dotStrategy, Co. v. Twitter Inc

• Court: District Court, N.D. California
• Decided: August 3, 2020
• Docket: 3:19-cv-06176
• Status: Unknown

Opinion

1 2 3 4 5 IN THE UNITED STATES DISTRICT COURT 6 FOR THE NORTHERN DISTRICT OF CALIFORNIA 7 8 DOTSTRATEGY CO, Case No. 19-cv-06176-CRB

9 Plaintiff, ORDER GRANTING IN PART AND 10 v. DENYING IN PART MOTION TO DISMISS 11 TWITTER INC., 12 Defendant.

13 Twitter, Inc., promises advertisers on its platform that they will only be charged when 14 “people” interact with the accounts or Tweets they are paying to promote. DotStrategy, Co., 15 believes it was charged for interactions with automated accounts (“bots”) and that Twitter failed to 16 refund it for those interactions even after it learned that the bot accounts were not, in fact, 17 controlled by “people.” DotStrategy has sued Twitter under California’s Unfair Competition Law. 18 Twitter moves to dismiss. The motion is denied as to dotStrategy’s allegations based on 19 interactions with bots. The First Amended Complaint adequately alleges that dotStrategy suffered 20 economic injury as a result of its reliance on Twitter’s false representation that advertisers would 21 only be charged for interactions with “people.” However, dotStrategy has not adequately alleged 22 that it was wrongfully charged for interactions with “fake” accounts that were nonetheless 23 controlled by people. Twitter’s motion is granted as to those allegations. 24 I. BACKGROUND 25 “Twitter is a social networking and microblogging service, enabling registered users to 26 read and post short messages called Tweets.” FAC (dkt. 58) ¶ 1. Twitter does not make money 27 by charging users for access to the platform. Id. ¶ 4. Instead, it sells advertising. Id. Advertisers 1 Twitter charges advertisers based on how many times users interact with the promoted 2 account or content. Id. ¶ 8. At various times it has represented that advertisers pay only for 3 interactions with “people.” Id. For example, in 2013, Twitter represented to advertisers that they 4 would “only be charged when people follow your Promoted Account or retweet, reply, favorite or 5 click on your Promoted Tweets.” Id. ¶ 37(c). Similarly, in 2014, Twitter claimed that advertisers 6 would “[p]ay only when people follow[ed] [their] account.” Id. ¶ 38(c). 7 DotStrategy is a marketing company which has advertised its services on Twitter. Id. ¶ 21. 8 Between October 2013 and December 2016, dotStrategy placed thirty-four ads on Twitter for 9 which it paid a total of $2,220.76. Id. ¶ 36. DotStrategy alleges that when it placed its ads, it 10 reviewed and relied on Twitter’s representations that advertisers would only be charged for 11 interactions with “people.” Id. ¶¶ 39, 75. 12 When it first began advertising with Twitter, dotStrategy agreed to the Twitter Advertising 13 Terms. Huffman Decl. (dkt. 67) ¶ 3.1 The Advertising Terms include two provisions relevant 14 here. First, they state that Twitter “[t]o the fullest extent permitted by law . . . disclaim[s] all 15 guarantees regarding . . . quality . . . of . . . any User Actions . . . .” Huffman Decl. Ex. B (dkt. 67- 16 2) ¶ 9. Second, they explain that “[c]harges are solely based on [Twitter’s] measurements for the 17 Program.” Huffman Decl. Ex. B ¶ 11. 18 A large number of accounts on Twitter are primarily controlled by bots rather than human 19 beings. FAC ¶ 9. In July 2018, Twitter deleted 70 million accounts “it had deemed spammy, 20 inactive, or which were displaying ‘erratic’ behavior that indicated they were likely bots.” Id. ¶ 49 21 1 DotStrategy does not oppose Twitter’s request for judicial notice of the Advertising Terms, and 22 Twitter correctly notes that the FAC incorporates the Advertising Terms by reference because it implicates the parties’ rights and duties under that document. See Coto Settlement v. Eisenberg, 23 593 F.3d 1031, 1038 (9th Cir. 2010). Twitter’s request for judicial notice of the Advertising Terms is therefore granted. See RJN (dkt. 68). The other documents Twitter requests notice of 24 are either “not subject to reasonable dispute” because