Opinion
Case No.: 07-cv-2198 RMW
09-06-2012
In re: EBAY LITIGATION
ORDER GRANTING DEFENDANT'S
MOTION FOR SUMMARY JUDGMENT
Plaintiffs bring this class action on behalf of users of the popular website eBay.com to recover damages for "search delays," the gap between the time a seller posts an item for sale and the time the listing appears in search results. Plaintiffs allege that defendant eBay, Inc. ("eBay") failed to disclose the existence of such delays, and seek a "pro rata" refund of the fees paid by sellers to list their items. eBay moves for summary judgment, arguing that search delays caused no actual harm, and in the alternative, that a pro rata refund does not provide a reasonable approximation of plaintiffs' alleged injuries. In light of this purported deficiency, eBay seeks dismissal of plaintiffs' claims as a matter of law. For the reasons below, the court grants eBay's motion for summary judgment and dismisses plaintiffs' claims with prejudice.
I. BACKGROUND
eBay is an online marketplace where millions of users buy and sell a nearly infinite assortment of items. When listing an item for sale on eBay, sellers select various options using the "Sell Your Item" form, including the selling format and listing duration. Hartrich Decl. ¶ 3. The most common selling formats are "auction-style" and "fixed price." Id. For a fixed-price listing, the seller sets a designated price (the "Buy It Now" price) at which a buyer can purchase an item immediately. Id. For an auction-style listing, the seller sets a starting price for bidding and may also select a reserve price, which is the minimum amount that the seller is willing to accept, and a "Buy It Now" price. Id. ¶ 4. Sellers in both formats may choose 1, 3, 5, 7 or 10-day listings (the "listing period"). Id. ¶ 6.
Since September 2008, eBay has also offered 30-day fixed-price listings. See Hartrich Decl. ¶ 6.
In order to list an item for sale, sellers must pay an "insertion fee." Id. ¶ 8. For fixed-price sales, the insertion fee is the same regardless of the duration of the listing. Id. For auction-style sales, the insertion fee is the same for 1, 3, 5 and 7-day listings, but sellers pay a $0.40 upgrade charge for 10-day listings. Id.
Sellers also select a listing start time. The default option is "start listing when submitted." However, even when a seller chooses this option, the listing is not instantly "keyword searchable" by prospective shoppers. All listings are subject to brief "search delays" while being automatically indexed into eBay's search database. See Jackman Decl., Ex. B at 28-29. In addition, some listings are subject to longer search delays as a result of eBay's "Trust and Safety" process, under which eBay screens certain listings for fraud and other violations of the website's rules. Id. at 39-40. According to eBay, the average length of a Trust and Safety search delay is six hours. Prowse Decl. ¶ 8. Such delays occur only at the beginning of the listing period. Sehlmeir Decl. ¶ 3. Thus, a one-day listing for a teddy bear would last twenty-four hours, but shoppers searching for "teddy bears" on eBay's website might only be able to locate that particular teddy bear during the last eighteen hours of the listing period.
Statements on eBay's website indicate that some search delays may last twenty-four hours or longer. See Eng Decl., Ex. 5 at 5.
In April 2002, eBay began sending sellers a confirmation email after they submitted a listing indicating that "it may take up to several hours for your listing to appear in eBay search results." See Jackman Decl., Ex. F at 17. eBay did not include a similar disclaimer on the "Sell Your Item" form until March 2007. See id at 16.
If a listing period ends without a sale, sellers may relist their items by choosing the "Relist" option from a drop down menu next to the unsold item. Hartrich Decl. ¶ 11. If the second listing is successful, eBay automatically issues an "insertion fee credit" to the seller's account in the amount of the original insertion fee (the "relisting credit program"). See id., Ex. A. The relisting credit program was available to all sellers until sometime in mid-2008, but is currently offered only for single-item, auction-style listings. Id. ¶ 11.
A. Plaintiffs' Damages Theory
Plaintiffs filed this action on April 20, 2007, alleging that the existence of search delays gave rise to claims for breach of contract, unfair competition and false advertising. Plaintiffs argued that damages could be assessed on a classwide basis by calculating a pro rata refund of the insertion fees paid by sellers. For instance, if a seller paid a $1.00 insertion fee for a one-day listing that was subject to a six-hour search delay, he or she would be entitled to $0.25 in damages.
On October 25, 2010, the court certified a class of "all eBay customers, beginning April 20, 2003, who advertised a listing for sale on eBay.com using the Sell Your Item form and using the 'start listing when submitted' default selection for listing start time." Dkt. No. 134 at 19. The court excluded sellers who successfully sold items at the "Buy It Now" price, reasoning that "[e]ven if no delay took place, a 'Buy It Now' listing may terminate before the end of the chosen duration, simply because the item sold at the 'Buy It Now' price. Thus, plaintiffs' damages methodology, which assumes that bargained-for listing time is lost due to a delay, cannot be applied to 'Buy It Now' listings where the item is successfully sold at the 'Buy It Now' price." Id. at 16.
