get that, we’d at least have needed beauty shots/descriptions of the Quiznos sandwich in the control group).

The Second Circuit favors admitting surveys, with objections going only to their weight unless their probative value is substantially outweighed by danger of unfair prejudice or confusion. The court found this survey relevant. It goes to consumer perception of value. If the two products seemed to cost the same, consumers might have been misled into thinking they’d get better value for their money. (As an exercise, Google either party’s name and “you get what you pay for.” It seems that the perception that Quiznos is pricier is fairly broadly shared, and that’s not necessarily bad for Quiznos.) So, the ad was arguably misleading because it omitted information explaining the price difference.

All I can say here is: Subway’s lawyers did a good job. There are lots of omissions in every ad, necessarily. One ordinarily doesn’t have to disclose matters not inherently related to the attribute one is touting: if you’re featuring quality, you don’t have to explain that your product costs more, though you would have to disclose quality-related things that might otherwise mislead consumers. Likewise, if you’re featuring cheapness, you don’t have to explain that your quality is at the low end, though you could have to disclose if your unit pricing actually meant that consumers were likely to spend the same total amount as they would on the competitor’s product.

Though the ad here said nothing about cost, consumers apparently made inferences about cost based on a quality/amount comparison—they inferred that the Quiznos would cost more—but they underestimated the differences. Questions: what then should Quiznos have disclosed, in Subway’s view? Given the result, “prices differ” isn’t going to work, but does Quiznos then have to conduct an average price survey before advertising about amount of meat? Is the underestimation material to consumers? Would any comparative ad touting Quiznos over Subway without disclosing prices have produced similar results?

This last question to me is why the absence of a control for Quiznos makes the survey very troublesome. Once consumers are induced to compare Subway and Quiznos, they may well be operating on assumptions not contained in the ad. The control group didn’t make that comparison; I would think that the only reliable result from the Survey is the Subway-Subway difference, and not the Subway-Quiznos gap. So the court was wrong to accept Steckel’s conclusion that the ad creates a perception of greater price comparability. Steckel argued that his control tested what consumers were likely to think without comparisons, but that means he was, without meaning to do so, testing the effect of any comparison on consumer judgments, not testing the effect of this particular allegedly misleading comparison. If consumers can lawfully be told “Subway has an Italian sandwich, and Quiznos does too; we at Quiznos think ours is better,” then one must test the effect of the more meat representations here, not the effect of being asked to compare. Again, once consumers are thinking comparatively, they will be thinking differently and that may induce them to change their price estimates—moving the anchor, in effect—regardless of what the ad says. The question of controls is a separate issue from whether the ad should be deemed to make any implicit representations about price.

Subway, however, succeeded in persuading the court that the ad might have been comparing apples and oranges, and thus needed to disclose more differences. The survey’s flaws were not so egregious that its prejudicial effect outweighed its probative value. The court even accepted the lack of a “don’t know” option because Steckel opined that “don’t know” “is often a cop-out to prevent people from doing the hard work in ... making some logical conclusion[,]” and that, “in this particular design, it would also encourage guessing in the control, and that would minimize any difference between test and control.” (There are cases to the contrary; I can see both sides of the argument, but without controlling precedent this ruling just makes surveys more vulnerable to what the court thinks the right result is.)

On to the merits: the court denied summary judgment. The Cheesesteak ad did not make establishment claims, because it didn’t purport to rely on any tests. So Subway had to show falsity; Quiznos responded that its claims were true: the Quiznos sandwich actually contained two times as much meat as the standard Subway sandwich. The court found numerous issues of material fact. The Quiznos audit raises questions about the “two times” claim, since 27.65% of 651 franchises failed to meet the 5 ounce standard, and 10.29% made sandwiches with less than 4 ounces of meat. In later tests, 74% of stores made sandwiches with less than 4.5 ounces, and a comprehensive survey showed that 44.14% of Prime Rib sandwiches contained less than 5 ounces, with 5.86% containing less than 4 ounces. (Incidentally, how do Subway stores do by this standard? Doesn’t Subway have to show that its franchises do better? If they are also subject to variation, Quiznos may still be delivering 2x the meat.) Moreover, the court found even these results highly questionable, given that another court deemed the process a “charade.”

The court also found that there was a material question of fact about the 2x meat claim because Subway offered a double portion of meat for an extra $1. (How to square this with the survey results, which seem to indicate that consumers understood the comparison?) Quiznos knew about the Subway option but opted to air an ad “comparing a sandwich it had specifically designed to contain a double portion of meat to a Subway sandwich with a single portion of meat.” Moreover, the Subway sandwich depicted in the ad had actually been discontinued while the ad was airing and replaced with the Steak and Cheese, which did contain at least 2.5 ounces of meat. So a reasonable jury could conclude that the ad was literally false.

Quiznos argued that this couldn’t constitute literal falsity, because the ad was ambiguous. But the unambiguous message was that Quiznos’ product contains at least twice the meat of Subway’s product. This was reinforced by the “men on the street” commentary that the Subway had “little meat” or “no meat.”

