Federal Trade Commission v. Directv, Inc. et al

Northern District of California, cand-4:2015-cv-01129

ORDER by Judge Haywood S. Gilliam, Jr. GRANTING IN PART AND DENYING IN PART {{396}} MOTION FOR JUDGMENT ON PARTIAL FINDINGS; FINDINGS OF FACT AND CONCLUSIONS OF LAW. Case Management Conference set for 9/4/2018 02:00 PM in Oakland, Courtroom 2, 4th Floor.

Interested in this case?

Current View

Full Text

5 1 2 3 4 UNITED STATES DISTRICT COURT 5 NORTHERN DISTRICT OF CALIFORNIA 6 7 FEDERAL TRADE COMMISSION, Case No. 15-cv-01129-HSG 8 Plaintiff, ORDER GRANTING IN PART AND DENYING IN PART MOTION FOR 9 v. JUDGMENT ON PARTIAL FINDINGS; FINDINGS OF FACT AND 10 DIRECTV, INC., et al., CONCLUSIONS OF LAW 11 Defendants. Re: Dkt. No. 396 12 On March 11, 2015, the Federal Trade Commission ("FTC") brought this action alleging Northern District of California United States District Court 13 that Defendants DIRECTV and DIRECTV, LLC (collectively, "DIRECTV" or "Defendant") 14 violated Section 5(a) of the FTC Act, 15 U.S.C. § 45(a), as well as the Restore Online Shoppers 15 Confidence Act ("ROSCA"), 15 U.S.C. §§ 8401 et seq. See Dkt. No. 1 ("Compl."). A court trial 16 was held in August 2017. At the close of the FTC's presentation of its case-in-chief, DIRECTV 17 moved for judgment on partial findings pursuant to Rule 52(c) of the Federal Rules of Civil 18 Procedure, on the ground that the FTC failed to prove essential elements of its claims. Dkt. No. 19 396. Based on the evidence presented at trial, the arguments of the parties and the applicable law, 20 the Court GRANTS IN PART DIRECTV's motion. This memorandum opinion will constitute 21 the Court's Findings of Fact and Conclusions of Law pursuant to Rules 52(a) and (c). 22 I. BACKGROUND 23 A. Allegations 24 In this action the FTC challenges DIRECTV's advertisement campaign for and sale of its 25 subscription satellite television services. The FTC commenced its investigation in April 2010 26 when it served DIRECTV with a civil investigative demand. The FTC then filed the complaint in 27 this action in March 2015. See Compl. In the complaint, the FTC sought equitable relief, 28 including a permanent injunction and equitable monetary relief. Id. at 11 ("Prayer for Relief"). 5 1 The FTC contends that from 2007 to the time of trial, DIRECTV failed to adequately 2 disclose in its advertising certain terms of purchase for its satellite television services, in violation 3 of § 5 of the FTC Act. In particular, the FTC alleges that DIRECTV failed to adequately disclose 4 that: (1) the introductory discounted price only lasts 12 months, after which the subscriber is 5 charged the then-prevailing rate; (2) the subscriber is subject to a 24-month commitment period; 6 (3) a subscriber who cancels before the end of the commitment period is assessed an early 7 cancellation fee of $20 per month for the remaining months in the commitment period; and 8 (4) subscribers receive a free premium channel package (e.g., HBO, Starz, Cinemax, and 9 Showtime) for three months, but must affirmatively cancel these premium channels before the end 10 of the three-month period to avoid monthly charges (the "premium channel negative option"). See 11 id. ¶¶ 12–14. The FTC further contends that DIRECTV failed to clearly and conspicuously 12 disclose the premium channel negative option on its website, and that it also failed to obtain Northern District of California United States District Court 13 express informed consent before charging consumers, in violation of ROSCA. See id. ¶¶ 35–42. 14 Based on these claims, the FTC seeks restitution in the amount of $3.95 billion. Dkt. No. 15 362 ("Tr. Vol. 1") at 33:5–16 (transcript of FTC's opening statement). 16 B. Procedural Posture 17 This matter was tried to the Court, sitting without a jury, from August 14, 2017, to August 18 25, 2017. After the close of the FTC's case-in-chief, DIRECTV moved orally for Judgment on 19 Partial Findings pursuant to Federal Rule of Civil Procedure 52(c). The Court suspended the trial 20 before DIRECTV presented its defense to allow DIRECTV to file its motion, as well as the 21 parties' respective proposed findings of fact and conclusions of law. See Dkt. No. 390 ("Tr. Vol. 22 10") at 1708:25–1713:6. 23 II. LEGAL STANDARD 24 Under Rule 52(c): 25 If a party has been fully heard on an issue during a nonjury trial and 26 the court finds against the party on that issue, the court may enter judgment against the party on a claim or defense that, under the 27 controlling law, can be maintained or defeated only with a favorable 28 2 5 finding on that issue. The court may, however, decline to render any 1 judgment until the close of the evidence. A judgment on partial findings must be supported by findings of fact and conclusions of 2 law as required by Rule 52(a). 3 Fed. R. Civ. P. 52(c). Thus, the Rule "authorizes the court to enter judgment at any time that it 4 can appropriately make a dispositive finding of fact on the evidence." Granite State Ins. Co. v. 5 Smart Modular Techs., Inc., 76 F.3d 1023, 1031 (9th Cir. 1996). In doing so, the Court draws its 6 own conclusions as factfinder and need not draw any inferences in favor of the non-moving party. 7 See Ritchie v. United States, 451 F.3d 1019, 1023 (9th Cir. 2006). As the plaintiff, the FTC bears 8 the burden of proof and must prove each element of its case by a preponderance of the evidence. 9 United States v. F/V Repulse, 688 F.2d 1283, 1284 (9th Cir. 1982). The Court should grant the 10 motion if it becomes apparent that plaintiff has failed to carry an essential burden of proof. Cf. 11 EBC, Inc. v. Clark Bldg. Sys., Inc., 618 F.3d 253, 272 (3d Cir. 2010) ("The rule's objective is to 12 'conserve[] time and resources by making it unnecessary for the court to hear evidence on Northern District of California United States District Court 13 additional facts when the result would not be different even if those additional facts were 14 established.'"). 15 III. JURISDICTION AND VENUE 16 The Court has subject matter jurisdiction pursuant to 28 U.S.C. §§ 1331, 1337(a), and 17 1345, and 15 U.S.C. §§ 45(a), 53(b). Venue is proper in this district under 28 U.S.C. §§ 1391(b), 18 (c)(2), and (d), and 15 U.S.C. § 53(b). 19 IV. FINDINGS OF FACT AND CONCLUSIONS OF LAW 20 A. The Parties 21 The FTC is an independent agency of the United States Government created by statute. 15 22 U.S.C. §§ 41–58. The FTC enforces Section 5(a) of the FTC Act, which prohibits unfair or 23 deceptive acts or practices in or affecting commerce. The FTC also enforces ROSCA, which 24 prohibits certain methods of negative option marketing on the Internet. The FTC is authorized to 25 initiate federal district court proceedings, by its own attorneys, and to seek injunctive relief for 26 violations of the FTC Act and ROSCA. 15 U.S.C. § 53(b). 27 28 3 5 1 DIRECTV is a provider of subscription satellite television services.1 Each of DIRECTV's 2 subscription packages offers a varying selection of channels, equipment, and video quality. See, 3 e.g., Ex. 172 (letter describing several available subscription packages). Over time, the terms of 4 DIRECTV's subscription packages, including the price for each package, have changed 5 periodically. See Ex. 76 (summary of subscription offers between 2008 and 2015, citing trial 6 exhibits describing the terms of each package), Ex. 617 (2014 "New Customer Offer Timeline" 7 graphic showing offer terms between 2008 and 2013). 8 DIRECTV's subscription television services are a "considered purchase," such that 9 consumers generally "go through anywhere between [a] 45- to 90-day process" when deciding 10 whether or not to purchase. Dkt. No. 367 ("Tr. Vol. 2") at 308:16–309:4 (Bentley). The 11 complexities of the product stem from the "sheer number of different tradeoffs and decisions that 12 need to be made" regarding the packages, "advanced receiver, size of the hard drive, movies to Northern District of California United States District Court 13 save," number of rooms, and whether advanced receivers were required in each room. Tr. Vol. 2 14 at 294:19–295:1 (Bentley); Dkt. No. 371 ("Tr. Vol. 4") at 643:10–11 (Guyardo) (DIRECTV is a 15 "high involvement" and "complex" purchase). Consumers who wish to subscribe to DIRECTV 16 must proceed through a lengthy and detailed subscription process, either on the phone with a 17 DIRECTV representative or on the web. See Tr. Vol. 2 at 302:1–310:14 (Bentley); Tr. Vol. 4 at 18 661:20–662:19 (Guyardo); Dkt. No. 372 ("Tr. Vol. 5") at 792:18–21 (Chen). 19 DIRECTV acquires subscribers using multiple sales channels that are broadly categorized 20 as either "Direct Sales" or "Indirect Sales." Tr. Vol. 4 at 667:3–668:9 (Guyardo); Ex. 9 at 44. 21 Direct Sales involves customers who subscribe either by phone through one of DIRECTV's toll 22 free numbers or on the DIRECTV website, directv.com. Tr. Vol. 2 at 298:19–25 (Bentley). The 23 majority of Direct Sales subscribers complete their sign up over the phone, and only a small 24 percentage subscribe on the web. Dkt. No. 389 ("Tr. Vol. 9") at 1441:23–1442:15 (Leever). 25 Indirect Sales are made through a number of third-party channels. Ex. 9 at 44. These 26 27 1 DIRECTV was acquired by AT&T Inc. on July 24, 2015. See Dkt. No. 337 at 17. 28 4 5 1 include Consumer Electronics (consumer electronics and other big box stores like Walmart or Best 2 Buy that sell DIRECTV in their stores, Tr. Vol. 2 at 299:10–23 (Bentley)); Telco (partnerships 3 with telephone companies that sell DIRECTV services on their networks in addition to phone and 4 internet services, Tr. Vol. 2 at 299:2–5 (Bentley)); National Strategic Partners (national third-party 5 dealers, Tr. Vol. 2 at 300:3–7 (Bentley)); Local Strategic Partners (local third-party dealers, Tr. 6 Vol. 2 at 300:9–14 (Bentley)); and Multi-Dwelling Units (multi-unit buildings that contract with 7 DIRECTV to make the service available to residences). Ex. 9 at 44. 8 Many third-party providers, including Telco partners, used their own advertising tactics. 9 See Tr. Vol. 2 at 225:21–226:20, 300:15–301:1 (Bentley); Dkt. No. 388 ("Tr. Vol. 8") at 1328:20– 10 23 (Pratkanis) (referring to Costco tactic). Third-party providers also had offers that were 11 different from Direct Sales offers. Tr. Vol. 2 at 299:6–9 (Bentley). For example, Telco partners 12 offered bundles with their telephone and Internet services and DIRECTV's television services, and Northern District of California United States District Court 13 the Telco partners would offer such services, including DIRECTV's, using their own sales 14 network and cross-selling to their existing customers. Tr. Vol. 2 at 300:15–301:1 (Bentley). 15 Likewise, there were over two thousand National Strategic Partners that would create and place 16 their own advertising for DIRECTV. Tr. Vol. 2 at 225:21–226:20 (Bentley). 17 B. The Challenged Advertisements 18 Between 2007 and the time of the trial, DIRECTV advertised its subscription television 19 service through several different media. DIRECTV aired television commercials, see, e.g., Exs. 20 472–478; disseminated tens of thousands of different print advertisements through newspapers and 21 direct mailings, Dkt. 386 ("Tr. Vol. 6") at 933:20–21 (Erdem) (over 40,000 ads); published over 22 150 different Internet banner ads, Ex. 1161; and used its website (directv.com), which was revised 23 several times during this period, Tr. Vol. 9 at 1448:12–20 (describing website redesigns). 24 i. Television Commercials 25 DIRECTV advertised on television in nationally-broadcast 30-second and one-to-two- 26 minute ads about DIRECTV called "direct response television." Tr. Vol. 2 at 220:8–11 (Bentley). 27 All television ads prompted consumers to call 1-800-DIRECTV and/or visit directv.com. Id. 28 5 5 1 ii. Print Advertising 2 DIRECTV's print advertising has generally included three broad categories: freestanding 3 inserts disseminated with Sunday newspapers (also known as Sunday Circulars), Alternative 4 Media, and Solo Mail. 5 a. Sunday Circulars 6 Sunday Circulars are multi-page advertisements inserted into the Sunday newspaper. Tr. 7 Vol. 2 at 312:7–9 (Bentley); Ex. 9 at 60. DIRECTV has circulated approximately 200–300 8 different types of freestanding newspaper inserts per month. Tr. Vol. 2 at 231:7–10 (Bentley). 9 The number of consumers who called DIRECTV in response to a Sunday Circular is 10 "extremely small." Tr. Vol. 5 at 883:6–17 (Gieselman). By way of example, in response to a 11 Sunday Circular distributed to 7 million consumers in or around August 2013, only three one 12 hundredths of a percent (.03%) of the people who received the Sunday Circular called DIRECTV, Northern District of California United States District Court 13 and only 32% of the people who called ultimately subscribed to DIRECTV during the call. Tr. 14 Vol. 5 at 879:21–883:18 (Gieselman); Ex. 384 at 7. 15 b. Alternative Media 16 Alternative Media includes "hundreds of different tactics" that generally are provided to 17 consumers with material or advertisements from other (and often many other) companies. Tr. Vol. 18 2 at 223:22–224:2, 224:9–13, 313:22–314:10, 315:17–316:17 (Bentley); Ex. 9 at p. 59. 19 Alternative Media advertisements are comprised of newspaper-delivered inserts, shared mail or 20 co-op mailings, mail inserts, or directories. Tr. Vol. 2 at 223:22–224:13 (Bentley); Ex. 9 at p. 59. 