their veracity “can be accurately and readily determined from sources whose accuracy cannot reasonably be questioned[,]” see Fed. R. Evid. 25 201(b), (b)(2); see also Moore v. Apple, Inc., 73 F. Supp. 3d 1191, 1197 & n.1 (N.D. Cal. 2014) (materials available online subject to judicial notice); Erickson v. Neb. Mach. Co., No. 15–cv– 26 1147–JD, 2015 WL 4089849, at *1 n.1 (N.D. Cal. July 6, 2015) (materials available on the Wayback Machine subject to judicial notice), or incorporated by reference because they are quoted 27 in or implicated by the FAC, see Coto, 593 F.3d at 1038; Daniels-Hall v. Nat’l Educ. Ass’n, 629 1 (quoting another source). Around the same time, 480 of dotStrategy’s Twitter followers were 2 deleted. Id. ¶ 50. After a Twitter account has been deleted, it is “as if the account never existed,” 3 making it difficult or impossible to find information about the account. Id. ¶ 51 (quoting another 4 source). 5 DotStrategy believes that Twitter wrongfully charged it for interactions with “fake 6 accounts that often [took] the form of an automated bot.” Id. ¶¶ 16–18. It has brought suit 7 claiming that Twitter’s misrepresentations violated the UCL. Id. ¶¶ 72–87. Twitter moves to 8 dismiss. See Mot. (dkt. 65). 9 II. LEGAL STANDARD 10 Pursuant to Federal Rule of Civil Procedure 12(b)(6), a complaint may be dismissed for 11 failure to state a claim upon which relief may be granted. Dismissal may be based on either “the 12 lack of a cognizable legal theory or the absence of sufficient facts alleged under a cognizable legal 13 theory.” Godecke v. Kinetic Concepts, Inc., 937 F.3d 1201, 1208 (9th Cir. 2019). A complaint 14 must plead “sufficient factual matter, accepted as true, to ‘state a claim to relief that is plausible on 15 its face.’” Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009) (quoting Bell Atlantic Corp. v. Twombly, 16 550 U.S. 544, 570 (2007)). A claim is plausible “when the plaintiff pleads factual content that 17 allows the court to draw the reasonable inference that the defendant is liable for the misconduct 18 alleged.” Id. When evaluating a motion to dismiss, the Court “must presume all factual 19 allegations of the complaint to be true and draw all reasonable inferences in favor of the 20 nonmoving party.” Usher v. City of Los Angeles, 828 F.2d 556, 561 (9th Cir. 1987). “[C]ourts 21 must consider the complaint in its entirety, as well as other sources courts ordinarily examine 22 when ruling on Rule 12(b)(6) motions to dismiss, in particular, documents incorporated into the 23 complaint by reference, and matters of which a court may take judicial notice.” Tellabs, Inc. v. 24 Makor Issues & Rights, Ltd., 551 U.S. 308, 322 (2007). 25 Claims for fraud must meet the pleading standard of Federal Rule of Civil Procedure 9(b), 26 which requires a party “alleging fraud or mistake [to] state with particularity the circumstances 27 1 constituting fraud or mistake.”2 Rule 9(b) “requires . . . an account of the time, place, and specific 2 content of the false representations as well as the identities of the parties to the 3 misrepresentations.” Swartz v. KPMG LLP, 476 F.3d 756, 764 (9th Cir. 2007) (internal quotation 4 marks omitted). 5 If a court does dismiss a complaint for failure to state a claim, it should “freely give leave 6 [to amend] when justice so requires.” Fed. R. Civ. P. 15(a)(2). A court nevertheless has 7 discretion to deny leave to amend due to “undue delay, bad faith or dilatory motive on the part of 8 the movant, repeated failure to cure deficiencies by amendments previously allowed, undue 9 prejudice to the opposing party by virtue of allowance of the amendment, [and] futility of 10 amendment.” Leadsinger, Inc. v. BMG Music Publ’g., 512 F.3d 522, 532 (9th Cir. 2008) 11 (alteration in original) (citing Foman v. Davis, 371 U.S. 178, 182 (1962)). 12 III. DISCUSSION 13 Twitter advances four arguments for dismissal. First, it argues that dotStrategy has not 14 alleged false statements or misrepresentations with the specificity required by Rule 9(b). Mot. 15 at 9–13. Second, it argues that dotStrategy has failed to adequately plead reliance on any 16 ostensible misstatements. Id. at 14–17. Third, it contends that dotStrategy’s claims are precluded 17 by disclaimers in the Advertising Terms. Id. at 13–14. Fourth, it argues that dotStrategy has 18 failed to establish UCL standing. Id. at 17–18. 19 A. False Statements 20 Twitter complains that the FAC fails to satisfy Rule 9(b) because dotStrategy alleges that 21 Twitter wrongfully charges for “fake,” “false,” or “spam” accounts without adequately defining 22 those terms. Id. at 10–13. Twitter is correct. DotStrategy clearly considers “bot” accounts (i.e., 23 those controlled entirely or primarily by a computer program rather than a human being) to be 24 fake. See FAC ¶ 17. But, as Twitter points out, and as counsel for dotStrategy confirmed during 25 oral argument, the FAC also alleges that a broader category of human-controlled Twitter accounts 26 are fake, without identifying the outer boundaries of this group. See, e.g., id. ¶ 63 (stating that 27 1 “fake Twitter Accounts . . . include[e] but [are] not limited to automated bots.”). Although the 2 FAC includes some vague suggestions about what makes a human-controlled account fake, see, 3 e.g., id. ¶¶ 17, 74, 76, 80, its allegations are insufficient to put Twitter on notice as to which of its 4 statements were ostensibly false. 5 DotStrategy responds that the terms “fake,” “false,” and “spam” cannot be insufficiently 6 precise, because Twitter itself has used those words to describe activity forbidden on its platform. 7 Opp’n (dkt. 70) at 5–6. This argument fails. The Twitter communications dotStrategy cites may 8 state that “fake” or “spam” accounts are not allowed, see id., but because they do not explain what 9 those terms mean, they do nothing to help satisfy Rule 9(b). DotStrategy suggests its lack of 10 precision should be excused because what makes an account “fake,” “false,” or “spam” in 11 Twitter’s eyes is a “matter[ ] within the opposing party’s knowledge.” Opp’n at 7 (quoting Moore 12 v. Kayport Package Express, Inc., 885 F.2d 531, 540 (9th Cir. 1989)). But the fact that Twitter 13 knows what it means when it uses these terms does not excuse dotStrategy’s obligation to identify 14 the categories of interactions it was wrongfully charged for. 15 Any theory of liability premised on interactions with human-controlled accounts fails for 16 the additional reason that the FAC identifies no statement promising that Twitter advertisers 17 would not be charged for interactions with “fake” accounts that were controlled by people. The 18 only alleged misstatements that could possibly be construed as making such a promise are general 19 proclamations about the benefits of advertising with Twitter. See, e.g., FAC ¶¶ 27, 38 (promising 20 that advertising on Twitter can “build an engaged audience to amplify your message” and “an 21 active community of advocates and influencers for your business”). Even assuming these claims 22 are more than non-actionable puffery, but see Reply (dkt. 71) at 9–10, they could not reasonably 23 be construed as a promise that advertisers would not be charged for engagements with human- 24 controlled accounts. A reasonable advertiser would understand that achieving its goals might 25 require some interaction with Twitter users who use the platform to disseminate spam, violate 26 Twitter’s terms of service, or otherwise qualify as “fake” despite being human. This is especially 27 true because according to dotStrategy’s own allegations, Twitter is rife with such users. See FAC 1 Finally, at oral argument, dotStrategy’s counsel suggested that a reasonable advertiser 2 would understand the word “people” to mean people who abide by Twitter’s rules. The Court 3 rejects this argument. It is aware of no definition of personhood that is contingent on abiding by a 4 corporation’s terms of service. 5 However, dotStrategy adequately alleges that Twitter falsely represented that advertisers 6 would not be charged for interactions with bots. The FAC identifies numerous statements to the 7 effect that advertisers would only be charged for interactions with “people.” See, e.g., FAC 8 ¶ 37(c) (“You’ll only be charged when people follow your Promoted Account or retweet, reply, 9 favorite or click on your Promoted Tweets.”). A reasonable advertiser would understand these 10 statements to mean that they would not be charged—or would be offered a refund—for 11 interactions Twitter knew involved an automated account. Unsurprisingly, Twitter concedes that 12 “these statements could be interpreted as a promise that Twitter will not charge for known bot- 13 engagement.” Mot. at 17. 14 And contrary to Twitter’s position, see Reply at 1, the FAC adequately alleges that 15 dotStrategy was charged and has not been refunded for interactions with bot accounts. In 16 particular, it alleges that dotStrategy’s @buzznnames Twitter account lost 480 (roughly 17%) of 17 its followers in the twenty-eight days preceding July 20, 2018. FAC ¶ 50. The FAC further 18 alleges that Twitter embarked on a campaign to purge its platform of bots during the same time 19 period, leading to the deletion of 70 million accounts, id. ¶ 49, and that a large number of 20 automated accounts are active on Twitter, id. ¶¶ 9, 56(a). Taken together, those facts plausibly 21 allege that at least some of the 480 deleted accounts must have been bots, that dotStrategy most 22 likely paid for interactions with some of those bots, and that Twitter failed to reimburse the money 23 paid for those interactions despite knowing they involved automated accounts. See id. ¶¶ 21, 48, 24 52. 25 True, dotStrategy has not identified what interactions with the 480 deleted accounts it was 26 charged for or which deleted accounts were bots. But that imprecision must be excused because 27 the information that could fill the gap is a “matter[ ] within the opposing party’s knowledge.” 1 provide data about the account after the deletion date. It is as if the account never existed.”3 FAC 2 ¶ 51 (quoting another source). Because only Twitter has access to the data that would allow 3 dotStrategy to identify the bot accounts it was wrongfully charged for, the FAC’s failure to 4 identify those accounts with particularity is not dispositive. See Rubenstein v. Neiman Marcus 5 Grp. LLC, 687 F. App’x 564, 568 (9th Cir. 2017). Even Twitter agrees that dotStrategy need not 6 “identify each account for which it believes it was improperly charged, and the precise date and 7 time . . . it was improperly charged for an advertising interaction by that account.” Reply at 5–6 8 (emphasis in original). 9 Twitter’s motion to dismiss is granted as to dotStrategy’s allegations that it was wrongfully 10 charged for interactions with “fake” accounts controlled by humans. But it is denied with respect 11 to allegations that dotStrategy was wrongfully charged for accounts controlled by bots. The rest 12 of this Order therefore addresses Twitter’s other arguments only as applied to the latter theory of 13 liability. 14 B. Reliance 15 Twitter asserts that dotStrategy fails to adequately allege reliance on the misrepresentations 16 regarding interactions with automated accounts. Mot. at 14–17. But the FAC identifies specific 17 false representations made by Twitter, see FAC ¶¶ 37–38, and alleges both that “[o]n or about 18 October 9, 2013, through December 3, 2016, Plaintiff reviewed various representations by Twitter 19 on its Promoted Products, including the representations identified [earlier in the FAC,]” id. ¶ 39, 20 and that it “relied” on those representations, id. ¶ 75. As dotStrategy correctly notes, this Court 21 has held virtually identical allegations adequate to allege reliance on false advertising in a case 22 involving UCL claims.4 See Bronson v. Johnson & Johnson, Inc., No. C 12–04184 CRB, 2013 23 WL 5731817, at *6 (N.D. Cal. Oct. 22, 2013) (holding allegations that purchasers relied on a 24

25 3 It makes no difference if dotStrategy could have accessed some of this information at some point before filing the instant suit. DotStrategy could not have known that it would need access to 26 this information for litigation before it became aware that it had been wrongfully charged for interactions with automated accounts. 27 4 Twitter argues Bronson is distinguishable because dotStrategy does not allege it relied on the 1 website or packaging that displayed “specific [false] representations” adequate to satisfy 2 Rule 9(b), even though the complaint did “not allege the specific date” when the consumers 3 reviewed the misrepresentations). Other decisions in this district have also held reliance 4 adequately pled when plaintiffs alleged that they reviewed and relied upon specific 5 misrepresentations during a comparable span of time.5 See, e.g., Bruton v. Gerber Prods. Co., No. 6 12–CV–02412–LHK, 2014 WL 172111, at *3, *13 (N.D. Cal. Jan. 15, 2014). 7 Twitter makes much of the fact that dotStrategy agreed to the Advertising Terms in 8 October 2013. Reply at 6–7. Twitter contends that this renders any misrepresentations it made 9 after October 2013 irrelevant. Id. That argument fails for two reasons. First, the FAC alleges that 10 Twitter misrepresented in 2013 that advertisers would not be charged for interactions with bots, 11 FAC ¶ 37(c), so even if statements made after that point are irrelevant, dotStrategy would still 12 adequately allege reliance. Second, the FAC alleges that dotStrategy continued to place ads, 13 thereby incurring additional charges, after October 2013. FAC ¶ 36. Misrepresentations after that 14 date are relevant because dotStrategy could have relied on them in deciding to place additional ads 15 at additional cost. Cases involving misrepresentations made after the plaintiff’s purchase was 16 complete are distinguishable. 17 Finally, to the extent Twitter argues dotStrategy’s reliance was unreasonable, that 18 argument fails for the reasons explained above. A reasonable consumer would understand 19 statements such as “You’ll only be charged when people follow your Promoted Account or 20 retweet, reply, favorite or click on your Promoted Tweets” to mean that Twitter would refund 21 charges for those interactions if it later learned they involved a bot. See FAC ¶ 37(c). 22 C. Contractual Disclaimers 23 Twitter argues that dotStrategy’s suit is precluded by two disclaimers from the Advertising 24 Terms: that “[Twitter] disclaim[s] all guarantees regarding [the] . . . quality . . . of . . . any User 25 Actions” and that “Charges are solely based on our measurements for the Program.”6 Mot. at 13 26 5 Twitter contends these cases are distinguishable because they involve food labeling but does not 27 explain why that factual distinction is legally relevant. See Reply at 7 n.2. 1 (internal citations omitted, internal alterations and emphasis in original). 2 As an initial matter, Twitter’s suggestion that extra-contractual misrepresentations can 3 never give rise to a UCL claim when corrected by contractual disclaimers is false. See Reply at 4 11–12. The Ninth Circuit has recognized that a UCL fraud claim can be based on misleading 5 representations in a solicitation even when the plaintiff later signed a contract with provisions 6 contradicting the earlier falsehoods. See Rubio v. Capital One Bank, 613 F.3d 1195, 1204–06 (9th 7 Cir. 2010). As a result, a plaintiff may state a claim for fraud under the UCL while failing to plead 8 a breach of contract claim. See id. Contrary to Twitter’s reading of the case, Reply at 11 n.6, 9 Rubio involved a fraud claim under the UCL, not just truth-in-lending laws, see 613 F.3d at 1204, 10 and has been extended to facts virtually identical to those presented here by courts in the Northern 11 District of California, see In re Facebook PPC Advert. Litig., No. 5:09–cv–03043–JF, 2010 WL 12 5174021, at *7–10 (N.D. Cal. Dec. 15, 2010). The question, then, is not whether Twitter’s 13 contractual terms corrected the false statements in its advertising, but whether dotStrategy’s 14 reliance on the false advertising was reasonable even in light of the contractual disclaimers. See 15 id. at *9. 