Unlike the Cheesesteak ad, the Ultimate Italian ad made establishment claims. Thus, it could be proved false by showing that the supporting study was not sufficiently reliable. Again, there were genuine issues of material fact, centered on the availability of a Subway double meat option.

With respect to the Cheesesteak ad, the extrinsic evidence of deception was an internet survey of 97 consumers, which found that the Quiznos Prime Rib sandwich “looked better” in the ad than what consumers had experienced, while the Subway Cheesesteak sandwich “looked worse.” The court found the survey relevant to Subway’s claim that the ad made Subway’s sandwich look less appetizing than what customers experienced, because it asked consumers whether they believed that the sandwiches shown “accurately depicted” what the sandwiches would look like if ordered. (Doesn’t that question disprove misleadingness/prove puffery, if consumers think the ad doesn’t depict reality?) The survey was not so flawed that its probative value was outweighed by the risk of prejudice or confusion. So: genuine issue of material fact on whether the ad misled about the appearance of the sandwiches.

Likewise, with the Ultimate Italian ad, “a reasonable jury could conclude that the omission of pricing information was misleading and deceptive to consumers because consumers were misled to believe that the two sandwiches were similarly priced but the Quiznos sandwich had more than twice the meat and thus was a better value,” or “that viewers were misled by Quiznos' failure to compare more comparable sandwiches.” Quiznos had its own survey, which found a “more meat” takeaway without a price message, but that just creates an issue for the jury.

Subway also argued that the internet contest violated the Lanham Act: (1) it was advertised as a “meat-no meat comparison,” claiming that Subway’s product had little or no meat; (2) Quiznos claimed superiority over a materially different product; (3) it solicited videos depicting Subway’s product as having no meat; (4) it compared Quiznos’ product to a discontinued Safeway product. (Except for (3), framing the issues this way seems to plead around §230: to the extent these claims can be resolved without looking at the user-generated videos, no problem. (3) might also be resolved by looking only at the solicitations, but do those count as “advertising or promotion”?)

“Viewing the Contest in its entirety,” the court found genuine issues of material fact. For example, the contest used the domain name meatnomeat.com, which is arguably a literal falsity because it “clearly implies” that a Subway sandwich has no meat. (Or perhaps it’s puffery.) Without extrinsic evidence, Subway will have to argue literal falsity to prevail against the contest. The four sample videos “designed to shape” the submissions also make no meat/less meat claims. Defendants argued that no reasonable consumer would have been deceived by these representations, but that’s for the jury.

Defendants argued that the sample videos and contestant videos weren’t “commercial advertising or promotion.” The samples weren’t commercial speech made for the purpose of influencing consumers to buy Quiznos’ products, but rather examples of how consumers could enter the contest, and they were only available to an extremely narrow range of viewers who went to the contest pages, and thus not disseminated sufficiently to the relevant consuming public. The court found these arguments unpersuasive. The contest was designed to influence consumers, and was part of an integrated Quiznos campaign; the ads were available to anyone with internet access, and thus sufficiently disseminated. Plus Quiznos posted the winning submission on VH1 and on a Times Square billboard.

Okay, but what about the contestant videos? §230 seems like the big barrier. Subway argued that defendants “went beyond the role of a traditional publisher” by “soliciting disparaging material” and “shaping the eventual content.” (Um, soliciting material and shaping content is what traditional publishers do. I know that’s hard to remember!) The court found that the critical inquiry was whether defendants merely published information provided by third parties or instea were “actively responsible for the creation and development of disparaging representations about Subway contained in the contestant videos.” Publishing includes reviewing, editing, altering, and posting content. The court held that it was unclear at this stage whether defendants had exercised the role of a traditional publisher.

Unless there are allegations that Quiznos or iFilm inserted disparaging content into submissions, it’s not unclear. What further facts do we need? The court relied on a 2004 case holding that a website goes beyond the traditional publisher’s role by actively encouraging, instructing, and participating in consumer complaints, such as by encouraging consumers to take pictures for posting on the website. As a result, the site “incurred responsibility for the information developed and created by consumers.” As I’m sure Eric Goldman will shortly elaborate, this is not a mainstream view of §230; the general question is: who provided the unlawful content? Otherwise a news site with a comments section asking people to provide their opinions—imagine a site with an editorial stance that occasionally asks leading questions, like, say, Fox—could be liable for defamatory comments.

Here, the court reasoned, defendants invited contestants to submit videos on the theme “why you think Quiznos is better.” Somehow the domain name meatnomeat.com is also relevant for implying that Subway has no meat. And the sample videos are also relevant because they arguably contain false representations. “Whether the Defendants are responsible for creating or developing the contestant videos is an issue of material fact, best submitted to the jury after viewing all of the relevant evidence.” A reasonable jury could conclude that the defendants “actively solicited disparaging representations about Subway and thus were responsible for the creation or development of the offending contestant videos.” Being responsible for the creation or development of offending content is sufficient for liability.

Naturally, the state-law consumer protection/unfair competition claims also survived.

No comments: