Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 1 of 28 Page ID #:2436

O 1

2 3 4 5 6

7 United States District Court 8 Central District of California 9 10 Case № 2:19-cv-08704-ODW (PLAx) 11 PRO WATER SOLUTIONS, INC.,

12 Plaintiff,

ORDER DENYING PLAINTIFF’S

13 v. MOTION TO CERTIFY CLASS [81] 14 ANGIE’S LIST, INC. et al.,

15 Defendants.

16 17 I. INTRODUCTION 18 Plaintiff Pro Water Solutions, Inc. brings this putative class action against 19 Defendants Angie’s List, Inc. (“Angie’s List”) and Angi Homeservices Inc. (“Angi”) on 20 behalf of itself and others who advertised their businesses on Angie’s List’s website. 21 (See Second Am. Compl. (“SAC”), ECF No. 52.) Defendants removed this case from 22 the Superior Court of California, County of Los Angeles to the Central District of 23 California on the basis of Class Action Fairness Act jurisdiction. (Notice of Removal 24 ¶¶ 8–37, ECF No. 1.) Presently before the Court is Pro Water’s Motion to Certify Class, 25 (Mot. Certify (“Motion” or “Mot.”), ECF No. 81), which, for the following reasons, is 26 DENIED.1 27

28 1 After carefully considering the papers filed in connection with the Motion, the Court deemed the matter appropriate for decision without oral argument. Fed. R. Civ. P. 78; C.D. Cal. L.R. 7-15. Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 2 of 28 Page ID #:2437

1 II. FACTUAL BACKGROUND 2 Pro Water is in the business of providing water treatment services, and this case 3 arises from its efforts to market its services by advertising on Angie’s List. Angie’s List 4 operates a website that homeowners2 use to locate, evaluate, contact, hire, and rate 5 businesses for various contracting jobs. (SAC ¶ 4.) From 2011 to 2019, Pro Water was 6 a registered business, or service provider, with Angie’s List. (Id. ¶ 16.) During this 7 time period, Pro Water also paid nonparty HomeAdvisor, Inc. to generate leads—that 8 is, to send it the contact information of homeowners who were seeking water treatment 9 services in exchange for a per-lead fee. (See SAC ¶ 75.) 10 Until recently, Angie’s List and HomeAdvisor were two separate companies. 11 After a series of corporate transactions in 2017, Angie’s List and HomeAdvisor both 12 became subsidiaries of a newly created corporate entity, Angi. (See id. ¶¶ 52–57.) Even 13 after this merger, Angie’s List and HomeAdvisor continue to function as separate 14 platforms providing distinct services to homeowners and service providers. (Traughber 15 Decl. ¶¶ 48–51.) To help clarify the role of each business in this controversy, the Court 16 continues to refer to Angie’s List and HomeAdvisor as two separate entities. 17 A. Angie’s List 18 Early iterations of Angie’s List functioned like an enhanced digital Yellow Pages. 19 (See SAC ¶ 107 (drawing this comparison).) Homeowners paid Angie’s List for access 20 to the “Directory,” a database of service providers such as Pro Water. (Traughber Decl. 21 ¶ 3.) For each participating service provider, the Directory, which is still a part of the 22 Angie’s List website, contains a description of the services a service provider offers, a 23 compendium of reviews of the service provider written by other Angie’s List 24 homeowners, and a method by which homeowners can contact the service provider for 25 more information or to hire the service provider. (See id. ¶¶ 6–7.) 26

2 Consumers other than homeowners did and do use Angie’s List, but to avoid potential confusion 27 over the use of the word “consumers,” the Court refers to those who visit Angie’s List seeking to pay 28 service providers to perform tasks as “homeowners.” (See Decl. Chase Traughber (“Traughber Decl.”) ¶ 2, ECF No. 84-5.)

2 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 3 of 28 Page ID #:2438

1 Angie’s List does not charge service providers to be listed in the Directory, but 2 service providers have the option to pay a fee to advertise in the Directory. (Id. ¶¶ 4– 3 5.) Specifically, a service provider can enter into an annual contract with Angie’s List 4 and pay a monthly fee in exchange for Angie’s List’s (1) giving the service provider 5 featured placement in the Directory, and (2) displaying the provider’s “coupons” to 6 homeowners viewing the Directory. (Id. ¶¶ 5, 9; SAC ¶ 27.) “Coupons” are special 7 offers available to homeowners who engage service providers through Angie’s List. 8 (SAC ¶ 99.) An advertiser’s purchase of a coupon on Angie’s List is roughly analogous 9 to purchasing ad space in a traditional Yellow Pages: with both, the service provider 10 pays money to the owner of the directory for featured placement. 11 In 2019, Angie’s List began giving service providers entering into new 12 advertising contracts “express guidance,” or estimates, regarding the number of 13 “contacts” the service provider could expect to receive from homeowners during the 14 first ninety days of advertising. (Opp’n Mot. Certify Class (“Opp’n”) 16, ECF No. 84; 15 Traughber Decl. ¶ 16; see Decl. Joseph Duffy (“Duffy Decl.”) Ex. A (Dep. Jeremy 16 Michael Stewart (“Stewart Dep.”)) 65:2-8, ECF Nos. 84-1, 84-2 (“The way the business 17 operated changed over time. Initially, Angie’s List advertisers would pay the 18 company . . . to appear in the directory. Over time we shifted to try to tie the amount 19 of advertising spend that an Angie’s List service professional had to the amount of 20 contact that they received.”).) A “contact” on Angie’s List occurs when a homeowner 21 reaches out to a service provider about a project, whether by viewing the Directory, 22 claiming a coupon deal, using the SR Path (discussed below), or otherwise. (Traughber 23 Decl. ¶ 14.) Each of these actions counts as a “contact,” whether or not the homeowner 24 eventually hires the service provider. (Id. ¶ 15.) 25 Angie’s List began providing estimates around the same time it introduced a 26 tiered pricing scheme, under which service providers could pay a higher monthly fee 27 for more prominent Directory placement and a higher estimated number of contacts. 28 (Id. ¶¶ 12–13.)

3 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 4 of 28 Page ID #:2439

1 B. HomeAdvisor 2 HomeAdvisor’s business model differs fundamentally from that of Angie’s List. 3 Whereas Angie’s List generates revenue by charging (1) homeowners for Directory 4 access and (2) service providers for advertising space, HomeAdvisor generates revenue 5 by attracting homeowners to its website and then converting those homeowners to 6 individual leads for paying service providers. (See Stewart Dep. 47:1–7; SAC ¶ 56.) 7 HomeAdvisor charges service providers a fee per lead; the service provider receives the 8 homeowner’s contact information and has the option to contact the homeowner about 9 the job. (See SAC ¶¶ 60, 75–77; Stewart Dep. 51:4–12 (“Angie’s List advertising 10 service professionals would pay . . . a fixed non-variable fee every month, and would 11 receive contact volume and a number of contacts without a change in that price, whereas 12 HomeAdvisor service professionals would pay on a per-lead basis . . . .”).) 13 C. SR Path 14 In 2018, after the merger of Angie’s List and HomeAdvisor, Angie’s List 15 introduced a new feature to its website: the SR Path. The SR (or “Service Request”) 16 Path was, and still is, an alternate way for homeowners to find suitable service providers 17 for their projects. (Traughber Decl. ¶ 27.) Defendants explain that Angie’s List 18 introduced SR Path because previously, Angie’s List’s website required homeowners 19 to sign up for an account to access the Directory and contact service providers. (See id. 20 ¶ 3.) Defendants explain that website visitor behavior had changed such that 21 homeowners were increasingly less likely to go to the trouble of creating an account 22 and would thus “bounce off”—that is, they would visit the Angie’s List homepage but 23 ultimately elect not to create an account or log in. (See Opp’n 7.) Angie’s List was 24 losing customers as a result of this behavior, and it introduced SR Path to provide “a 25 lower friction way for consumers to fill out information and be matched to service 26 professionals.” (Stewart Dep. 37:14–16.) 27 The SR Path is available on the Angie’s List homepage for any visitor to use, 28 without the need to create an account or pay a fee. (Traughber Decl. ¶¶ 27–28.) A

4 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 5 of 28 Page ID #:2440

1 homeowner experiences the SR Path as a page or a pop-up in which the homeowner 2 provides their contact information and answers a series of questions regarding the nature 3 and timing of their project. (Id. ¶ 31.) Based on this information, the SR Path provides 4 up to four “results”—that is, four suggested service providers. (Id.) The homeowner is 5 then presented the option to contact the service provider. 6 Angie’s List does not charge service providers any separate fee for appearing in 7 the search results of the SR Path; the SR Path was simply one of the several ways 8 Angie’s List connected service providers to homeowners in exchange for the service 9 providers’ periodic payments. (See Stewart Dep. 188:4–24.) 10 Defendants assert that the SR Path effectively serves it purpose; they attribute the 11 increase in overall contacts to service providers on the Angie’s List website—30% 12 year-over-year in some instances—at least in part to the SR Path. (Duffy Decl. Ex. B 13 (“Traughber Dep. ISO Opp’n”) 231:1–24, ECF No. 84-3.) Relatedly, over time the SR 14 Path has become the primary means by which Angie’s List service providers receive 15 contacts, increasing from about 33% of all contacts to around 65% in 2020. (Decl. Paul 16 Cullen (“Cullen Decl.”) Ex. A (“Stewart Dep. ISO Mot.”) 189:25–190:23, ECF 17 Nos. 81-2, 81-3; Cullen Decl. Ex. B (“Traughber Dep. ISO Mot.”) 255:7–256:6, ECF 18 No. 81-4.) Correspondingly, Angie’s List has made certain changes to its website to 19 emphasize the SR Path and de-emphasize the Directory. For example, in the past, the 20 SR Path search results page—the one that listed the four recommended service 21 providers for the homeowner—provided a link to the Directory below the search results. 22 In March 2021, Angie’s List removed that link, meaning that homeowners on the SR 23 Path could no longer switch directly to viewing the Directory instead. (Traughber Decl. 24 ¶ 34.) 25 In contrast to Defendants’ assertions that the SR Path increased overall contacts 26 to service providers, one of Pro Water’s contentions in this suit is that the SR Path 27 diverted homeowners away from the Directory and thus interfered with the 28 effectiveness of its advertising spend with Angie’s List. (See SAC ¶ 113.)

5 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 6 of 28 Page ID #:2441

1 D. Dual Advertisers; Shared SR Path Search Results 2 During the times relevant to this lawsuit, Pro Water was a “dual advertiser”— 3 that is, it paid Angie’s List periodic fees to advertise there, and it also paid 4 HomeAdvisor lead generation fees on a per-lead basis. (Mot. 7–8.) Because of its role 5 as a dual advertiser, Pro Water learned that, after the merger, Angie’s List and 6 HomeAdvisor began to coordinate their activities such that the search results on Angie’s 7 List’s SR Path did not consist entirely of Angie’s List service providers (that is, service 8 providers paying Angie’s List periodic advertising fees). Instead, up to three of the four 9 SR Path search results could be HomeAdvisor service providers (that is, service 10 providers paying HomeAdvisor for homeowner contact information on a per-lead 11 basis). (Traughber Decl. ¶ 32.) When a HomeAdvisor service provider appeared in the 12 SR Path search results, Angie’s List would transmit the contact information of the 13 searching homeowner to HomeAdvisor, who would then transmit that information to 14 the service provider that appeared in the results. The service provider would then pay 15 HomeAdvisor for the lead. (Traughber Dep. ISO Opp’n 109:1–25.) To be clear: 16 Angie’s List service providers who appeared in SR Path did not receive the contact 17 information of the searching homeowner; Angie’s List service providers paid Angie’s 18 List fees for advertising and featured placement on the Angie’s List website, not for the 19 contact information of homeowners who were potential clients.3 (See Decl. Brad 20 Belnap ¶¶ 29–33, ECF No. 81-11.) 21 Not all categories of service provider on Angie’s List corresponded to a category 22 of service provider on HomeAdvisor, so there were instances where an SR Path’s search 23 results showed no HomeAdvisor service providers whatsoever (because no such 24 providers existed in the category being searched). (Traughber Decl. ¶ 39.) Similarly, 25 3 The record on this point is somewhat muddled, but it appears that, at least conceptually, the SR Path 26 is a distinct feature from Angie’s List’s “Request-A-Quote” feature, which it introduced to its site in 2018. (Traughber Decl. ¶ 14.) The Request-A-Quote feature allows homeowners to submit 27 information about their jobs to service providers and obtain individualized quotes for the jobs. Angie’s 28 List service providers whom homeowners contact using the Request-A-Quote feature do receive the homeowner’s contact information. (Traughber Dep. ISO Opp’n 265:5–15.)

6 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 7 of 28 Page ID #:2442

1 if HomeAdvisor had a corresponding category of service provider but could not provide 2 suitable possibilities based on the parameters the homeowner entered into the SR Path, 3 the search results would display only providers from Angie’s List. (Id. ¶ 33.) 4 Dual-advertiser service providers such as Pro Water who contracted with both 5 Angie’s List and HomeAdvisor were subject to a unique wrinkle in this process that the 6 parties refer to as double-dipping or double-charging. Specifically, when a dual 7 advertiser service provider appeared in the SR Path search results and a homeowner 8 contacted that service provider, HomeAdvisor would charge that service provider a lead 9 generation fee. (Traughber Dep. 149:12–150:24 ISO Opp’n.) From the service 10 provider’s perspective, the service provider was being charged twice—once to advertise 11 on Angie’s List to appear in that homeowner’s SR Path search results, and then again 12 by HomeAdvisor for the lead generated by the provider’s appearance in that same 13 homeowner’s SR Path search results. (Id.; see Mot. 7.) Defendants assert that a 14 “robust” refund process was available for dual advertisers who had been double charged 15 in this way. (Traughber Dep. ISO Opp’n 148:18–149:9, 150:1–6; Traughber Decl. 16 ¶ 37.) Pro Water disagrees; its rebuttal is based on the experience of a former Angie’s 17 List employee who declares that many service providers contacted her because of 18 problems with the refund process. (Decl. Amber Stewart ¶¶ 10–14, ECF No. 93-11.) 19 When the former employee tried to raise the issue, management threatened her with 20 write-ups and termination. (Id. ¶¶ 15–16.) 21 III. PROCEDURAL BACKGROUND 22 Based on the foregoing facts, Pro Water asserts that Angie’s List breached 23 contracts and violated statutes by engaging in five principal categories of conduct: 24 sharing Pro Water’s information with HomeAdvisor; introducing SR Path onto the 25 Angie’s List website; including HomeAdvisor service providers in the SR Path search 26 results; leading service providers to believe that the only way they could appear in the 27 SR Path search results was by advertising through Angie’s List; and double-charging 28

7 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 8 of 28 Page ID #:2443

1 dual advertisers.4 Pro Water’s Complaint asserted four claims variously aimed at 2 obtaining relief for these alleged wrongs. After three rounds of motions to dismiss and 3 amended pleadings, what remains is a single claim for “unfair” business acts or 4 practices under the California unfair competition law (“UCL”), Business and 5 Professions Code section 17200. The allegations pertaining to the first category of 6 conduct—Angie’s List’s sharing Pro Water’s information with HomeAdvisor—related 7 principally to contract claims which are now dismissed. The remaining basis for the 8 UCL claim is Angie’s List’s introducing and maintaining the SR Path on its website, 9 corresponding to the remaining four categories. 10 On April 15, 2022, Plaintiff moved for class certification, seeking certification of 11 the “California Class,” defined as all California-based service providers who paid to 12 advertise on Angie’s List’s website during the limitations period, and the “California 13 HomeAdvisor Subclass,” defined as any member of the California Class who is also a 14 dual advertiser with HomeAdvisor. (Mot. 1.) 15 IV. LEGAL STANDARD 16 Class certification is appropriate only if the parties demonstrate each of the four 17 requirements of Federal Rule of Civil Procedure (“Rule”)23(a) and at least one of the 18 requirements of Rule 23(b). Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 613–14, 19 621 (1997). Under Rule 23(a), a class action is certifiable only if (1) the class is so 20 numerous that joinder of all members is impracticable; (2) there are questions of law or 21 fact common to the class; (3) the claims or defenses of the representative parties are 22 typical of the claims or defenses of the class; and (4) the representative parties will fairly 23 and adequately protect the interests of the class. Fed. R. Civ. P. 23(a); see, e.g., Allen 24 v. Verizon Cal., Inc., No. SACV 08-0774 DOC (MLGx), 2010 WL 11583099, at *2–4 25

26 4 The Court omits discussion of the contract claim that is unique to Pro Water. The Court previously found that portion of the contract claim to be well-pleaded, but that claim relates to Pro Water’s 27 allegation that Angie’s List improperly charged it for a month of advertising services without 28 providing those services. (Order Mot. Dismiss SAC 13–14, ECF No. 59.) That claim is unique to Pro Water and is not a part of the present Motion.

8 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 9 of 28 Page ID #:2444

1 (C.D. Cal. Aug. 12, 2010) (considering the numerosity, commonality, typicality, and 2 adequacy requirements of Rule 23(a)). 3 As for Rule 23(b), Pro Water relies solely on Rule 23(b)(3), which states that a 4 class may be maintained where “questions of law or fact common to class members 5 predominate over any questions affecting only individual members,” and a class action 6 would be “superior to other available methods for fairly and efficiently adjudicating the 7 controversy.” Fed. R. Civ. P. 23(b)(3); see, e.g., Allen, 2010 WL 11583099, at *4–11 8 (considering the predominance and superiority requirements of Rule 23(b)(3)). 9 “Rule 23 does not set forth a mere pleading standard.” Wal-Mart Stores, Inc. v. 10 Dukes, 564 U.S. 338, 350 (2011). Rather, “[a] party seeking class certification must 11 affirmatively demonstrate . . . compliance with the Rule,” id., by a preponderance of the 12 evidence, Olean Wholesale Grocery Coop., Inc. v. Bumble Bee Foods LLC, 31 F.4th 13 651, 665 (9th Cir. 2022). When faced with conflicting or insufficient evidence on the 14 issue of class certification, courts must “judge the persuasiveness and not merely the 15 admissibility of evidence bearing on class certification.” Henson v. Fid. Nat’l Fin. Inc., 16 300 F.R.D. 413, 417 (C.D. Cal. 2014) (citing Ellis v. Costco Wholesale Corp., 657 F.3d 17 970, 982 (9th Cir. 2011)). This “may ‘entail some overlap with the merits of the 18 plaintiff’s underlying claim.’” Amgen Inc. v. Conn. Ret. Plans & Tr. Funds, 568 U.S. 19 455, 465–66 (2013) (quoting Dukes, 564 U.S. at 351). 20 “[A] court’s class-certification analysis must be ‘rigorous’ . . . .’” Id. at 465. 21 This is true of analysis under Rules 23(a) and 23(b) alike. Comcast Corp. v. Behrend, 22 569 U.S. 27, 33 (2013), cert denied, 578 U.S. 1023 (2016). Nevertheless, “Rule 23 23 grants courts no license to engage in free-ranging merits inquiries at the certification 24 stage. Merits questions may be considered to the extent—but only to the extent—that 25 they are relevant to determining whether the Rule 23 prerequisites for class certification 26 are satisfied.” Amgen, 568 U.S. at 466. If the court concludes that the moving party 27 has met its burden of proof, then the court has broad discretion to certify the class within 28

9 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 10 of 28 Page ID #:2445

1 the framework of Rule 23. Zinser v. Accufix Rsch. Inst., Inc., 253 F.3d 1180, 1186 2 (9th Cir. 2001). 3 V. EVIDENTIARY OBJECTIONS 4 As part of its Reply, Pro Water raises objections to evidence supporting 5 Defendants’ opposition. (Pl.’s Obj., ECF No. 93-1.) The Court took the matter under 6 submission one week after Pro Water filed its Reply and objections, which was also one 7 week before the hearing on the Motion was scheduled to take place. (See ECF No. 94.) 8 Then, on May 23, 2022, two weeks after the Reply was filed and on the day the hearing 9 would have taken place, Defendants filed a response to Pro Water’s evidentiary 10 objections, along with objections of its own to evidence supporting Pro Water’s reply. 11 (Resp. Obj., ECF No. 95; Defs.’ Obj., ECF No. 96.) 12 Defendants’ materials are untimely. The Court assumes solely for the purpose of 13 argument that it is permissible to submit materials in rebuttal to a reply; even so, in the 14 absence of a Court order or request for permission, filing such materials two weeks after 15 the reply itself was due, and one week after the Court has already taken the matter under 16 submission, is considered untimely. See Phuong Bich Dzuong v. Deutsch Bank Nat’l 17 Tr. Co., No. SACV 16-536-JLS (KSx), 2016 WL 11000445, at *1 (C.D. Cal. June 23, 18 2016) (striking untimely brief); cf. Christian v. Mattel, Inc., 286 F.3d 1118, 1129 19 (9th Cir. 2002) (“The district court has considerable latitude in managing the parties’ 20 motion practice and enforcing local rules that place parameters on briefing.”). For this 21 reason, the Court STRIKES Defendants’ objection-related materials. (ECF Nos. 95, 22 96.) 23 Pro Water’s objections are more in the character of procedural requests to strike 24 evidence than evidentiary objections. First, Pro Water objects to the Declaration of 25 John Borgo, Angi’s Vice President of Product/Strategy Analysis, and the attachments 26 thereto (together, the “Borgo Materials”), on the grounds that Defendants failed to 27 identify Borgo sufficiently in advance of the class certification deadline to allow Pro 28 Water to depose him. Pro Water fails to show prejudice arising from Defendants’ delay,

10 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 11 of 28 Page ID #:2446

1 and accordingly, Pro Water’s objections are OVERRULED. Pro Water further objects 2 to the attachments to the Traughber Declaration, on the grounds that Defendants 3 produced them untimely. Pro Water likewise fails to show prejudice and this objection 4 is OVERRULED. 5 VI. DISCUSSION 6 The following discussion applies primarily to certification of the Class. The 7 Court addresses certification of the Subclass at the end of this section. 8 A. Theory of UCL Claim; Restitution Model 9 As a preliminary matter, as discussed above, following the settling of the 10 pleadings, Pro Water’s case proceeds on one claim—unfair conduct under the California 11 UCL—and four principal theories: that Angie’s List (1) introduced SR Path onto its 12 website, (2) included HomeAdvisor service providers in the SR Path search results, 13 (3) caused service providers to believe that the only way they could appear in the 14 SR Path search results was by advertising through Angie’s List; and (4) double-charged 15 dual advertisers. Preliminarily, Pro Water’s Motion does not clearly identify which of 16 these theories forms the basis of its bid for class certification. To make this 17 determination and properly distill the class certification inquiry, the Court looks to Pro 18 Water’s proposed restitution model. Pro Water provides the following example of its 19 restitution model: 20 Assume that an Angie’s List [Service Provider] (i.e., one of the CA class 21 members) receives 100 contacts from advertising on the Angie’s List website in any given month. Assume also that half of those contacts came 22 from the SR [P]ath and the other half from other aspects of the Angie’s List 23 website, such as the directory. If half of the jobs coming from the SR [P]ath overall were directed to HomeAdvisor service professionals, that means 24 that half of the opportunities to obtain a contact were taken away from the 25 hypothetical Angie’s List [Service Provider]. That Angie’s List Service 26 Provider advertiser should have been provided an additional 50 contacts in that given month, but for the fact that Defendants gave those opportunities 27 instead to HomeAdvis[o]r advertisers. That, in this hypothetical, results in 28 a 1/3 devaluation of the Angie’s List Service [Provider’s] advertising

11 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 12 of 28 Page ID #:2447

spend. Utilizing this method, the actual number of contacts received by 1 Angie’s List [Service Providers], and the corresponding number of 2 contacts that were essentially directed to HomeAdvis[o]r advertisers, one 3 can compute the dilution of advertising spending on a classwide basis. 4 (Mot. 21–22 (emphasis added).) The italicized language is key and demonstrates that 5 Pro Water’s restitution model is meant to put Angie’s List service providers in the place 6 they would have been had the SR Path search results consisted entirely of Angie’s List 7 service providers. Moreover, Pro Water’s restitution model has two numerical inputs. 8 The first input is the percentage of a service provider’s contacts that came from SR Path 9 (as opposed to the Directory), and the second input is the percentage of SR Path search 10 results that displayed Angie’s List service providers. 11 Based on these observations, it is clear that the restitution model is based solely 12 on the theory that Defendants engaged in unfair business practices by including 13 HomeAdvisor service providers in the SR Path search results—theory number (2), and 14 only theory number (2), as designated above. To be very clear on this point: the 15 restitution model does not attempt to address the double-charging problem unique to 16 the Subclass, nor does it attempt to address any of the other unfair conduct Pro Water 17 ascribes to Angie’s List. Thus, Pro Water seeks certification on one and only one of its 18 theories: that Defendants treated the entire Class unfairly by allowing HomeAdvisor 19 service providers to appear in the SR Path search results on the Angie’s List website. 20 With the benefit of a clear understanding of the nature of the claim Pro Water seeks to 21 certify, the Court proceeds to consider whether certification is appropriate. 22 B. Commonality 23 “To take advantage of Rule 23’s procedure for aggregating claims, plaintiffs 24 must . . . establish ‘there are questions of law or fact common to the class’ . . . .” Olean, 25 31 F.4th at 663 (citing Fed. R. Civ. P. 23(a)). “A common question ‘must be of such a 26 nature that it is capable of classwide resolution—which means that determination of its 27 truth or falsity will resolve an issue that is central to the validity of each one of the 28 claims in one stroke.’” Id. (quoting Dukes, 564 U.S. at 350); Tyson Foods, Inc. v.