21 DIRECTV sent out "anywhere from 2[00] to 300 [Alternative Media] tactics every month" 22 in multiple channels of distribution. Tr. Vol. 2 at 312:5–6, 321:6–10 (Bentley) ("there may be 300 23 different creative versions of [Alternative Media] in a single month. . ."); Ex. 9 at 59. Alternative 24 Media is high-volume, low-return marketing that targets the same households week after week. 25 Tr. Vol. 2 at 320:21–321:10, 322:18-323:16 (Bentley) (one alternative media advertisement 26 disseminated 10 million pieces to generate approximately 300 subscriptions). 27 28 6 5 c. Solo Mail 1 Solo Mail is a multipage letter sent in a standard envelope to targeted, credit pre-approved 2 households. Tr. Vol. 2 at 311:23–312:9, 315:9–16, 321:2–10 (Bentley); Ex. 9 at 58. DIRECTV 3 would use predictive models to determine which households to target. Tr. Vol. 2 at 321:11–19 4 (Bentley). DIRECTV generally sent 15 to 18 million Solo Mail letters per month. Tr. Vol. 2 at 5 320:12–16 (Bentley). 6 iii. Banner Ads 7 Banner ads are animated images that appear on third-party websites. Tr. Vol. 5 at 801:3–6 8 (Chen). When a consumer clicks on the banner ad, she is routed to directv.com. Id. 9 iv. directv.com 10 Consumers may subscribe to DIRECTV's satellite television services through directv.com. 11 Few consumers who enter DIRECTV's web flow complete the purchase on the website. See, e.g., 12 Northern District of California Tr. Vol. 5 at 794:13–18 (Chen) (rate of subscription for prospects who entered the website and United States District Court 13 placed a package in their cart ranged between 6.5 and 8.9%.); Ex. 1146 at 1. 14 Based on unique toll-free numbers on the website and available data regarding visits to 15 directv.com, out of all subscriptions arising from a customer who visited directv.com, 16 approximately 70% of those consumers called a toll-free number and completed their subscription 17 with a telephone customer service representative. Tr. Vol. 2 at 305:14–25 (Bentley). 18 C. The FTC Failed to Prove a Violation of Section 5(a) of the FTC Act as to Any of 19 Defendant's Print, Television or Electronic Banner Advertisements 20 Section 5(a) of the FTC Act prohibits "[u]nfair methods of competition in or affecting 21 commerce, and unfair or deceptive acts or practices in or affecting commerce." 15 U.S.C. § 45(a). 22 A practice is deceptive in violation of § 5(a) if: (1) "there is a representation, omission, or 23 practice"; (2) that "is likely to mislead consumers acting reasonably under the circumstances"; and 24 (3) "the representation, omission, or practice is material." F.T.C. v. Gill, 265 F.3d 944, 950 (9th 25 Cir. 2001). Even if it contains some truthful disclosures, a representation still "may be likely to 26 mislead by virtue of the net impression it creates." F.T.C. v. Cyberspace.Com LLC, 453 F.3d 27 1196, 1200 (9th Cir. 2006); see also Gill, 265 F.3d at 956 (evaluating the "overall 'net 28 7 5 1 impression'" of the representations). And failure to disclose material information may cause an 2 advertisement to be false or deceptive within the meaning of the Act even where the advertisement 3 does not state false facts. Simeon Management Corp. v. F.T.C., 579 F.2d 1137, 1145 (9th Cir. 4 1978). Actual deception is not required for a Section 5 violation. Trans World Accounts, Inc. v. 5 F.T.C., 594 F.2d 212, 214 (9th Cir. 1979). 6 Here, the FTC does not contend that any of the over 40,000 advertisements it purports to 7 challenge contained affirmatively false representations. See Dkt. No. 337 (FTC's section of joint 8 pretrial statement) at 34 ("Because, in a failure to disclose case, the deception does not turn on 9 literal falsity, the FTC need not demonstrate that reasonable consumers are likely to take away 10 some false message about [the] advertised offer."); Dkt. No. 412 (transcript of argument on 11 motion for partial findings) at 23 ("This is an omission case, not a falsity case."). Nor does the 12 FTC contend in any systematic or detailed way that the advertisements entirely omitted the Northern District of California United States District Court 13 information it claims was necessary to ensure that the statements in the advertisements were not 14 misleading. See Tr. Vol. 1 at 8:3–7 (FTC's opening statement) ("When you look at the 15 advertisements, you will see disclosures, if they exist at all, flashing briefly on the screen, hidden 16 behind hyperlinks, buried in fine print, or obscured by dense paragraphs of legal text.") (emphasis 17 added). Instead, the FTC's theory is that Defendant's advertisements were likely to mislead 18 reasonable consumers because, in its view, critical details of Defendant's offer were not disclosed 19 prominently enough. Tr. Vol. 1 at 15:10–13 ("In this case we are talking about omissions; 20 omissions by inadequately disclosing material terms after the advertised price has been 21 prominently featured in DIRECTV's advertisements.") 22 The FTC fails to meet its burden as to Defendant's print, television and electronic 23 advertisements for at least two reasons. First, as to the very small number of advertisements 24 actually analyzed by its expert witnesses, the FTC failed to establish that there was any misleading 25 "net impression," or even to identify clearly what it claims the net impression was. Neither a 26 facial review of these advertisements nor the FTC's extrinsic evidence established that these 27 materials are likely to mislead a reasonable consumer. Second, the FTC also failed to articulate 28 8 5 1 what common net impression is conveyed by the over 40,000 challenged advertisements (which 2 spanned several different formats), or to explain how and why that impression would be likely to 3 mislead a reasonable consumer. 4 i. The FTC's Theory 5 In the complaint, the FTC framed its allegation with regard to the two-year requirement as 6 follows: In numerous instances, in connection with the advertising, 7 marketing, promotion, offering for sale, and sale of Defendants' subscription service, Defendants have represented, directly or 8 indirectly, expressly or by implication, that their programming packages can be purchased by paying the advertised monthly prices, 9 typically for a twelve-month period. In numerous instances in which Defendants have made the 10 representation set forth in Paragraph 28 of this Complaint, Defendants have failed to disclose, or to disclose adequately, to 11 consumers certain material terms and conditions of the offer, including but not limited to: 12 A. The mandatory two-year agreement period, which carries Northern District of California United States District Court an early cancellation fee, for the subscription service; and 13 B. The significantly higher price for programming packages, typically $25 to $45 per month higher, during the mandatory 14 second year of the consumer's agreement. This additional information would be material to consumers in 15 deciding to purchase Defendants' subscription services. 16 Compl. ¶¶ 28–29. 17 As to the premium channel offer terms, the FTC alleged that: 18 In numerous instances, in connection with the advertising, marketing, promotion, offering for sale, and sale of Defendants' 19 subscription service, Defendant have represented, directly or indirectly, expressly or by implication, that consumers could obtain 20 certain premium channels for free for a certain period of time, typically three months. 21 In numerous instances in which Defendants have made the representation set forth in Paragraph 31 of this Complaint, 22 Defendants have failed to disclose, or disclose adequately, to consumers the material terms and conditions related to the costs of 23 the offer, including: A. That Defendants automatically enroll consumers in a 24 negative option continuity plan with significant charges; B. That consumers must affirmatively cancel the negative 25 option continuity plan before the end of the trial period to avoid charges; 26 C. That Defendants use consumers' credit or debit card information to charge consumers for the negative option 27 continuity plan; and 28 9 5 D. The costs associated with the negative option continuity 1 plan. 2 Compl. ¶¶ 31–32. 3 In its pretrial statement, the FTC articulated its theory this way: "DIRECTV's 4 advertisements consistently create an overall 'net impression' that the company's service may be 5 obtained for substantially less than consumers ultimately must pay over the course of their 6 mandatory 24-month contract." Dkt. No. 337 at 3. Now, in its opposition to the motion for 7 judgment on partial findings, the FTC frames its theory as being based on the principle that 8 "prominently advertising a half-truth while omitting or hiding qualifying material information 9 renders the net impression of the advertisements misleading." Dkt. No. 401 at 2 and n.2 10 (emphasis in original) (citing F.T.C. v. AMG Svc's, Inc., 29 F.Supp.3d 1338, 1351 (D. Nev. 2014) 11 and Cyberspace.com, 453 F.3d at 1201). 12 ii. Net Impression Analysis Northern District of California United States District Court 13 In assessing whether the FTC has met its burden of establishing liability under § 5(a), the 14 Court must first determine what "net impression" the advertisements at issue create, so that it can 15 then decide whether this net impression was likely to mislead reasonable consumers. See AMG 16 Svc's, Inc., 29 F.Supp.3d at 1368 n.9 (explaining that "the FTC Act requires the court to determine 17 what 'net impression' the [allegedly misleading] loan note creates"); F.T.C. v. Commerce Planet, 18 Inc., 878 F.Supp.2d 1048, 1063 (C.D. Cal. 2012), aff'd 642 Fed. Appx. 680 (9th Cir. 2016) 19 ("District courts consider the overall, common sense 'net impression' of the representation or act 20 as a whole to determine whether it is misleading."). In determining what net impression an 21 advertisement creates, the Court considers the face of the advertisement, and may also consider 22 extrinsic evidence. See Commerce Planet, 876 F.Supp.2d at 1062–78 (considering face of 23 advertisements as well as extrinsic evidence in assessing whether net impression was misleading; 24 "[a]lthough a facial examination of the sign-up pages sufficiently demonstrates that the website 25 marketing of [company] was misleading to a reasonable consumer, the Court may consider 26 27 28 10 5 1 extrinsic evidence as corroborating evidence"). 2 2 Generally, the FTC can succinctly describe its position as to what the net impression is and 3 explain why that net impression is likely to mislead in its view. See F.T.C. v. Johnson, 96 4 F.Supp.3d 1110, 1139–1140 (D. Nev. 2015) (describing FTC's position that "a consumer 5 proceeding through the website pages would have a net impression that they were getting a free 6 CD for a minimal shipping and handling or download fee," while in truth "the sites collected 7 preliminary information without informing the consumer that they would be enrolled in a core 8 membership with a negative option and additional upsell memberships with negative options"). 9 In attempting to determine the net impression of the advertisements at issue, the Court is 10 confronted with an immediate challenge: the FTC's evidence spans at least four tiers, which 11 together encompass tens of thousands of discrete advertisements used by DIRECTV over an eight- 12 year period. Northern District of California United States District Court 13 First, the FTC's expert witness, Dr. Tülin Erdem, conducted an online survey based on a 14 print advertisement from 2013. Exs. 2374, 2375 (survey URL); Ex. 244 (underlying 15 advertisement); Tr. Vol. 6 at 928:3–934:6. 16 Second, Dr. Erdem testified that she chose this print advertisement for her survey based on 17 a review of 116 advertisements used by DIRECTV between 2007 and 2015, and said she found 18 the chosen ad to be "sufficiently representative." Tr. Vol. 6 at 934:7–21, 956:3–977:20, 996:17– 19 997:10; Dkt. No. 393.3 Dr. Erdem claimed that the results of her survey could be generalized to 20 draw conclusions about other print advertisements beyond the one she tested. Tr. Vol. 6 at 956:8– 21 2 22 Defendant contends, citing two out-of-circuit district court cases, that "courts have consistently held that, where it is not reasonably clear from the face of the advertisement that reasonable 23 consumers take away the alleged implied false message, extrinsic evidence is required to prove deception." Dkt. No. 396 at 6 (citing United States v. Bayer Corp., 2015 WL 5822595 at *11 and 24 *13 (D.N.J. Sept. 24, 2015) and F.T.C. v. Nat'l Urological Grp., Inc., 645 F.Supp.2d 1167, 1193 (N.D. Ga. 2008) (internal quotation marks omitted)). Neither the Ninth Circuit nor any district 25 court in this circuit has construed the requirements of the FTC Act in the manner Defendant urges, and the Court does not find these cases persuasive on this issue. 3 26 Dr. Erdem testified that she reviewed "exactly 116 print ads" in detail. Tr. Vol. 6 at 934:22– 935:2. At the Court's request, the parties submitted a stipulation listing the trial exhibit number 27 for each of the items reviewed by Dr. Erdem, which included 119 print advertisements. Dkt. No. 393. 28 11 5 1 957:10, 986:1–995:21. Another of the FTC's expert witnesses, Dr. Anthony Pratkanis, testified 2 similarly that the characteristics of two print advertisements, Exhibits 238 and 273, were 3 generalizable to "hundreds" of other advertisements he reviewed. Tr. Vol. 8 at 1251:21–1266:3. 4 Third, the FTC introduced a total of approximately 228 print advertisements, 17 television 5 commercials, and 33 electronic banner ads into evidence. Dkt. No. 401-1 (FTC's proposed 6 findings of fact and conclusions of law) at ¶¶ 41, 46, 50, 54, 56, 59, 62–63 (listing exhibit numbers 7 of admitted exhibits in each category). Witnesses testified regarding approximately 47 of these 8 exhibits. A large number—approximately 231—were introduced by stipulation without live 9 witness testimony. Dkt. No. 391. 10 Fourth, and finally, a much larger number of advertisements that were disseminated during 11 the period for which the FTC claims DIRECTV customers are entitled to restitution—over 12 39,000—were never introduced into evidence at all. Tr. Vol. 6 at 933:20–21; cf. Dkt. Nos. 360-1, Northern District of California United States District Court 13 363-1, 364-1, 368-1, 369-1, 376-1, 381-1, 382-1, 383-1, 392-1, 391, 395, 415 (lists of exhibits 14 admitted at trial and by stipulation). 15 All told, the FTC's theory of the case thus requires the Court to attempt to determine the 16 "net impression" of more than 40,000 advertisements, across print, television, and electronic 17 formats. Simply to state this fact is to highlight the extraordinary ambition—and daunting 18 challenges—inherent in the FTC's theory. The district court in Johnson noted similar challenges 19 in that case, where the FTC "ask[ed] the Court to draw conclusions as to all of Defendants' sites, 20 presented to the Court or not, based on the exhibits presented and [its expert's report]": 21 Immediately the Court is confronted with a problem. As previously mentioned, the FTC has provided over one-hundred exhibits of grant 22 websites and the expert report from [expert]. However, for the purposes of summary judgment, the Court cannot rely solely on the 23 representations of the FTC and their expert as to what the websites represent. . . . In its briefing, the FTC picks examples of claims from 24 across its grant website exhibits to suit their arguments. It is not clear to the Court, however, whether the FTC's examples are 25 representative of the grant website experience as a whole. . . . The Court cannot adopt the FTC's approach of using selected examples 26 of claims picked from across the entire universe of the FTC exhibits, seemingly without a clear methodology, and draw conclusions as to 27 every one of Defendants' sites. 28 12 5 1 Johnson, 96 F.Supp.3d at 1118, 1121. 2 The challenges described by the Johnson court are substantially multiplied here, given the 3 enormous number of advertisements at issue, the diversity of format of those advertisements, and 4 the varying nature of the representations made in the advertisements over time. 5 Accordingly, the Court will proceed here in two stages. First, the Court will examine the 6 single print advertisement analyzed by Dr. Erdem to determine whether the net impression created 7 by that advertisement would be likely to mislead consumers acting reasonably under the 8 circumstances. Second, the Court will consider whether, even if the net impression of this 9 advertisement was misleading, that impression can be generalized (1) to any of the other 10 advertisements in evidence or (2) to tens of thousands of other advertisements, the vast majority of 11 which were never admitted in evidence, spanning an eight-year period across different media. 12 iii. Print Advertisements Northern District of California United States District Court 13 a. Analysis of print advertisement tested by Dr. Erdem (Ex. 244)4 14 1. Facial review 15 Exhibit 244, which Dr. Erdem used as the basis for her web-based survey, is a one-page 16 advertisement with the heading "Best Offer Ever" in gold, silver and yellow letters against a blue 17 background: 18 19 20 21 22 23 24 25 26 4 27 This exhibit was admitted by stipulation. Dkt. No. 391. The parties' joint exhibit list describes it as "2013 print ad, offer ends 10/02/2013, 'Best Offer Ever.'" Dkt. No. 353-1 at 34. 28 13 5 1 2 3 4 5 6 7 8 9 10 11 12 Northern District of California United States District Court 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 Ex. 244.5 28 14 5 1 As would be expected for a complex multi-option product like a satellite television 2 subscription, the advertisement contains a large amount of information about various 3 programming packages, pricing and equipment options, product quality and installation logistics. 4 Roughly the middle third of the advertisement describes three tiers of programming packages: the 5 Entertainment Package, the Choice Package and the Xtra Package. Each package includes a 6 different number of channels and a different number of on demand titles. Each package also 7 includes HBO, Starz, Showtime and Cinemax "free for three months." 8 The text box for each package includes a price "for 12 months" "after instant savings." In 9 each instance, that price is crossed out by a red line, and a lower price appears in white text against 10 a red circle with the phrase "Act Now!" at the top. So, for example, the crossed-out price for the 11 Entertainment Package is $29.99/month, and the "Act Now!" price is $24.99/month. The price in 12 the red bubble has a caret next to it, which directs the viewer to the following bolded text in the Northern District of California United States District Court 13 first line of a section of disclosures in black text against a white background at the bottom of the 14 page: "BILL CREDIT/PROGRAMMING OFFER: IF BY THE END OF PROMOTIONAL 15 PRICE PERIOD(S) CUSTOMER DOES NOT CONTACT DIRECTV TO CHANGE 16 SERVICE THEN ALL SERVICES WILL AUTOMATICALLY CONTINUE AT THE 17 THEN-PREVAILING RATES."6 Because most of the text in the disclosure box is not bolded, 18 not in all capital letters, and not partially underlined, this information stands out visually. See Ex. 19 1119 (FTC ".com Disclosures" guidance) at 17 ("A disclosure in a color that contrasts with the 20 background emphasizes the text of the disclosure and makes it more noticeable."). The text in the 21 disclosure box is smaller than most of the text in the advertisement, though type in different sizes, 22 5 23 The Court notes for clarity of the record that the reproduction of Exhibit 244 in this order is substantially smaller than the actual size of the advertisement as it was distributed to potential 24 consumers in paper form. See Exhibit 2026A (paper version of the advertisement measuring approximately 8.5" by 11"). The resolution of the original Exhibits 244 and 2026A is also 25 substantially clearer than the resolution of the reproduction in this order. The Court has inserted a reproduction of Exhibit 244 simply to provide context for this order's narrative description of the 26 advertisement's content. 6 Brad Bentley, who has served as a sales and marketing executive at DIRECTV and then at its 27 parent company AT&T for 17 years, referred to the section at the bottom of the advertisement as the "offer details" section. Tr. Vol. 2 at 217:9–230:25, 238:23–239:3, 265:22–266:3. 28 15 5 1 fonts and colors is used throughout the ad. The next line in the disclosure box lists the current 2 non-promotional package price for each option (for example, the non-promotional price at the time 3 for the Entertainment Package was $54.99 per month). 4 Bolded text in black print against a white background immediately below the three boxes 5 describing the packages explains that "ALL DIRECTV OFFERS REQUIRE 24-MONTH 6 AGREEMENT.**" This same bolded language appears again in smaller text (again black 7 against a white background) in the bottom third of the advertisement, below three smaller boxes 8 describing installation details, HD DVR features and options for bundling TV, internet and phone 9 service. The two asterisks direct the viewer to the following bolded text in the disclosure box at 10 the bottom of the page: "24-MONTH AGREEMENT: EARLY CANCELLATION WILL 11 RESULT IN A FEE OF $20/MONTH FOR EACH REMAINING MONTH." Again, this 12 information is one of the few pieces of text in the disclosure section that is bolded, in all capital Northern District of California United States District Court 13 letters and partially underlined, making the information stand out visually. 14 Near the bottom of the page, above the disclosure section, large text says "Upgrade to 15 DIRECTV! Call 1-866-951-9617 or visit directv.com." 16 Based on its facial review of Exhibit 244, the Court finds that the net impression a 17 reasonable consumer would take away from this advertisement is that a promotional price applies 18 for 12 months, that a 24-month agreement is required, that an early cancellation fee of $20 per 19 month will apply if the customer cancels before the end of the 24-month period, that services will 20 continue at the end of any promotional period at the then-applicable regular rate unless the 21 customer contacts DIRECTV to change the services, and that the customer needs to call a toll-free 22 number or go to DIRECTV's website to subscribe. These provisions are adequately disclosed 23 throughout the advertisement, and while the ad contains a substantial amount of information, a 24 reasonable consumer would understand that this is because subscription satellite television service 25 is a complex product with a number of options for price, level of service, package features and 26 other components. See Tr. Vol. 2 at 308:16–309:7 (subscribing to satellite television services 27 involves a large "number of different tradeoffs and decisions that need to be made," making 28 16 5 1 subscribing a "considered purchase") (Bentley); Tr. Vol. 9 at 1439:14–24 (testimony of former 2 DIRECTV executive Karen Leever that satellite television services are "a complex product" 3 requiring customers to make a number of decisions to customize their package). Because the 4 advertisement adequately discloses the details that the FTC claims were omitted due to a lack of 5 prominent disclosure, the net impression of the advertisement on its face would not be likely to 6 mislead a reasonable consumer. 7 7 2. Extrinsic evidence 8 The Court further finds that none of the extrinsic evidence introduced by the FTC meets its 9 burden of showing that Exhibit 244, or any of the other print, electronic or television 10 advertisements analyzed by the FTC's experts, would be likely to mislead a reasonable consumer. 11 a. Expert Testimony 12 1. Dr. Erdem Northern District of California United States District Court 13 The primary extrinsic evidence upon which the FTC relies in support of its claim is a web- 14 based survey conducted by Dr. Tülin Erdem, a professor of business administration and marketing 15 at New York University. Tr. Vol. 6 at 902:19–21. Dr. Erdem described her assignment as 16 "evaluating whether DIRECTV's website and print advertisements clearly and conspicuously 17 disclosed certain items about the contract, price terms, et cetera." Id. at 910:17–20. Dr. Erdem 18 testified that "in order to evaluate this hypothesis whether they are indeed clear and conspicuous 19 or not, I had to look at the baseline understanding as well as whether it can be improved with some 20 targeted modifications." Id. at 910:22–25. 21 Dr. Erdem created a web-based survey to test this hypothesis, using the actual print ad as a 22 7 23 As noted above, Exhibit 244 disclosed the then-current promotional price for each package. The FTC argues that "no DIRECTV print ad in evidence stated the actual price a consumer would pay 24 during the second year of the 24-month contract because the regular price is subject to change." Dkt. No. 401 at 6. The uncontradicted evidence at trial explained why: DIRECTV could not know 25 in advance what the second-year price would be, because it had to take into account consistent increases in its own cost structure before it could set that figure. See Tr. Vol. 3 at 406:20–407:6 26 (Bentley). DIRECTV's costs to obtain content rose from 8 to 10 percent annually over a three year period, but the company could only pass on a 3 to 5 percent increase to its customers for 27 business reasons. Id. at 331:1–333:5 (Bentley). The evidence at trial also established that annual price increases occur across the pay television industry. Id. at 333:12–20. 28 17 5 1 control stimulus and creating a modified test version of the ad as a comparative stimulus. Id. at 2 927:1–929:25, 936:23–937:22.8 In the modified version, Dr. Erdem changed the manner in which 3 certain information regarding price and subscription terms appeared in the ad. Id. at 936:23– 4 937:22. She then asked survey respondents a number of questions, and concluded that her 5 modified version of the ad increased "consumer understanding" of these terms. Id. at 912:8– 6 913:20. 7 Dr. Erdem's survey suffers from several flaws that essentially negate its probative value as 8 to the central issue in the case: whether DIRECTV advertisements' are likely to mislead 9 reasonable consumers based on the net impression they create. 10 First, Dr. Erdem's survey was designed to test consumers' comprehension of certain terms 11 and to determine whether their understanding could be enhanced with minor modifications to 12 DIRECTV's disclosures. See id. at 926:20–927:24, 963:2–24, 1010:1–15. But saying that Northern District of California United States District Court 13 changing the presentation of certain information may result in better recollection of that 14 information simply does not support the conclusion that the information was likely to mislead as 15 presented in its original form. Dr. Erdem acknowledged that she did not perform a deception 16 study, or do any copy testing to determine "what is the overall impression of the consumer from 17 the ad." See id. at 965:2–19, 990:1–22. While the FTC suggests that deception studies are only 18 relevant in Lanham Act cases, see id. at 1013:14–14:10, it never explains why. Dr. Erdem's 19 generic testimony about principles of "shrouding" also failed to establish that the ad was likely to 20 mislead a reasonable consumer. See id. at 924:10–13 ("[F]ine print is a classic example of that 21 8 22 FTC counsel used a slide deck as a demonstrative visual aid during Dr. Erdem's testimony. See, e.g., Tr. Vol. 6 at 938:22–24 (referring to "slide 47"). The FTC submitted the slide deck as 23 "Attachment A" to its proposed findings of fact, and Dr. Erdem's control and test versions of the Exhibit 244 print advertisement as "Attachment B" to that filing. Dkt. Nos. 401-2 and 401-3. 24 However, the FTC never actually introduced the control and test stimuli in evidence, which it appears to recognize by attempting to submit this material as an "attachment" to its proposed 25 findings. See Dkt. No. 401-2 at 44 (depicting "Modifications to Original DIRECTV Print Ad" by showing original and treatment versions side by side, and referencing "Trial Exhibit 913"). 26 Exhibit 913 was not offered for admission at trial, nor was it included in any of the parties' stipulations. Accordingly, the Court bases its analysis on the testimony Dr. Erdem offered at trial 27 and the Court's recollection of the demonstrative slides, and rejects the FTC's apparent (and unacknowledged) attempt to supplement the record after the close of the evidence. 28 18 5 1 kind of shrouding or a practice that makes cognitive load of consumers high, and a majority of 2 consumers don't read the fine print."). Accordingly, the questions asked and answered by Dr. 3 Erdem's survey provide little, if any, support for the FTC's claim that the tested advertisement 4 (Ex. 244) would be likely to mislead a reasonable consumer. Compare Commerce Planet, 878 F. 5 Supp. 2d at 1068 (expert's testimony focused on whether most customers would understand 6 negative option and continuity program subscription aspects of website; analysis "focuse[d] on 7 what the user can perceive and what the user should do"). 8 Moreover, Dr. Erdem acknowledged that as part of her survey results, she grouped 9 participants who responded that they didn't "know" about specific terms together with those who 10 responded that they didn't "recall" or "remember" the terms. Tr. Vol. 6 at 966:6–968:11. The 11 FTC argues that this approach "is the right one for a deceptive-omission case, where consumers 12 are misled in part as a result of their lack of knowledge." Dkt. No. 401 at 14 (emphasis in Northern District of California United States District Court 13 original). But the FTC's argument fails to recognize that Dr. Erdem's survey did not test, and 14 therefore cannot establish, why a consumer responded that he or she did not know or recall the 15 terms. Because Dr. Erdem did not ask follow-up questions or have subcategories for responses, 16 she acknowledged that she did not know whether these survey participants saw the key terms but 17 just could not recall what they had seen in the print advertisement earlier, or whether they had the 18 wrong impression about these terms based on the advertisements. Tr. Vol. 6 at 966:22–968:19, 19 970:2–16 (Erdem). Accordingly, this characteristic of Dr. Erdem's survey further supports the 20 Court's conclusion that the results lack substantial probative value on the question of whether 21 Exhibit 244 was likely to mislead a reasonable consumer. 22 Finally, the Court finds that given the centrality of the disclosures in the text at the bottom 23 of the ad to the issues in this case, Dr. Erdem's web study did not adequately replicate the 24 experience of a prospective customer viewing the ad in paper form, as it was actually distributed. 25 Defendant introduced a hard copy of the advertisement in its original paper form. Ex. 2026A. 26 With the paper version in hand, a prospective customer could review those disclosures simply by 27 looking at the bottom of the page. By contrast, in Dr. Erdem's survey, the survey taker had to 28 19 5 1 scroll down the screen to see those disclosures (in other words, the disclosures were not visible 2 unless the survey taker scrolled down). Tr. Vol. 6 at 973:7–15 (Erdem). Further, the Court's 3 personal comparison at trial of the paper and electronic survey versions of the ad reflected that the 4 disclosure text in the survey version was substantially less clear and legible than the paper copy. 5 Compare Ex. 2026A with Ex. 2371 (URL link for survey).9 The fact that respondents could 6 "zoom [in] on any part," Tr. Vol. 6 at 975:2–4 (Erdem), does not remedy this problem, nor does it 7 matter that none of the participants in Dr. Erdem's pretest said that the text was not legible, Tr. 8 Vol. 6 at 1013:1–3. Survey participants were not asked if they could read the disclosure block, Tr. 9 Vol. 6 at 1015:6–11, and the Court finds that it would not be apparent to a participant that the 10 specific content of the disclosure block could be relevant (in fact, crucial) to the survey exercise. 11 For all of these reasons, the Court finds that Dr. Erdem's survey has little to no probative 12 value on the question of whether the 2013 advertisement admitted as Exhibit 244 was likely to Northern District of California United States District Court 13 mislead a consumer acting reasonably under the circumstances. 14 2. Dr. Pratkanis 15 The Court finds Dr. Pratkanis' "social influence analysis" to be similarly unpersuasive on 16 the question of whether a reasonable consumer likely would be misled by Defendant's 17 advertisements. Dr. Pratkanis acknowledged that he did no copy testing or other empirical testing 18 of how consumers perceive Defendant's advertisements. Tr. Vol. 8 at 1291:13–1293:12. He also 19 did not analyze Defendant's advertisements to determine their net impression, or do any content 20 analysis that would support the conclusion that his conclusions were generalizable to all of the ads 21 at issue over the eight-year period at issue. Tr. Vol. 8 at 1304:6–1306:24, 1311:6–1315:24. 22 Instead, Dr. Pratkanis testified that general principles of social influence such as "throwing 23 the lowball" and "free gift" can increase the "likelihood" that someone would make a purchase. 24 Dr. Pratkanis' "fundamental finding" was that DIRECTV employed a "lowball" initial offer to get 25 26 9 Because all that was entered in evidence was the URL links, rather than an actual electronic copy 27 of the survey, it is unclear how the FTC intends to preserve the evidence presented at trial for the record. 28 20 5 1 consumers to later agree to a less desirable offer. Id. at 1249:12–1250:8. During his testimony, 2 Dr. Pratkanis discussed two print ads: another variant of the one-page "Best Offer Ever" flyer 3 from 2013 (Ex. 238) and a 2014 two-page advertisement headed "Get 2 YEARS OF SAVINGS" 4 (Ex. 273). 5 Exhibit 238 describes the Entertainment Package, and has a slashed-out price of $29.99 6 "FOR 12 MONTHS," next to the notation "With 24-mo. agreement.**" A red bubble just above 7 the slashed-out price contains large white text reading "ACT NOW! $24.99/MO." with a caret. As 8 with Exhibit 244, the print ad Dr. Erdem tested, the caret and asterisks refer the viewer to a 9 disclosure section of black text against a white background that occupies approximately the 10 bottom one-third of the advertisement. The text in the disclosure section that corresponds to the 11 caret begins in the first line of that section, and contains the following bolding, underlining and 12 capitalization: "BILL CREDIT/PROGRAMMING OFFER: IF BY THE END OF Northern District of California United States District Court 13 PROMOTIONAL PRICE PERIOD(S) CUSTOMER DOES NOT CONTACT DIRECTV 14 TO CHANGE SERVICE THEN ALL SERVICES WILL AUTOMATICALLY CONTINUE 15 AT THE THEN-PREVAILING RATES." Later text in that line lists the "current prices" of 16 each package. The text in the disclosure section that corresponds to the asterisks is bolded and in 17 all capital letters, and says "24-MONTH AGREEMENT: EARLY CANCELLATION WILL 18 RESULT IN A FEE OF $20/MONTH FOR EACH REMAINING MONTH." 19 Exhibit 273, like Exhibit 244, describes three different packages. It contains the same red 20 bubble price format, slashed-out prices, and references to the disclosure section via carets and 21 asterisks. The formatting and content of the disclosures described above with regard to Exhibit 22 244 (including underlining, bolding and capitalization) is identical in Exhibit 273, except that the 23 Court assumes that this was a two-sided advertisement, such that the second page of text and the 24 disclosure section would appear on the back of the piece.10 25 For the reasons described in its analysis of Exhibit 244 above, the Court finds based on a 26 10 27 No witness testified to how this exhibit, which was admitted by stipulation, see Dkt. No. 391, was laid out in the document potential DIRECTV customers actually saw. 28 21 5 1 facial review that neither Exhibit 238 nor Exhibit 273 creates a net impression that would be likely 2 to mislead a reasonable consumer. 3 Moreover, the Court accords minimal weight to Dr. Pratkanis' opinions in light of this 4 determination that the advertisements at issue were not facially deceptive. Dr. Pratkanis' general 5 testimony regarding the "likelihood of the impressions that [a consumer] would take," Tr. Vol. 8 6 at 1347:1–1348:8, is unpersuasive for several reasons. First, while Dr. Pratkanis testified, for 7 example, that a consumer would "have a tendency and a likelihood" to perceive the slashed-out 8 price as the "real price or the regular price," he acknowledged doing no study or survey to 9 determine whether customers in fact held that belief. Id. at 1314:8–1315:15. Nor did Dr. 10 Pratkanis do any empirical testing of any advertisement, id. at 1301:4–8, survey customers to ask 11 what messages they took away from ads, id. at 1303:16–1304:24, or do a content analysis of any 12 advertisements, id. at 1305:11–1306:4. Dr. Pratkanis has performed this kind of testing in other Northern District of California United States District Court 13 instances, but did not do so here because the FTC did not ask him to. Id. at 1303:16–1304:24. 14 Similarly, Dr. Pratkanis conceded that he could not measure how purported "lowball" 15 representations would affect consumers' understanding: 16 Q: But likelihood is what? Is it more likely than not? Does that mean more than 50 percent who sign up don't know? 17 A: It would be an increase over base rate. Q: Do you know what the base rate is? 18 A: No. 19 Tr. Vol. 8 at 1312:2–6. 20 Third, as Defendant points out, the "lowball theory" also does not map well to FTC's 21 theory in this case, because that theory presumes that consumers do understand the price that they 22 will pay before the time of purchase. Id. at 1320:12–1321:12; see id. at 1348:4–8 (agreeing with 23 the Court's statement that "if they are exposed to the true terms up front there can't be a lowball 24 by definition"). Here, the Court has found that the true terms are adequately and accurately 25 disclosed in the advertisements Dr. Pratkanis testified about. For at least these reasons, the Court 26 finds that nothing in Dr. Pratkanis' testimony persuasively supports the FTC's contention that the 27 print advertisements at issue in this case created a misleading net impression in violation of the 28 22 5 1 FTC Act. 2 3. The Cases Upon Which the FTC Relies Reinforce the Conclusion That the Agency Failed to Meet Its Burden of Proof 3 The cases the FTC relies upon in support of its argument that these advertisements (and, by 4 extension, the thousands of others at issue in this case) were misleading by omission are readily 5 distinguishable. In those cases, the courts found that advertisements created a misleading 6 impression, even when the relevant information was technically disclosed in some form, because 7 the overall presentation would lead a reasonable consumer to conclude that the terms were 8 materially different than the true terms. 9 For example, in Cyberspace, the defendants mailed millions of solicitations offering 10 internet access to individuals and small businesses. 453 F.3d at 1198. The solicitations included a 11 check, usually for $3.50, addressed to the recipient and referencing the recipient's phone number 12 Northern District of California on the "re" line. Id. The check was "attached to a form resembling an invoice designed to be United States District Court 13 detached from the check by tearing at the perforated line." Id. "The back of the check and invoice 14 contained small-print disclosures revealing that cashing or depositing the check would constitute 15 agreement to pay a monthly fee for internet access, but the front of the check and the invoice 16 contained no such disclosures." Id. Under those circumstances, the Ninth Circuit readily rejected 17 defendants' argument that "the fine print notices they placed on the reverse side of the check, 18 invoice, and marketing insert preclude liability" under the FTC Act. Id. at 1200. This was 19 because the "mailing created the deceptive impression that the $3.50 check was simply a refund or 20 rebate rather than an offer for services," and gave the "overall impression that the check resolves 21 some small, outstanding debt." Id. at 1200–1201. 22 Similarly, in AMG Services, defendants' loan note contained two sets of representations: a 23 Truth In Lending Act ("TILA") box, and additional fine print disclosures. 29 F.Supp.3d at 1366. 24 The TILA box "represent[ed] that borrowers [were] obligated to repay a fixed sum equal to the 25 principal plus a single finance charge." Id. (emphasis added). But terms in the fine print 26 "contradict[ed] the terms disclosed in the TILA box," in that the repayment amount was not fixed, 27 28 23 5 1 the borrower was automatically enrolled in a ten pay-period "loan renewal" plan that imposed 2 multiple finance charges, and the "renewal plan" increased the borrower's cost of credit from $90 3 to $675. Id. at 1367. Given these contradictions, the court had no trouble concluding that "[i]t 4 requires no citation of authority to demonstrate that the 'net impression' of a boldfaced 5 representation, which states that the borrower is responsible to repay a fixed sum, is misleading 6 when the fine print indicates that the boldfaced sum is not fixed." Id. at 1368; see also id. at 1351 7 (finding that "the net impression of the Loan Note Disclosure is likely to mislead borrowers acting 8 reasonably under the circumstances because the large prominent print in the TILA Box implies 9 that borrowers will incur one finance charge while the fine print creates a process under which 10 multiple finance charges will be automatically incurred unless borrowers take affirmative action"); 11 1373 (recommending that summary judgment be granted because "the representation [was] clearly 12 misleading and the defendant relie[d] exclusively on the fine print to correct the Northern District of California United States District Court 13 misrepresentation"). 14 Johnson also illustrates the usual type of representation found to create a misleading net 15 impression. In that case, the FTC argued that a consumer reviewing defendant's website "would 16 have the net impression that they were getting a free CD for a minimal shipping and handling or 17 download fee," but failed to inform the consumer "that they would be enrolled in a core 18 membership with a negative option and additional upsell memberships with negative options." 96 19 F.Supp.3d at 1139–1140. Essentially, the FTC argued that "the disclosures were presented in a 20 way that made consumers unlikely to expect them (because the claims 'free' and 'risk free' were 21 so prominent) and were unlikely to see them (because the disclosures were inadequate or non- 22 existent)." Id. at 1140. The court agreed, finding it clear that the sites did not advertise a 23 membership program, and concluding that customers would be likely "to believe that they are 24 ordering a CD for a one-time fee and would not have reason to believe they were enrolling in 25 membership programs with monthly charges." Id. at 1145. The Court noted that "[i]nformation 26 about the trial memberships, and their associated costs, [was] hidden in disclosures that consumers 27 would be justified in believing they did not have to examine closely." Id. at 1146. 28 24 5 1 Significantly, however, the Johnson court noted that the circumstances before it were 2 different than a squarely-presented offer regarding subscription services like the one at issue in 3 this case: 4 [T]here is a significant distinction between sites that clearly offer 5 subscription-based or membership-based services to a consumer with a trial period and negative option, on one hand, and sites that 6 offer to ship a product for a one-time payment with a bundled membership trial period and negative option on the other. In the 7 first situation, the consumer is made aware of an ongoing obligation for use of the seller's service. Generally speaking, Defendants are 8 correct that a consumer's failure to cancel their membership in that situation would be the fault of the consumer. The Court's concern 9 in this case is that the sites do not clearly offer a membership program. . . . That is a significantly different situation from one in 10 which a consumer willingly receives the benefit of a subscription or membership program and forgets to cancel. 11 Id. at 1141-42. 12 Northern District of California Cyberspace and AMG Services also do not support the FTC's argument as to the probative United States District Court 13 weight of Dr. Erdem's analysis, Dkt. No. 401 at 14, and the critically different facts of those cases 14 highlight the weakness of the government's evidence in this case. The Ninth Circuit's comment in 15 Cyberspace that the defendant's proffered survey "did not probe whether the notices were 16 sufficiently conspicuous to draw the survey subjects' attention in the first place," 453 F.3d at 17 1201, addressed a circumstance in which (1) the supposed disclosure was on the back of what 18 appeared to be a check; and (2) the misleading impression was that a reasonable consumer would 19 not understand that by cashing the check she would be enrolling in a subscription service. Id. at 20 1200–1201. Similarly, as explained above, the deception in AMG Services was that the net 21 impression made by the advertisements at issue fundamentally and comprehensively contradicted 22 the true terms of the offered loan. In contrast, here Dr. Erdem's survey treated respondents who 23 did not recall the terms being studied the same as those respondents who said they did not know 24 what the terms were, without providing any persuasive basis for concluding that either answer 25 reflected a misleading net impression. See Tr. Vol. 6 at 966:22–970:16. 26 In sum, the facts in this case are plainly different from those in cases in which the FTC has 27 28 25 5 1 successfully established that a misleading net impression existed. Nothing in one part of Exhibit 2 244 (or Exhibits 238 or 273) contradicts any other part of the advertisement. The advertisement 3 makes plain that it essentially describes an "offer for services," Cyberspace at 1200, accurately 4 describes (within the limitations of a one- or two-page flyer) the terms of the offer, and invites an 5 interested consumer to follow up to select a specific package and subscribe through a phone- or 6 website-based process. The Court finds it self-evident that this case bears no resemblance to cases 7 in which a defendant represents in an advertisement that the consumer will receive a benefit (like 8 the check in Cyberspace, the fixed interest rate in AMG Services, or the free CD in Johnson), only 9 to turn around and impose a concealed and unwanted burden (like the hidden subscriptions and 10 requirements in each of those cases). Compare Ex. 1119 (FTC ".com Disclosures" guidance) at 11 18 ("Consumers who are trying to complete a task and obtain a specific product or service may not 12 pay adequate attention to a disclosure that does not relate to the task at hand. This can be Northern District of California United States District Court 13 problematic if, for example, an advertiser is selling a product or service together with a negative 14 option trial for a different product or service.") (emphasis added). 15 Nor has the FTC shown that this advertisement was a "deceptive door opener," in which 16 DIRECTV "induce[d] first contact through deception, even if the buyer later becomes fully 17 informed before entering the contract." Dkt. No. 401-1 at 146 (citing Resort Car Rental Sys., Inc. 18 v. F.T.C., 518 F.2d 962, 964 (9th Cir. 1975)). In Resort Car Rental, defendant used the slogan 19 "Dollar-A-Day" for its car rental business, but apparently charged more than that. Id. Under 20 those circumstances, the Ninth Circuit agreed with the FTC that "the trade name, 'dollar-a-day' by 21 its nature has a decisive connotation for which any qualifying language would result in a 22 contradiction in terms." Id. That principle is inapplicable here, because (1) the Court has found 23 that nothing in Exhibit 244 (or Exhibits 238 or 273) contradicts the true terms of DIRECTV's 24 provision of services; and (2) for a complex product like subscription satellite television services, 25 a reasonable consumer would understand the limitations of how information is presented in a one- 26 or two-page flyer. See Tr. Vol. 2 at 243:2–23, 311:11–317:11 (testimony of Brad Bentley that the 27 amount of "real estate" in a particular form of advertisement affects the layout of information 28 26 5 1 included, but that "we made sure all the offer details were available on each piece"). 2 Consistent with these observations, the Court finds that the FTC has failed to establish that 3 Exhibit 244, or Exhibits 238 or 273, would be likely to mislead a reasonable consumer. 4 b. Even had the FTC shown that these examples created a misleading net impression, it failed to show how that impression is generalizable to hundreds 5 or thousands of additional advertisements. 6 1. Expert Testimony 7 Dr. Erdem and Dr. Pratkanis' testified about the detailed content of only three 8 advertisements. Dr. Erdem's survey concerned a single 2013 print advertisement, which she chose 9 as the basis for her test after reviewing a set of nearly 120 DIRECTV ads culled by her staff. Tr. 10 Vol. 6 at 989:12–25. She did not conduct any copy or content analysis, id. at 989:18–22, but 11 rather reviewed this small set of advertisements "across different denominations like the content, 12 the language, the fine print, whether it was a circular, etc." and "decided [the 2013 advertisement] Northern District of California United States District Court 13 was sufficiently representative." Id. at 956:3–957:10. According to Dr. Erdem, the 14 advertisements "typically" had fine print, a promotional price, and package choices. Id. at 996:6– 15 13. She further testified that "the single page format was easier to adopt for an online survey." Id. 16 at 957:6–7. In this way, the advertisement was chosen, at least in part, because of the limitations 17 of the survey methodology and not because the advertisement itself was the most representative. 18 Even if the Court had found, contrary to the analysis above, that Exhibit 244 was likely to 19 mislead a consumer acting reasonably under the circumstances, Dr. Erdem presented no 20 persuasive explanation in support of her assertion that this single 2013 print advertisement was 21 somehow representative of even the advertisements she reviewed, let alone the universe of 40,000 22 or more advertisements at issue in this case.11 23 The Court finds that the several years' worth of advertisements accused by the FTC in this 24 case took different forms and depended on the specific offer in place at the time. See Ex. 617 25 (timeline of offers). The ads contain everyday low pricing or promote introductory pricing for 12 26 11 27 The evidence at trial established that this single 2013 advertisement would have resulted in only about 300 subscriptions, and even fewer activations. Tr. Vol. 2 at 323:13–16 (Bentley). 28 27 5 1 months; require no commitment or state that a 24-month agreement is required; have no premium 2 channel offer or offer premium channels free for 3 months; explain rebate credits (when rebates 3 were in effect) or, later, the instant rebate prices; and disclose additional terms about equipment 4 pricing, packages, and cancellation in other areas and the disclosure block. See, e.g., Exs. 78, 84, 5 85 (2007); Exs. 92, 95 (2008); Ex. 113 (2009); Exs. 124, 139 (2010); Exs. 172, 2014 (2011); Exs. 6 175, 178/2707, 185 (2012); Exs. 203, 209, 212, 217 (2013); Ex. 281 (2014).12 Indisputably, there 7 is substantial variation in DIRECTV's print advertising with respect to, among other things, 8 tactics, page count, content, and the frequency, placement, and manner of disclosures in the ads 9 Dr. Erdem reviewed. Compare, e.g., Ex. 2026A (paper copy of single page ad used in Erdem 10 study, described above), with Ex. 2707 (original four-page Sunday Circular of Ex. 178 with no 11 slashed out pricing; disclosing "FOR 12 MONTHS" and "2-year" agreement immediately under 12 each price point; and disclosing first and second-year savings in large font on inside left page and Northern District of California United States District Court 13 again, in red, on back cover), with Ex. 124 (advertising "50% OFF FOR ONE YEAR!" at top and 14 side; disclosing promotional price is "FOR 12 MONTHS" in red bubble), with Ex. 92 (two-sided 15 advertisement disclosing promotional price is "FOR 12 MONTHS" with the "Reg." package price 16 in red bubbles on front and back), with Ex. 172 (four-page direct mail letter disclosing "FOR 12 17 MONTHS" and "2-yr agreement" immediately under each price point; disclosing first and second- 18 year savings in second paragraph of first page, and under each promotional price on later pages; 19 disclosing value of premium channels next to premium channel copy), with Ex. 209 (four-page ad 20 disclosing promotional price is "FOR 12 MONTHS" 12 times on front and back covers; disclosing 21 first and second-year savings under each package on back cover; disclosing value of free premium 22 channels next to premium channel copy on inside right page), with Ex. 113/2011A (slashing out 23 regular package price rather than prior promotional price), with Ex. 139/2013A (advertising 24 12 25 As one example of the FTC's tendency to gloss over pertinent details, counsel in opening statement highlighted an advertisement from 2007 and asked "Where does it say what the higher 26 second year price will be?" Tr. Vol. 1 at 11:12–13 (referencing and displaying Ex. 84). But the evidence ultimately established, without contradiction, that this offer did not require a two-year 27 commitment, and that it contained standard rather than promotional pricing. Tr. Vol. 2 at 345:3– 25 (comparing Ex. 84 to Ex. 617 (timeline of offers)) (Bentley). 28 28 5 1 "PACKAGES START AT $29.99 Everyday Low Price" without promoting any specific package 2 or premium channels). This variation among DIRECTV's ads precludes generalizing Dr. Erdem's 3 survey results to the relative handful of print ads she reviewed, let alone to the some 40,000 she 4 did not.13 5 Contrary to the FTC's claim, neither DIRECTV nor the Court is suggesting that the FTC 6 had to introduce all of the more than 40,000 advertisements into evidence. See Dkt. No. 401 at 15. 7 But the FTC does have to explain why conclusions about a handful of advertisements can be 8 applied to derive a uniform net impression for an extremely large number of others that vary 9 significantly in format, content and emphasis. The FTC simply cannot meet this burden by 10 characterizing the advertisements at a high level of generality and asserting that conclusions 11 regarding one advertisement apply uniformly to tens of thousands of others. See Dkt. No. 401-1 12 (FTC's proposed findings of fact and conclusions of law) at ¶¶ 82-91 (referencing inclusion of Northern District of California United States District Court 13 "some type of 'call to action'" in advertisements, as well as general thematic and formatting 14 consistency); Tr. Vol. 6 at 956:16–20 (Erdem) ("[T]he common denominations were some kind of 15 promotional package -- I mean, premium channels package and promotional price, some kind of 16 call to action, we call these kind of ads call to action ads, and the templates were similar."). The 17 Court does not find Dr. Erdem's opinion that she would "expect no differences" in the empirical 18 test results across all of the dramatically different types of advertisements at issue in this case, Tr. 19 Vol. 6 at 996:21-997:10, to be persuasive or adequately supported. The Court similarly finds 20 unpersuasive and inadequately supported Dr. Pratkanis' assertion that his conclusions about 21 22 13 This conclusion applies with even more force to the television commercials. Those 23 commercials plainly differ in content, purpose, approach and format from print ads, and often relied primarily on overt humor to attract viewers' attention and interest them enough to call a toll- 24 free number or visit the website to learn more. See Tr. Vol. 5 at 872:7–875:7 (testimony of former DIRECTV executive Jon Gieselman explaining that commercials focused on different main 25 messages, with the main "proof points" being "technology superiority, programming superiority [and] customer service superiority"); Ex. 443 (television commercial featuring a talking horse 26 discussing benefits of DIRECTV); Tr. Vol. 2 at 260:13–18, 266:4–16 (the purpose of DIRECTV's advertisements was to "create some interest and start a dialogue by motivating the consumers to 27 pick up the phone and call DIRECTV or to go on the DIRECTV website," which is colloquially known as the "call to action" in DIRECTV's advertising) (Bentley). 28 29 5 1 Exhibits 238 and 273 can be generalized to a broader universe of Defendant's advertisements. 2 2. Other Extrinsic Evidence 3 The Court also finds that none of the other extrinsic evidence relied upon by the FTC 4 meets its burden of establishing that any of Defendant's advertisements was likely to mislead a 5 reasonable consumer. 6 a. Eye Tracking Studies 7 The FTC introduced evidence that in 2010, DIRECTV contracted with Realeyes North 8 America, a market research firm, to perform eye tracking studies on four DIRECTV advertising 9 circulars. Hellstrom Depo. 9:19–23; Ex. 705 at 1–2.14 DIRECTV engaged Realeyes to help its 10 team understand how consumers read circulars. Ex. 378 at 31; Ex. 408. During this study, 11 participants were shown a DIRECTV print advertisement on a large computer screen. Tr. Vol. 7 12 at 1221:8 (Hellstrom Depo. at 25:25–26:07); Ex. 718 at 16. Realeyes then tracked participants' Northern District of California United States District Court 13 "eye gaze data" for 30 to 90 seconds, aggregated that data, and created a "heat map" to show 14 where participants focused their "visual attention" on DIRECTV's advertisement. Tr. Vol. 7 at 15 1221:8 (Hellstrom Depo. at 62:16-63:12); Ex. 718 at 16; Ex. 720 at 3. 16 The four DIRECTV ads tested by Realeyes were each four-page circular ads. Each of the 17 circular ads was comprised of three components—a front cover, a two-page inside spread, and a 18 back cover. Hellstrom Depo. 51:12–16. Realeyes conducted individual tests of all three 19 components of each circular ad (twelve tests in all), and generated a report for each test. 20 Hellstrom Depo. 50:24–51:16; Exs. 708, 709, 710, 713, 714, 715, 718, 719, 720, 723, 724, 725. 21 Each report contained a bar graph called "Attention Allocation," as well as heat maps. See, e.g., 22 Ex. 708. DIRECTV prepared an internal document called "Heat Mapping Research: Learnings & 23 Recommendations" which presented the results of the Realeyes eye tracking studies. Ex. 408. 24 These eye tracking studies are not persuasive on the relevant issue: whether the studied 25 advertisements, or any of Defendant's advertisements, were likely to leave a reasonable consumer 26 14 27 The FTC played excerpts of Mr. Hellstrom's deposition at trial. Dkt. No. 387 ("Tr. Vol. 7") at 1221:5–8. 28 30 5 1 with a misleading net impression. These types of studies do not measure consumer sentiment, 2 persuasion, or understanding of an advertisement. Tr. Vol. 8 at 1339:24–1340:5 (Leever); see also 3 id. at 1339:15–23 (Pratkanis) (answer to question "And you can't measure any miscomprehension 4 or any wrong impression from an ad based on an eye tracking study?" was "Generally, that's 5 correct. It might give you some input into understanding that, but in general that's correct."). The 6 fact that study participants' eyes may be more drawn to bright colors and pictures than to written 7 details during a simulation test conducted on a computer simply does not provide persuasive 8 support for the FTC's claim that advertisements describing a multifaceted offer for a plainly 9 complex product like subscription satellite television services were misleading.15 10 b. RIO – Subscriber Call Data 11 DIRECTV's RIO customer service database includes a note field where the agent writes 12 down information about a customer's telephone call. Tr. Vol. 8 at 1280:14–1281:9 (Pratkanis). Northern District of California United States District Court 13 DIRECTV developed proprietary text-mining software that goes through the note field and assigns 14 a primary tag to the call, corresponding to the primary reason for the phone call. (Id.). There are 15 about 1,120 different primary tags. Tr. Vol. 8 at 1281:10–12 (Pratkanis). 16 Dr. Pratkanis analyzed the RIO data by selecting 17 primary tags he characterized as 17 encompassing claims by customers, at the fourth and thirteenth months of service, expressing 18 surprise, confusion, or dislike over bill increases. Tr. Vol. 8 at 1281:17–1282:10 (Pratkanis). The 19 17 tags fall into four categories: billing, 24-month commitment, early cancellation fee, and 20 possible remedies for a price increase. Id. Dr. Pratkanis's RIO analysis was based on a random 21 sample drawn from the RIO database. Tr. Vol. 8 at 1285:12–23 (Pratkanis). 22 Dr. Pratkanis opined that spikes in the RIO data at months 4 and 13 "would confirm the 23 kind of things you would expect if the free gift as a negative option is not fully disclosed." Tr. 24 25 15 This conclusion applies equally to the "click-tracking studies" introduced by the FTC. See, e.g., 26 Ex. 347; Tr. Vol. 8 at 1340:6–9 (Pratkanis). A survey participant's response to an invitation to "click on the three parts of the [advertisement] that are most compelling to you," Ex. 347 at slide 27 30, simply is not probative of whether the advertisement was likely to leave a reasonable consumer with a misleading net impression. 28 31 5 1 Vol. 8 at 1282:19–1283:3. But as Dr. Pratkanis recognized, month 4 is when subscribers wishing 2 to cancel their premium packages because the free period ended would naturally call to do so. Id. 3 at 1336:5–7. And in month 13, "one possibility" is that subscribers call to explore other package 4 options once their introductory pricing ends. Id. at 1336:8–15. Critically, Dr. Pratkanis did no 5 content analysis to identify callers at these stages who called because they claimed to have been 6 misled about the conditions of their subscriptions. Id. at 1336:17–24. While Dr. Pratkanis said he 7 excluded tags like "cancel premiums" or "cancel HBO" from his analysis, id. at 1348:18–1349:6, 8 the Court finds the RIO analysis to be unpersuasive evidence of consumer deception given Dr. 9 Pratkanis' understandable inability to determine anything about why subscribers were actually 10 calling in the months characterized as "spikes." 11 c. Customer Experience Steering Committee Documents 12 The Court also heard evidence regarding findings made by DIRECTV's "Customer Northern District of California United States District Court 13 Experience Steering Committee" ("CX Committee"). The CX Committee was a cross-functional 14 group of DIRECTV executives led by Rasesh Patel. Tr. Vol. 2 at 274:16–20 (Bentley), Tr. Vol. 6 15 at 1020:4–6 (Patel). The evidence showed that the CX Committee was created in 2012 as part of a 16 "company-wide focus" to "improve the customer experience." Tr. Vol. 6 at 1025:4–7 (Patel). 17 The initiative focused on both current and prospective DIRECTV customers. Tr. Vol. 6 at 18 1027:16–18 (Patel). The CX Committee reviewed customer research and satisfaction data and 19 operational data. Tr. Vol. 6 at 1029:19–1030:2 (Patel). The CX Committee's goal was to 20 "identify customer pain points. . . and then really prioritize the ones that were most meaningful in 21 terms of impact." Tr. Vol. 6 at 1030:7–9 (Patel). The CX Initiative was "very self-critical" and 22 "discerning" in identifying "every opportunity. . . to improve the customer experience." Tr. Vol. 6 23 at 1052:18–24 (Patel). By 2013, the CX Committee had identified numerous customer "pain 24 points," which it categorized and prioritized in a "Customer Pain Points – Review and 25 Prioritization" discussion document dated May 7, 2013. Tr. Vol. 6 at 1031:8–22 (Patel); Ex. 589. 26 The CX Committee identified a number of these as "top recommended pain points" based on the 27 "degree of impact" a particular pain point might have on metrics such as customer satisfaction, 28 32 5 1 call volume, and numerous other dimensions. Tr. Vol. 6 at 1043:10–18 (Patel); Ex. 589 at 7. The 2 CX Committee initiative lasted three years, until 2015. Tr. Vol. 6 at 1029:3–7 (Patel). 3 Nothing about the CX Committee materials introduced at trial persuasively supports the 4 FTC's claim that Defendant's advertisements misled millions of customers between 2007 and 5 2015 (or were likely to do so). Instead, Mr. Patel credibly testified that the CX Committee 6 engaged in a good-faith and comprehensive effort to evaluate DIRECTV's operations, identify 7 potential problem areas and recommend improvements. The evidence at trial established that a 8 number of such improvements were implemented. For example, in October 2012, DIRECTV 9 moved from an offer requiring customers to redeem a rebate via mail or online to an "instant" 10 rebate. Ex. 617 at 5. This policy ensured that every customer would receive the discounted price 11 starting with the first month's bill, and addressed an issue noted by the CX Committee. Tr. Vol. 2 12 at 293:11–21 (Bentley). This change resulted in an $80 million reduction in revenue each year. Northern District of California United States District Court 13 Id. at 294:3–11 (Bentley). As another example, around the same time, DIRECTV adopted a 14 "Simplified Bill" that incorporated a number of changes with the goal of making the bill easier to 15 understand. Ex. 569. And third, another pain point that the CX Committee identified was that 16 customers were unable to refuse the premium channel promotional offer at the point of sale or 17 request a future cancellation date for premium channels. Ex. 589 at 4–9. In response to this pain 18 point, DIRECTV implemented changes to its premium offer policies and processes so that 19 customers could decline the premium channel offer at the point of sale, or at any specified future 20 date without having to remember to call to cancel on the last day of the trial period. Tr. Vol. 7 at 21 1053:1–11 (Patel). 22 Mr. Patel also credibly explained that if consumers believed that the terms in DIRECTV's 23 advertisements did not match the terms disclosed when they sought to subscribe, that feeling of 24 deception would have been reflected in the findings of the CX Committee and in DIRECTV's data 25 collected on pain points, consumer research, sales calls, closing rates, activation rates, and churn. 26 Tr. Vol. 8 at 1047:3–10451:9 (Patel). The evidence did not support the FTC's theory in this 27 regard. DIRECTV's "churn rate" (the rate at which customers leave DIRECTV's platform) stayed 28 33 5 1 constant at approximately 1.5% from 2007 to 2013. Tr. Vol. 4 at 650:20–651:7 (Guyardo); Ex. 2 991 at 5. DIRECTV improved its market share from 2008 to 2013, when "the DIRECTV 3 subscriber base grew by 3.5 million subscribers. . . while all of [DIRECTV's] other competitors 4 combined only grew by 400,000." Tr. Vol. 4 at 651:22–652:5 (Guyardo); Ex. 991 at 6. A 5 subscriber's average time on the platform has also steadily increased over time. The average 6 customer now stays with DIRECTV for "close[] to seven years"—years after the initial 24-month 7 commitment period. Tr. Vol. 2 at 347:8–18 (Bentley).16 Finally, DIRECTV's Net Promoter Score 8 increased over time from 16 to 34 and from "second best in the industry to best in class in the 9 industry." Tr. Vol. 7 at 1077:16–20 (Patel).17 10 DIRECTV's investment of substantial resources in analyzing its operations, candidly 11 identifying areas for improvement, and following through on a number of improvements does not 12 support a finding that the company violated the FTC Act. If anything, this evidence further Northern District of California United States District Court 13 underscores how this case is very unlike the straightforward deception cases upon which the FTC 14 purports to rely. 15 d. Other Evidence 16 Finally, none of the FTC's other evidence persuasively met its burden of proof of showing 17 that DIRECTV's print advertisements were likely to deceive a reasonable consumer. The FTC's 18 attempt to rely almost entirely on pretrial deposition testimony regarding a large number of 19 exhibits never even discussed at trial is telling, and does not account for the detailed and credible 20 trial testimony that substantially undercut its position. See Dkt. No. 401-1 at ¶¶ 411–453 (citing 21 16 22 DIRECTV witnesses testified that, due to upfront "subscriber acquisition costs" approaching $1,000 per subscriber, the company does not break even until, on average, a customer remains on 23 the platform for 28 to 30 months. Tr. Vol. 2 at 330:3–6 (Bentley); Tr. Vol. 8 at 881:25-882:2 (Gieselman). This means that it is "absolutely paramount" to DIRECTV's business model to 24 retain the average customer for longer than the break-even point. Tr. Vol. 2 at 347:24–348:14 (Bentley). 17 25 "Net Promoter Score," or "NPS," is a calculation of promoters (consumers who are satisfied or would recommend a service) minus detractors (consumers who are dissatisfied or would not 26 recommend a service), the sum of which measures customer satisfaction. Tr. Vol. 3 at 363:12– 364:11 (Bentley); Ex. 571 at 2. NPS is not a metric unique to DIRECTV; rather, it is a "self- 27 critical" standardized metric that can be compared within and across industries. Tr. Vol. 3 at 364:6–11 (Bentley). 28 34 5 1 virtually no trial testimony). Moreover, the FTC's approach appears to have been to seize on 2 language superficially helpful to its case in documents, while consistently ignoring pivotal context 3 showing the actual meaning of the evidence. 4 For example, the FTC called DIRECTV executive Brad Bentley in its case-in-chief, and 5 asked him a number of questions about a study prepared by McKinsey consultants as part of the 6 CX Committee project. Tr. Vol. 2 at 279:17–282:5 (Bentley); Ex. 573 (July 10, 2012 "discussion 7 document" slide deck). FTC counsel focused on a slide called "NPS drops early--opportunity to 8 improve selling and on-boarding." Tr. Vol. 2 at 279:17–282:5 (Bentley); Ex. 573 at 18. 9 Government counsel suggested that the slide showed that "33 percent of DIRECTV customers felt 10 misled during the sign-up process." Tr. Vol. 2 at 281:4–9. But on redirect, Mr. Bentley made 11 clear that because this exercise surveyed customers in the first 90 days of their subscriptions, this 12 figure had nothing to do with the issues that are the basis of the FTC's complaint: second-year Northern District of California United States District Court 13 pricing, 24-month commitment or early cancellation fees. Id. at 288:15–289:14. Moreover, Mr. 14 Bentley persuasively explained that the biggest complaint among onboarding customers at this 15 time was the first month's bill not aligning with the customer's expectations, an issue DIRECTV 16 addressed by replacing its mail-in rebate model with an automatic instant rebate model. Id. at 17 291:13–294:14 (customers reported feeling misled in survey because, although they "had to 18 redeem [their] rebate" in order to obtain the introductory price, "oftentimes consumers wouldn't 19 redeem [the rebate] quick[ly] enough for it to be reflected on the first month bill."). 20 iv. Conclusion 21 The evidence at trial conclusively established that the FTC failed in its case-in-chief to 22 meet its burden of proving a Section 5 claim based on any of DIRECTV's non-website 23 advertisements. The FTC's ambition in attempting to show that over 40,000 advertisements were 24 likely to deceive substantially exceeded the strength of its evidence: this case did not involve the 25 type of strong proof the Court would expect to see in a case seeking nearly $4 billion in restitution, 26 based on a claim that all of DIRECTV's 33 million customers between 2007 and 2015 were 27 necessarily deceived. Tr. Vol. 1 at 34:3–16 (FTC's opening statement). Because the FTC 28 35 5 1 definitively failed to prove its case as to Defendant's non-website advertisements, the motion for 2 judgment on partial findings is GRANTED as to the FTC's Section 5 claims to the extent they are 3 based on those advertisements. 4 D. The Court Will Defer Judgment With Regard to the Claims Based on the Website 5 For now, the Court reaches a different conclusion with regard to the Section 5 and ROSCA 6 claims based on the various iterations of Defendant's website. While the FTC's case as to the 7 websites was far from overwhelming, the Court finds it appropriate to defer a determination on 8 these claims until after the close of the evidence. 9 i. Website version tested by Dr. Erdem (Exs. 1052, 2033 and 2371) 10 Dr. Erdem conducted an online survey based on a modified mockup version of the August 11 2013 version of DIRECTV's website (known as the "Old Flow"). Ex. 2371; Tr. Vol. 6 at 929:1– 12 11, 937:23–938:8; Tr. Vol. 4 at 683:17–684:4; Tr. Vol. 9 at 1448:1–20 (Poling-Hiraldo and Northern District of California United States District Court 13 Leever). Dr. Erdem testified that she used the August 2013 Old Flow version of the website 14 because she understood that this was the only interactive live version the FTC was able to recover. 15 Tr. Vol. 6 at 938:2–5. FTC expert witness Dr. Nathan Good prepared the "UX" (or user 16 experience) version of the August 2013 Old Flow that Dr. Erdem used as the basis for the 17 treatment stimulus. Tr. Vol. 1 at 114:8–20, 121:8–25 (Good). A number of the links that were 18 live on the actual website were not live in Dr. Good's UX version (meaning that nothing would 19 happen when those links were clicked). Id. at 114:16–20, 121:8–126:14; see also generally Ex. 20 1052. 21 Dr. Erdem claimed that the results of her survey could be generalized to draw conclusions 22 about other versions of the website beyond the one she tested. Tr. Vol. 6 at 956:3–957:15, 986:1– 23 995:25. She testified that she would "expect" the results to be the same for other versions of the 24 website, but she did not test any other versions. Tr. Vol. 6 at 1003:16–23. 25 In the Old Flow, to complete a purchase, consumers would navigate through a landing 26 page, programming package selection page, receiver selection page, a shopping cart page, and 27 several checkout pages. See Ex. 1045; see also Ex. 2033 at .pdf p. 2 (Landing Page), 5–8 28 36 5 1 (Programming Package Selection Page), 28 (Receiver Selection Page), 47–49 (Shopping Cart 2 Page), 107 (Checkout Page: Installation Address), 111 (Checkout Page: Account Info), 112 3 (Checkout Page: Billing Address), and 121–123 (Confirmation Page). 4 Based on a facial review of the Old Flow, the Court cannot say conclusively at this stage 5 that the FTC has failed to prove that this version of the website could create a net impression that 6 was likely to deceive a consumer acting reasonably under the circumstances. In the Old Flow, on 7 the landing page of the website, near the introductory discounted price, DIRECTV disclosed that 8 the discounted price was for 12 months and required a 24-month agreement, see Ex. 2033 at 1; 9 Dkt. No. 370 ("Tr. Vol. 3") at 510:7–10 (Poling-Hiraldo). However, other significant details only 10 became visible if the prospective customer moused over a small info hover, labeled only "Offer 11 Details," at the bottom of the landing page under the "View All Packages" button. See Ex. 2033 at 12 2 (disclosures that all offers require a 24-month agreement, the amount of the discount for the first Northern District of California United States District Court 13 12 months, and the $20/month early cancellation fee appeared only behind info hover). 14 15 16 17 18 When a prospective customer clicked the "View All Packages" button and provided her zip 19 code, she would move to the package selection page. Id. at 3–4. On that page, near the 20 introductory discounted price, DIRECTV disclosed that the discounted price was for 12 months 21 with a 24-month agreement, and disclosed the regular package price, see Ex. 2033 at 4. 22 23 24 25 26 27 28 37 5 1 2 3 4 5 6 7 8 9 10 On the bottom of each of the package selection page, receiver selection page, and cart 11 page, DIRECTV disclosed that all offers require a 24-Month Agreement, next to a blue 12 "Additional Offer Details" hyperlink, see Ex. 2033 at 7, 48, 92; Tr. Vol. 3 at 511:17–512:11 Northern District of California United States District Court 13 (Poling-Hiraldo). 14 15 16 Again, however, significant details appeared only if a prospective customer clicked the blue 17 "Additional Offer Details" link on the package selection page, receiver selection page, and cart 18 page. Clicking this link led users to a light box (i.e., a textbox overlay that appeared on the same 19 page and dimmed the background, Tr. Vol. 1 at 84:10–14 (Stahl); Tr. Vol. 3 at 451:10–15 20 (Mandel)) containing a long list of details about the offer, including that the customer had to call 21 to change or cancel the premium channels after three months or be charged the then-prevailing 22 rate; the regular cost of the premium channels for three months; the regular package prices; the 24- 23 month agreement; and the $20/month early cancellation fee, see Ex. 2033 at 23-26; Tr. Vol. 3 at 24 512:12–513:25 (Poling-Hiraldo). While this box is titled "Additional Information," much of this 25 information is actually central to a potential customer's ability to evaluate and understand the 26 terms of the subscription. 27 28 38 5 1 2 3 4 5 6 At the top of the shopping cart page, DIRECTV disclosed the regular package price in 7 black, and the first year discount immediately underneath it, in green, as a bill credit labeled for 8 "Months 1-12," see Ex. 2033 at 50. On the cart page, immediately next to the package, the 9 website included blue language reading "What is my agreement?" Id. at 49. A prospective 10 customer who clicked on this language would see a box reading "You agree to maintain 11 DIRECTV service with any base package of $29.99/mo. or above for 24 consecutive months." 12 Northern District of California Id.; Tr. Vol. 3 at 514:23–515:12, 516:15–21 (Poling-Hiraldo). 18 United States District Court 13 14 15 16 At the top of the Cart, next to the "Package & Programming" header, DIRECTV provided 17 a "View Monthly Cost" tab that opened a chart showing all costs for each of the 24 months in six 18 month blocks, including that the regular price of the premium channels would be billed starting in 19 month four, and that the regular package price would be billed starting in month 13, see 2033 at 20 78–79; Tr. Vol. 3 at 514:1–22, 518:17–519:4 (Poling-Hiraldo). 21 In the Cart, under a "Premium Channels" header, DIRECTV disclosed the regular price of 22 the premium channels in black, and immediately underneath them, in green, a bill credit labeled: 23 "FREE for 3 months: HBO, SHOWTIME, STARZ, and Cinemax", see Ex. 2033 at 59. Again, the 24 25 18 DIRECTV referred to this mechanism as a "tool tip." A tool tip is a method DIRECTV used to 26 disclose additional information on the same webpage when the consumer clicked on a prompt or icon signaling there was additional information (e.g., a blue "i" enclosed in a circle). Tr. Vol. 3 at 27 515:24–25, 516:7–11 (Poling-Hiraldo). Unlike a hyperlink, a tool tip does not redirect a user to another webpage. Id. at 516:7–11. 28 39 5 1 disclosure that the customer must call to cancel the premium channels after three free months or be 2 charged the then-prevailing rate only appeared if the prospective customer moused over an info 3 hover next to the premium channel bill credit in the shopping cart on the cart page. See Ex. 2033 4 at 58; Tr. Vol. 3 at 517:10–20 (Poling- Hiraldo). 5 6 7 8 9 10 Finally, on the checkout page of the Old Flow, before the consumer could place an order 11 and transmit financial information to DIRECTV, the consumer was required to affirmatively click 12 the orange button affirming: "I Accept. Submit My Order." Ex. 2033 at 111. Next to the orange Northern District of California United States District Court 13 button was a blue hyperlink labeled "By clicking 'Submit Order' I agree to these Terms & 14 Conditions." Id. Clicking on this link brought up a light box that detailed the 24-month 15 agreement and the $20/month early cancellation fee. See Ex. 1021 at 124-25; Tr. Vol. 3 at 519:5– 16 15 (Poling-Hiraldo). 17 18 19 It was technically possible for a consumer to proceed through the entire purchase process 20 without ever clicking or hovering over any of the hyperlinks or symbols described above. See Ex. 21 1060 at 0:01–5:20, Ex. 1069 at 0:01–5:21; Tr. Vol. 9 at 1482:22–1483:3, 1513:11–1514:15 22 (Leever). 23 ii. Section 5 24 Based on a facial review of the Old Flow website, as well as the evidence introduced at 25 trial, the Court will defer a finding as to what net impression that website would leave with a 26 consumer acting reasonably. The manner in which some key terms were presented behind 27 hyperlinks, info hovers and tool tips leaves the Court unable to conclude at this stage that 28 40 5 1 judgment on partial findings is appropriate. The Court likewise defers this analysis as to the other 2 versions of the website placed in evidence by the FTC. 19 3 iii. ROSCA 4 ROSCA prohibits certain methods of negative option marketing on the Internet. A 5 "negative option feature" is "an offer or agreement to sell or provide any goods or services, a 6 provision under which the customer's silence or failure to take an affirmative action to reject 7 goods or services or to cancel the agreement is interpreted by the seller as acceptance of the offer." 8 16 C.F.R. § 310.2(w). ROSCA prohibits charging consumers for goods or services in Internet 9 transactions through a negative option feature, unless the seller "provides text that clearly and 10 conspicuously discloses all material terms of the transaction before obtaining the consumer's 11 billing information." 15 U.S.C. § 8403(1). Additionally, ROSCA requires sellers to obtain a 12 "consumer's express informed consent before charging the consumer's credit card, debit card, Northern District of California United States District Court 13 bank account, or other financial account for products or services through such transaction." Id. 14 § 8403(2). 15 Consistent with the above discussion of the Old Flow, the Court will defer ruling on this 16 claim until the close of the evidence. The Court is not in a position to conclude on the record 17 18 19 DIRECTV's web flow underwent two major redesigns during the relevant time period. The 19 design that existed until late 2009 was referred to as the "Wizard Flow." In late 2009, the Old Flow replaced the Wizard Flow, and the "New Flow" completely replaced the Old Flow in 2014. 20 Tr. Vol. 4 at 683:17–684:4 (Poling-Hiraldo); Tr. Vol. 9 at 1448:12–20 (Leever). The New Flow was publically split-tested with the Old Flow from January to July of 2014. During that period, 21 some consumers were directed to the Old Flow, and some were directed to the New Flow. Tr. Vol. 9 at 1493:7–1494:2 (Leever). After more than a year of development, the New Flow took 22 over 100% of prospect traffic in July 2014. Tr. Vol. 9 at 1486:21–1487:5 (Leever). The Wizard Flow, Old Flow, and New Flow differed significantly with respect to, among other things: the 23 design and layout of information (e.g., whether the packages were presented vertically or horizontally); the number and placement of hyperlinks, tool tips, or info hovers that DIRECTV 24 used to disclose information about its offers and product; and the icons and prompts signaling those hyperlinks, tool tips, or info hovers. Compare Ex. 1000 (9/4/09 Wizard Flow), with Ex. 25 2033 (8/13 Old Flow), with Ex. 1065 (9/17/15 New Flow); see also Tr. Vol. 9 at 1448:22–1451:21 (Leever). DIRECTV's web flow designs also differed with respect to the flow of webpages. For 26 example, in the New Flow, between the package selection step and the receiver selection step, the consumer was directed to a "Choose add-on packages" step that was not in the Old Flow. At the 27 top of this step, DIRECTV provided information and disclosures regarding the free premium channels offer. See Ex. 1065 at 10:58–11:14 (9/17/15 capture). 28 41 5 1 presented thus far that the FTC cannot meet its burden, given the manner in which information 2 about the negative option feature appeared on the website behind hyperlinks, info hovers or tool 3 tips. This decision to defer ruling until the close of the evidence applies to all versions of the 4 website in existence in 2011 or later that were introduced in evidence by the FTC.20 5 E. Remedy 6 Because the Court has granted Defendant's motion for partial findings to the extent the 7 FTC's case is based on anything other than Defendant's website, the scope of the maximum 8 potential recovery in this case has been substantially curtailed.21 Even in the event the FTC might 9 eventually be able to prevail on what remains of the case (a question the Court need not answer at 10 this stage), the Court believes that there are a number of problems with the FTC's equitable relief 11 theories that suggest it will have difficulty meeting its burden of providing an adequate basis for 12 the relief it seeks. Northern District of California United States District Court 13 The Ninth Circuit has held that § 13(b) of the FTC Act empowers the Court to award "any 14 ancillary relief necessary to accomplish complete justice, including restitution" and injunctive 15 relief. Commerce Planet, 815 F.3d at 598 (citing F.T.C. v. Pantron I Corp., 33 F.3d 1088, 1102 16 (9th Cir. 1994)). This holding was premised on the principle that "[t]he equitable jurisdiction to 17 enjoin future violations of § 5(a) carries with it the inherent power to deprive defendants of their 18 unjust gains from past violations. . . ." Commerce Planet, 815 F.3d at 599. The Ninth Circuit has 19 adopted a two-step approach to determining the amount of unjust gains. Under the first step, "the 20 FTC bears the burden of proving that the amount it seeks in restitution reasonably approximates 21 the defendant's unjust gains." Commerce Planet, 815 F.3d at 603. In Commerce Planet, the 22 Ninth Circuit held that unjust gains in a case like that one are "measured by the defendant's net 23 revenues (typically the amount consumers paid for the product or service minus refunds and 24 chargebacks), not by the defendant's net profits." Id. If the FTC makes this showing, at the 25 20 26 ROSCA was enacted on December 29, 2010. 21 See Tr. Vol. 9 at 1441:23–1442:15 (only approximately 20% of direct sales subscribers either 27 subscribed via DIRECTV.com or subscribed via a phone number that indicated they started the subscription process on the website) (Leever). 28 42 5 1 second step "the burden then shifts to the defendant to show that the FTC's figures overstate the 2 amount of the defendant's unjust gains." Id. at 604. "[T]he purpose of such an award is to prevent 3 the defendant's unjust enrichment by recapturing the gains the defendant secured in a transaction." 4 Id. at 603 (quotation omitted). 5 Here, the Court has significant questions about the FTC's ability to prove that the amount 6 it seeks in restitution reasonably approximates the Defendant's unjust gains, even if it prevails to 7 some extent with regard to its Section 5 claim based on the website. Expert witness Dr. Daniel 8 Rascher testified in support of the FTC's claim that $3.95 billion represents "a reasonable estimate 9 (under Commerce Planet) of the unjust gains DIRECTV collected from consumers." Dkt. 337 at 10 11. To calculate this amount, Dr. Rascher examined customer billing data for certain types of 11 programming packages identified by the FTC. Tr. Vol. 10 at 1552:5–10, 1554:11–17. He then 12 calculated an estimate of unjust gains related to second year pricing, Tr. Vol. 10 at 1599:09– Northern District of California United States District Court 13 1600:07 ($3.68 billion); Ex. 1321, and the premium channel offer, Tr. Vol. 10 at 1605:12–16 14 ($268 million); Ex. 1322.22 15 Dr. Rascher testified that he calculated DIRECTV's "unjust gains" based on the 16 assumption that "all DIRECTV subscribers. . . thought they would pay the same amount in the 17 second year as they paid in the first," Tr. Vol. 10 at 1630:15–20 (emphasis added), and that he 18 applied "that same kind of presumption" with respect to unjust gains for the premium channel 19 offer, id. at 1681:11–25. He thus presumed that all DIRECTV subscribers in his approximation of 20 unjust gains were misled, and that they were all misled in the same way. Dr. Rascher testified that 21 he was instructed that the FTC would prove what DIRECTV customers "thought" or "believed" 22 their deal with DIRECTV to be and thus supply the core predicate for his analysis. Tr. Vol. 10 at 23 1616:13–18, 1618:08–25 (Rascher). Dr. Rascher testified that the FTC advised him that he could 24 "presume reliance," which he said he then "operationalized" as a presumption that all subscribers 25 purchased DIRECTV service with the "expectation" that the second-year price would be the same 26 22 27 Dr. Rascher had never previously testified in a false advertising case before this trial. Tr. Vol. 10 at 1608:9–11. 28 43 5 1 as the first year price (and he adopted similar presumptions as to the other terms at issue). Tr. Vol. 2 10 at 1554:5–1556:6. 3 The Court acknowledges the Ninth Circuit's holding that once the FTC proves a defendant 4 made widely-disseminated material misrepresentations, it is "entitled to a presumption that all 5 consumers who purchased [defendant's product] did so in reliance on the misrepresentations." 6 Commerce Planet, 815 F.3d at 604 (citing F.T.C. v. Figgie Int'l, 994 F.2d 595, 605-06 (9th Cir. 7 1993)).23 But here, the Court believes the FTC has an uphill climb to prove its "expectation" 8 theory, because it will be difficult to establish based on the presentation of various terms in several 9 different iterations of Defendant's website that a reasonable consumer uniformly would expect to 10 pay the same price in year two that she paid in year one, as Dr. Rascher assumed. See Tr. Vol. 10 11 at 1555:08–16 (to determine the price that a customer would pay in the but-for world, Dr. Rascher 12 simply used the price the customer had paid in the first year based on the billing data). In other Northern District of California United States District Court 13 words, Dr. Rascher simply presumed that customers had an expectation that they would pay the 14 same introductory price for two years, based only on an instruction from the FTC. Id. at 1618:08– 15 25, 1667:02–09 (Rascher). The Court believes that this issue, and the breadth of the FTC's 16 interpretation of the "presumption of reliance," creates a significant possibility that at the close of 17 the evidence the FTC will be unable to establish a reasonable approximation of damages, or that 18 its approximation will be substantially or even entirely rebutted by the Defendant. In any event, 19 the Court will confront this concern if the FTC is able to prove liability as to what remains of its 20 case.24 21 22 23 23 Defendants cite the Second Circuit's decision in F.T.C. v. Verity Int'l, 443 F.3d 48, 69 (2d Cir. 24 2006), in support of its argument that no presumption of reliance applies here. Dkt. No. 396 at 23. The Ninth Circuit has never adopted the reasoning of Verity, and in fact distinguished that case in 25 an unpublished memorandum disposition. See F.T.C. v. Publishers Bus. Svcs., 540 Fed. Appx. 555, 556–557 (9th Cir. 2013) (reversing as inconsistent with Ninth Circuit law district court's 26 holding, relying in part on Verity, that defendants' gain rather than consumers' loss was proper measure of equitable damages). 24 27 The Court will also have to consider whether injunctive relief is warranted, and assess the appropriate measure of recovery under ROSCA if the FTC proves that claim. 28 44 5 1 V. CONCLUSION 2 Accordingly, the Court GRANTS IN PART Defendant's motion for judgment on partial 3 findings. To the extent the motion is not granted, the Court declines to enter judgment until the 4 close of the evidence. 5 The Court SETS a case management conference for September 4, 2018 at 2:00 p.m. to 6 discuss (1) a schedule and plan for completing what remains of the trial; and (2) whether the 7 parties believe that renewed settlement discussions would be productive in light of the Court's 8 findings and conclusions. 9 IT IS SO ORDERED. 10 Dated: 8/16/2018 11 ______________________________________ HAYWOOD S. GILLIAM, JR. 12 United States District Judge Northern District of California United States District Court 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 45