With respect to other sellers, the court found that plaintffs' damages theory was "plausible" as a basis for class certification, but did not reach any conclusion as to its merits. See id. In subsequent proceedings, the parties agreed that "getting the damages issue resolved" would be the next step in the case. Jackman Decl., Ex. J at 21.
On January 27, 2012, eBay filed the instant motion, arguing that the "pro rata theory" fails to present a viable method for calculating classwide damages. See Mot. at 1. In support of its motion, eBay submitted the expert report of Dr. Stephen Prowse, the declarations of two eBay employees and other evidence. Pursuant to the court's scheduling order, plaintiffs had nearly five months to conduct expert and fact discovery before submitting their response. See Dkt. No. 158. Despite that opportunity, plaintiffs did not engage an expert witness to validate their damages model. In their opposition motion, which was filed on July 9, 2012, plaintiffs imply that expert testimony is unnecessary because "the viability of [their] damages theory is, first and foremost, a legal question." Opp. at 1.
II. DISCUSSION
In order to sustain a claim for breach of contract, a plaintiff must show "appreciable and actual damage." Aguilera v. Pirelli Armstrong Tire Corp., 223 F.3d 1010, 1015 (9th Cir. 2000); Patent Scaffolding Co. v. William Simpson Constr. Co., 256 Cal. App. 2d 506, 511 (Cal. Ct. App. 1967) ("A breach of contract without damage is not actionable."). Similarly, a claim for restitution under the UCL and FAL must be based on a "specific amount found owing." Colgan v. Leatherman Tool Group, Inc., 135 Cal. App. 4th 663, 699 (Cal. Ct. App. 2006). The fact that the amount of damages is not "susceptible of exact proof" does not necessarily bar recovery. Long Beach Drug Co. v. United Drug Co., 13 Cal. 2d 158, 174 (Cal. 1939). However, even if actual damages are difficult to ascertain, a plaintiff must provide "substantial evidence" showing that his measure of damages reasonably approximates the sum necessary to "restore [him] to the status quo ante." Colgan, 135 Cal. App. 4th at 700 (overturning a restitution order where plaintiffs failed to offer evidence establishing "either the dollar value of the consumer impact or the advantage realized by [the defendant]"); compare Acree v. General Motors Acceptance Corp., 92 Cal. App. 4th 385, 398 (Cal. Ct. App. 2001) (affirming an approximation of damages based on expert testimony and other evidence providing the jury "with a reasonable basis for computing damages resulting from [the defendant's breach of contract]").
As a preliminary matter, the parties hotly dispute the nature of their contractual arrangement. Plaintiffs contend that the agreement includes "the specific listing duration offered in the Sell Your Item form and chosen by each member of the class." Opp. at 4. In other words, plaintiffs claim that they paid insertion fees for the ability to make their products available to consumers for a particular amount of time. eBay argues that a listing is simply "an agreement to provide sellers an opportunity to sell their listed items." Rep. at 3. eBay thus decries the notion that search delays, without more, deprive plaintiffs of an expected benefit.
The court finds that it is not necessary to resolve this issue because even under plaintiffs' interpretation, there is no evidence that a pro rata refund is a fair estimate of plaintiffs' actual harm. For sellers who achieved a successful sale, plaintiffs did not even attempt to show, for example, that the delays impacted the price at which the item ultimately sold, the success of future sales, or the seller's business in general. Indeed, the only evidence plaintiffs submitted as to damages is the declaration of class representative Steven Pollack, who indicated that he was harmed by "delays which prevented consumers from shopping, searching and browsing for my item, which diminished the visibility of my company's listing for the promised duration, and reduced my company's access to the eBay marketplace." Eng Decl., Ex. 1 ¶ 10. This conclusory statement says nothing about Mr. Pollack's actual damages, let alone those allegedly sustained by other class members, and is thus insufficient to support plaintiffs' claims. See Ware v. Rodale Press, Inc., 322 F.3d 218, 226 (3d Cir. 2003) (affirming dismissal of contract claim where evidence of damages evinced "little more than the opinion of [the plaintiff]"); Peterson v. Cellco Partnership, 164 Cal. App. 4th 1583, 1591 (Cal. Ct. App. 2008) (affirming dismissal of UCL claim where plaintiffs failed to show "actual economic injury"). Without expert testimony or other evidence demonstrating that sellers who sold their listed items would have fared better in the absence of search delays, the court cannot submit plaintiffs' damages theory to a jury. See Weinberg v. Whatcom County, 241 F.3d 746, 751 (9th Cir. 2001) ("[S]ummary judgment is appropriate where appellants have no expert witnesses or designated documents providing competent evidence from which a jury could fairly estimate damages.").
For sellers who failed to sell a listed item, this case presents a closer question. Plaintiffs urge the court to compare such sellers to an individual who pays a newspaper to run a classified advertisement for five days, but receives only four. Under similar circumstances, at least one California court has awarded a pro rata refund of the contract price. See Ammari Elecs v. Pac. Bell Directory, 2011 Cal. App. Unpub. LEXIS 8800 (Cal. Ct. App. Nov. 15, 2011) (upholding a pro rata refund for plaintiffs who paid to place ads in telephone directories for a specified period of time where "due to massive distribution failures, a substantial percentage of directories were either not delivered or were delivered significantly late").
While this analogy has some appeal, it ultimately fails to persuade. First, any inference that search delays caused the item not to sell is undermined by uncontroverted evidence that approximately of listings subject to search delays resulted in a successful sale—a rate of success that is higher than listings that involved no delay. See Prowse Decl. ¶ 7. Second, there is a significant difference between a classified ad and an eBay listing because, as eBay puts it, "all hours of a listing do not have equal value." Rep. at 9-10. eBay's expert report shows that only about of bids are placed in the first six hours of a listing, the average length of a "Trust and Safety" search delay, while of bids occur in a listing's last six hours and of bids occur in the final hour alone. See Proswse Decl. ¶¶ 34, 37. This is because eBay's default search algorithm prioritizes both auction-style and fixed-price listings as they near the end of the listing period. See id. ¶¶ 26-28. Thus, unlike a classified ad, which has the same probability of being seen by a potential customer throughout its duration, an eBay listing receives significantly more exposure towards the end of its duration. Put another way, the initial hours of a listing, which is the only time search delays occur, offer far less "advertising value" than a listing's final hours. Therefore, if the court were to award a refund to unsuccessful sellers on a pro rata basis—a theory that assumes the hours lost to search delays are worth the same amount as the other hours in a listing— it would effectively be granting plaintiffs a windfall. See, e.g., Baggett v. Hewlett-Packard Co., 2009 U.S. Dist. LEXIS 95241, at *10 (C.D. Cal. Sept. 29, 2009) (dismissing UCL claim where plaintiff sought damages "beyond that which he was promised").
In March 2008, eBay apparently changed its default search prioritization to "Best Match." See Prowse Decl. ¶ 27. At oral argument, eBay indicated, and plaintiffs did not dispute, that a listing's end time remains one of the most important factors in prioritizing search results.
Plaintiffs also failed to rebut eBay's claim that, at least for eligible sellers, the relisting credit program "fully addresses plaintiffs' purported concerns." Mot. at 18. As these sellers would, in essence, receive two listings for the price of one, a pro rata refund would unquestionably provide them with more than they bargained for.
--------
The fluctuating incremental value of an eBay listing renders the instant case distinguishable from plaintiffs' cited authorities, in which courts have awarded a pro rata refund for partial performance of a contract where the parties attributed a specific value to particular increments of performance. See, e.g., American Surety Co. v. United States, 368 F.2d 475, 479 n. 8 (9th Cir. 1966) (affirming an award of restitution for part performance of a drilling contract based on the price of "$9 per linear foot fixed by the contract"); E. A. Robey & Co. v. City Title Ins. Co., 261 Cal. App. 2d 517, 523 (Cal. Ct. App. 1968) (finding that the plaintiff was entitled to an award of $8,000, "that portion of the purchase price attributable to the property for which it received no title"). Of course, had plaintiffs retained an expert, they might nevertheless have been able to establish a "reasonable approximation" of the value of the time lost to search delays. See Acree, 92 Cal. App. 4th at 398. But given that plaintiffs have relied on their pro rata theory for at least two years and were afforded five months to respond to eBay's motion, their inability to provide any proof that a pro rata refund is a reasonable estimate of their damages is fatal.
In opposition, plaintiffs argue that the court should simply exercise its discretion to "fashion remedies to prevent, deter and compensate for" violations of the UCL and FAL. Opp. at 8 (quoting Cortez v. Purolator Air Filtration Products Co., 23 Cal. 4th 163, 176 (Cal. 2000)). However, as the California Court of Appeal has recently held, a court's discretion to grant equitable relief under those statutes "does not extend beyond the boundaries of the parties' evidentiary showing." Colgan, 135 Cal. App. 4th at 700. Having failed to produce evidence in support of their damages calculation, plaintiffs cannot rely on general equitable principles to sustain their claims.
Plaintiffs also contend that evidence of actual damages is unnecessary because nominal damages are available in California for breach of contract. While this argument has some support under California law, see, e.g., Sweet v. Johnson, 169 Cal. App. 2d 630, 632 (Cal. Ct. App. 1959), it has been expressly rejected by the Ninth Circuit. See Aguilera, 223 F.3d at 1015 (a breach of contract claim in California requires "appreciable and actual damage"). In a recent unpublished decision, the Court of Appeal noted this tension but held that "Aguilera binds us, as it did the district court." Ruiz v. Gap, Inc., 380 Fed. Appx. 689, 692 (9th Cir. 2010) (affirming grant of summary judgment for failure to show damages in a breach of contract action). Although the court finds this issue somewhat troubling, it is also bound to follow Augilera. Moreover, the class was explicitly certified on plaintiffs' pro rata theory, and given the extensive focus on damages during the last several years of litigation, the court is hesitant to consider a new measure of damages now. Accordingly, the court declines to sustain this action on the basis of nominal damages alone.
III. ORDER
For the foregoing reasons, the court grants eBay's motion for summary judgment and dismisses plaintiffs' claims with prejudice.
_______________
Ronald M. Whyte
United States District Judge