12 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 13 of 28 Page ID #:2448

1 Bouaphakeo, 577 U.S. 442, 453 (2016) (“[A] common question is one where the same 2 evidence will suffice for each member to make a prima facie showing [or] the issue is 3 susceptible to generalized, class-wide proof.” (internal quotation marks omitted)). “By 4 contrast, an individual question is one where members of a proposed class will need to 5 present evidence that varies from member to member.” Olean, 31 F.4th at 663 (citing 6 Bouaphakeo, 577 U.S. at 453). “What matters to class certification . . . is not the raising 7 of common ‘questions’ . . . but rather, the capacity of a class-wide proceeding to 8 generate common answers apt to drive the resolution of the litigation. Dissimilarities 9 within the proposed class are what have the potential to impede the generation of 10 common answers.” Dukes, 564 U.S. at 350. 11 Pro Water asserts, and Defendants do not contest, that the sole potential common 12 question this putative class action presents is whether Defendants’ conduct was “unfair” 13 under the California UCL. (Mot. 16 (“[W]hether this is an ‘unfair’ business practice is 14 a central question for every class member, and the answer will not differ based upon 15 each class member’s experience.”); Opp’n 13 (“Plaintiff proffers a single ‘common’ 16 question which is also the ultimate mixed question of law and fact in this case . . . .”); 17 see Reply 3–8, ECF No. 93.) 18 Proceeding with this sole potential common question, the Motion fails on the 19 commonality requirement because (1) variations among Class members change the 20 outcome of the “unfair” analysis; (2) the Subclass is uniquely positioned and shares no 21 common questions with the Class; and (3) Pro Water’s claim is unmeritorious as a 22 matter of law. 23 1. Variations among Class members with respect to “unfairness” 24 First, multiple types of variations among Class members create differences in the 25 outcome of the “unfair” analysis under the California UCL, defeating commonality. 26 Pro Water asserts, and Defendants do not contest, that case law provides two tests 27 for determining whether a business practice is “unfair” in this case: the “public policy” 28 test and the “balancing” test. See Cel-Tech Commc’ns, Inc. v. L.A. Cellular Tel. Co.,

13 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 14 of 28 Page ID #:2449

1 20 Cal. 4th 163, 186–87 (1999). Under the “public policy” test, a plaintiff must show 2 “that a practice violates public policy as declared by ‘specific constitutional, statutory 3 or regulatory provisions.’” Rubio v. Cap. One Bank, 613 F.3d 1195, 1205 (2010) 4 (quoting Gregory v. Albertson’s, Inc., 104 Cal. App. 4th 845, 854 (2002)). 5 Here, although Pro Water suggests that its claim passes muster under either test, 6 (Mot. 16), Pro Water fails to substantiate its suggestion that the public policy test 7 applies. First, Pro Water asserts that Angie’s List’s conduct is unfair under the public 8 policy test because Angie’s List violated antitrust laws. (Reply 5.) But Pro Water does 9 not elaborate on the nature of those antitrust violations; instead, it merely proffers the 10 self-reinforcing assertion that, because the UCL is directed toward enforcement of 11 antitrust laws, a claim based on the UCL must inherently be based on the public policy 12 set forth in antitrust laws. In making this argument, Pro Water suggests that any UCL 13 violation is necessarily also a violation of substantive antitrust law. The Court rejects 14 this suggestion and correspondingly rejects Pro Water’s argument. See Gregory, 15 104 Cal. App. 4th at 854 (requiring plaintiff to “tether[]” UCL claim based on violation 16 of public policy to “specific constitutional, statutory or regulatory provisions”). 17 Pro Water further asserts that Angie’s List’s behavior is unfair under the public 18 policy test because Angie’s List violated “the laws this Court cited in ruling on 19 Defendants’ third motion to dismiss.” (Reply 5.) Yet Pro Water does not actually 20 identify any of the laws the Court purportedly cited. It is Pro Water’s burden to provide 21 the Court with enough information about its theory of unfairness for the Court to make 22 a reasoned judgment about class certification. Pro Water fails in this burden with 23 respect to the public policy test. 24 This leaves the balancing test. Under the balancing test, a UCL plaintiff 25 demonstrating unfair conduct must show that the harm to the consumer outweighs the 26 utility of the practice. Rubio, 613 F.3d at 1205 (citing Lozano v. AT & T Wireless Servs., 27 Inc., 504 F.3d 718, 735–36 (9th Cir. 2007)). “The test of whether a business practice is 28 unfair ‘involves an examination of [the practice’s] impact on its alleged victim,

14 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 15 of 28 Page ID #:2450

1 balanced against the reasons, justifications and motives of the alleged wrongdoer.’” 2 S Bay Chevrolet v. Gen. Motors Acceptance Corp., 72 Cal. App. 4th 861, 886 (1999).5 3 Here, Pro Water contends that the question of UCL unfairness can be resolved in 4 one stroke for the entire Class. The Court disagrees and finds that UCL unfairness is 5 not a common question. 6 First and foremost, service providers who were given estimates regarding their 7 expected number of contacts are in a fundamentally different position with respect to 8 the unfairness of Angie’s List’s conduct when compared to those who were not. As 9 discussed, the theory on which Pro Water seeks certification is that Angie’s List acted 10 unfairly by allowing HomeAdvisor service providers to appear in the SR Path search 11 results, and that Angie’s List service providers lost contacts as a result of this practice. 12 But where an Angie’s List service provider was given an express estimate of the number 13 of contacts it would receive as part of its contractual relationship with Angie’s List, 14 analysis of the fairness of Angie’s List’s conduct fundamentally changes. 15 Certainly, service providers who exceeded their estimates, on one hand, and those 16 who fell short of their estimates, on the other hand, are not similarly situated with 17 respect to UCL unfairness. The issue is whether the question “[d]oes the harm to service 18 providers outweigh the utility of the practice?” will generate common answers across 19 the Class. Rubio, 613 F.3d at 1205. But Angie’s List service providers whose actual 20 contacts exceeded the estimate have significantly less a basis than other Class members 21 for arguing that they were harmed at all by the practice. Regardless of how Angie’s 22 List sources its SR Path search results, such a service provider appears to have received 23 the benefit of its bargain with Angie’s List. Such a service provider likewise has a weak 24

25 5 Some courts recognize a second element to the balancing test. In addition to requiring the plaintiff to show that the harm to the consumer outweighs the utility of the practice, these courts further require 26 that the plaintiffs show that the practice is “immoral, unethical, oppressive, unscrupulous, or substantially injurious to the consumer.” See Drum v. San Fernando Valley Bar Ass’n, 182 Cal. App. 27 4th 247, 257 (2010). Regardless, the Court need not make this legal determination; as discussed 28 herein, certification fails because the first element generates commonality and predominance problems without reference to the putative second element.

15 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 16 of 28 Page ID #:2451

1 basis for arguing that the practice has little utility; if the Angie’s List platform is 2 working well and a service provider is receiving contacts as estimated, that service 3 provider cannot meaningfully argue that one isolated aspect of the platform does not 4 have utility. 5 Conversely, Angie’s List service providers whose actual contacts fall short of the 6 estimate have an entirely different argument with respect to fairness. Service providers 7 aggrieved in this manner have a significantly stronger basis for arguing that including 8 HomeAdvisor service providers in SR Path search results harmed the aggrieved 9 provider in a way that outweighed the utility of the practice to Angie’s List. 10 Moreover, the service providers in these two categories—those whose estimates 11 were met, and those whose estimates were not met—stand in a fundamentally different 12 position with respect to fairness when compared to service providers who never 13 received any estimates at all. Such service providers are unable to point to estimates to 14 tether their fairness argument to any sort of express criteria. These service providers 15 would be left to argue unfairness based on other factors altogether, fundamentally 16 changing the analysis and its result. The percentage of service providers who received 17 estimates from Angie’s List is nontrivial and carves out a significant portion of the 18 Class, defeating commonality.6 19 Yet another schism among Class members cuts across the aforementioned rifts 20 and further divides the Class into three groups: (1) those who began their relationship 21 with Angie’s List before the SR Path was introduced and experienced a decrease in 22 overall contacts as a result of the SR Path; (2) those who began their relationship with 23 Angie’s List before the SR Path was introduced and experienced an increase in overall 24 6 Pro Water asserts in passing that the fact that some Class members were given estimates is “no reason 25 to deny certification, because the class definition can easily be amended to exclude class members who received such agreements.” (Reply 8.) But because the proportion of Class members who 26 received estimates is not at all insubstantial, excluding these would fundamentally change the nature and composition of the Class. Moreover, some Class members did not receive estimates at the 27 beginning of the Class period and began receiving estimates at some point during the class period. 28 Nowhere in its briefs does Pro Water articulate a plausible proposal for dealing with such class members or, more generally, for amending the class definition in so fundamental a way.

16 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 17 of 28 Page ID #:2452

1 contacts as a result of the SR Path, and (3) those who began their relationship with 2 Angie’s List after the SR Path was introduced. As Pro Water itself detailed in its Second 3 Amended Complaint, it and other longtime service providers have come to rely heavily 4 on the benefits and goodwill associated with advertising on the Angie’s List website, 5 such that terminating their relationship with Angie’s List is not a viable option. (See 6 SAC ¶¶ 9–15.) Longtime service providers in group (1) have a colorable argument that 7 it was unfair of Angie’s List to introduce a feature that served to divert homeowners to 8 non-Angie’s List service providers, especially when its service providers are so heavily 9 dependent on Angie’s List. On the other hand, the more fortunate service providers in 10 group (2) have significantly less basis on which to argue that the particular way that 11 Angie’s List chose to populate its SR Path search results was unfair. And none of these 12 observations apply in the first instance to service providers in group (3) because these 13 providers joined Angie’s List after the SR Path had already gone live and therefore had 14 no expectations regarding the effect of the SR Path on any prior contact count. This 15 difference fundamentally changes the nature and result of the multifactor, all-things- 16 considered UCL unfairness analysis as to these service providers. 17 If part of the unfairness analysis includes considering the “impact on [the] alleged 18 victim,” S. Bay Chevrolet, 72 Cal. App. 4th at 886, then the Court must consider the 19 length and depth of a service provider’s relationship with Angie’s List as part of the 20 unfairness analysis, because these details change the extent to which the service 21 provider is impacted. These further variations make the central “unfairness” question 22 even less likely to generate common answers. Dukes, 564 U.S. at 351; see Herskowitz 23 v. Apple, Inc., 301 F.R.D. 460, 469 (C.D. Cal. 2014) (finding no commonality in case 24 where plaintiff accused Apple of double-charging for downloads, where individualized 25 inquiries would be required to determine which class members (1) genuinely clicked 26 “buy” only once but were charged twice, and (2) clicked “buy” twice despite the 27 appearance of a pop-up warning them about the potential double charge and never 28 applied for a refund).