16 DotStrategy’s reliance on Twitter’s misstatements was reasonable despite the contractual 17 disclaimers because the contractual provisions Twitter relies on are not irreconcilable with 18 dotStrategy’s understanding that it would not be charged for interactions with bots. Disclaiming 19 the “quality” of “User Actions” is not a clear warning that those users might not be people. See 20 Mot. at 13. A reasonable advertiser could understand the disclaimer to refer to other aspects of 21 user interactions, such as whether users would retweet promoted content with positive rather than 22 negative messaging. That understanding would be particularly reasonable given Twitter’s other 23 representations guaranteeing that advertisers would not pay for interactions with automated 24 accounts. See FAC ¶ 37(c). Similarly, that Twitter has discretion to determine charges for 25

26 Promoted Products credits within the time period required under Section 11 below.” Mot. at 13. The “time period required” is sixty days from the date of the charge. Reply at 12 n.8. Because 27 Twitter does not argue that dotStrategy’s claims must be referred to this internal arbitration 1 advertising does not mean it can charge for interactions with automated accounts when it explicitly 2 promised not to do so elsewhere. An advertiser could reasonably believe that Twitter would 3 determine the amount of advertising charges in a manner consistent with its other representations. 4 In short, Twitter cannot rely on ambiguous disclaimers to correct unambiguous 5 misrepresentations. 6 D. UCL Standing 7 A private plaintiff only has standing to bring a UCL claim if it “has suffered injury in fact 8 and has lost money or property as a result of the unfair competition.” Cal. Bus. & Prof. Code 9 § 17204. Standing under the UCL is therefore limited to plaintiffs who have suffered economic 10 harm, rendering it meaningfully narrower than Article III standing. Kwikset Corp. v. Superior 11 Court, 246 P.3d 877, 886 (Cal. 2011). 12 Twitter advances three arguments that dotStrategy has failed to meet this standard. Two 13 are redundant of arguments considered above. First, Twitter argues that “because the FAC does 14 not adequately explain what Twitter’s alleged unlawful activity is, . . . it fails to establish that 15 Plaintiff paid money ‘as a result of’ any unlawful activity.” Mot. at 17 (emphasis in original). 16 This argument fails because, as explained above, dotStrategy has adequately alleged that Twitter 17 falsely represented that advertisers would only be charged for interactions with people. Second, 18 Twitter argues that because dotStrategy “does not identify which alleged misrepresentations it 19 relied on” it has failed “to establish the ‘causal connection’ UCL requires.” Mot. at 18 (emphasis 20 in original). This argument fails because dotStrategy has adequately pled reliance. 21 That leaves only Twitter’s contention that dotStrategy has not demonstrated economic 22 injury. See Mot. at 17–18. DotStrategy alleges that it “paid for ads for which it would not have 23 agreed to pay anything at all had it known the truth about Twitter’s misconduct.” FAC ¶ 77. That 24 is sufficient to allege economic injury. See Kwikset Corp., 246 P.3d at 892 (plaintiffs who 25 purchased lockets falsely labeled “Made in U.S.A.” suffered economic injury because they would 26 not have bought the lockets had they known the truth). 27 Twitter argues that dotStrategy received the benefit of its bargain because it gained more 1 assumption that “a public[ly] visible increase in . . . follower count” is the only goal an advertiser 2 || might have in promoting its products on Twitter. See Mot. at 18. Presumably human followers 3 are more valuable to advertisers than automated ones, because humans, unlike bots, sometimes 4 || purchase goods and services. If anything, that difference seems more meaningful than, for 5 example, a product’s domestic origin. See Kwikset Corp., 246 P.3d at 892. 6 || IV. CONCLUSION 7 For the foregoing reasons, the motion to dismiss is granted in part and denied in part. 8 IT ISSO ORDERED. co f 9 Dated: August 3, 2020 CHARLES R. BREYER 10 United States District Judge 11 12

15 16

= 17

Z 18 19 20 21 22 23 24 25 26 27 28