17 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 18 of 28 Page ID #:2453

1 2. Unique position of Subclass 2 The existence of the Subclass as a subset of the Class further decreases the 3 likelihood that the central “unfairness” question will generate common answers across 4 the Class As discussed, the Subclass consists of the dual advertisers: those who paid 5 Angie’s List periodic fees for advertising and featured placement and who also paid 6 HomeAdvisor for individual homeowner leads. The Class’s argument, of course, is that 7 it was unfair of Angie’s List to include HomeAdvisor service providers in the SR Path 8 search results. But the Subclass members are the very HomeAdvisor service providers 9 who were appearing in the SR Path search results. The Subclass members thus paid for 10 and benefitted from the very thing the Class is arguing that Angie’s List should not have 11 done. 12 For this reason, the Subclass, were it to receive restitution based on the model 13 Pro Water offers for the entire Class, would receive what can be fairly characterized as 14 a windfall. Monetary recovery under the California UCL is limited to restitution, which 15 is “the return of the excess of what the plaintiff gave the defendant over the value of 16 what the plaintiff received.” Pulaski & Middleman, LLC v. Google, Inc., 802 F.3d 979, 17 988 (9th Cir. 2015) (quoting Cortez v. Purolator Air Filtrations Prods. Co., 23 Cal. 18 4th 163, 174 (2000)). Here, the members of the Subclass, by virtue of their relationship 19 with HomeAdvisor, already received the benefits of having been included in SR Path 20 search results. Yet, under Pro Water’s proposed classwide restitution model, Subclass 21 members would receive a partial refund of their Angie’s List advertising dollars, meant 22 to put them where they would be if Angie’s List had never included HomeAdvisor 23 service providers in SR Path search results. To reap a benefit, and then to receive 24 restitution based on not having received that benefit, is to collect an inappropriate 25 windfall. See Rest. (3d) of Restitution & Unjust Enrichment (2011) (“Restatement”) 26 § 44(3)(b) (stating that restitution may be limited or denied to avoid 27 “inappropriate windfall”); Lanovaz v. Twinings N. Am., Inc., No. 12-cv-02646-R, 28 2015 WL 729705, at *2 (N.D. Cal. Feb. 19, 2015) (citing to Restatement in determining

18 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 19 of 28 Page ID #:2454

1 available restitution remedy under California consumer protection statutes, including 2 UCL). 3 Ultimately, though, this is a Rule 23 analysis, not a merits analysis, so the Court 4 need not determine whether Pro Water or any particular Subclass member is in fact 5 barred from recovery on the basis of windfall (or perhaps waiver, or any other theory 6 or approach). The point here is that this unique feature of the Subclass significantly 7 changes a part of the “unfairness” analysis. That change must be considered along with 8 all the other facts and circumstances of a service provider’s relationship to Angie’s List, 9 and accordingly, whether Angie’s List’s conduct was “unfair” is not a question that is 10 likely to “generate common answers apt to drive the resolution of the litigation.” 11 Dukes, 564 U.S. at 351. 12 3. Lack of merit of Pro Water’s claim 13 The third and final observation defeating commonality is that the claim of Pro 14 Water itself, as delineated by the restitution model it presents and the corresponding 15 theory of the case, is unmeritorious as a matter of law. 16 Under California law, courts considering whether a business practice is “unfair” 17 to consumers7 under the UCL apply the test articulated under section 5 of the Federal 18 Trade Commission Act. Camacho v. Auto. Club of S. Cal., 142 Cal. App. 4th 1394, 19 1403 (2006). Under this test, “a business practice is ‘unfair’ if (1) the consumer injury 20 is substantial; (2) the injury is not outweighed by any countervailing benefits to 21 consumers or competition; and (3) the injury could not reasonably have been avoided 22 by consumers themselves.” Klein v. Chevron U.S.A., Inc., 202 Cal. App. 4th 1342, 23 1376, n.14 (2012). 24 Here, at root, Pro Water asserts that Angie’s List’s including HomeAdvisor 25 service providers in the SR Path search results was unfair, because Angie’s List both 26

7 As Pro Water is a consumer of Angie’s List’s services, rather than a service provider platform that 27 competes directly with Angie’s List, Pro Water’s UCL claim is a “consumer” UCL claim, not a 28 business competitor UCL claim. See Cel-Tech, 20 Cal. 4th at 187 (recognizing this distinction and its importance).

19 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 20 of 28 Page ID #:2455

1 (1) led its service providers to believe that it would feature only its own service 2 providers on its website, and (2) allowed HomeAdvisor service providers to be included 3 in SR Path search results without informing its own service providers that there was 4 another way to be featured in the SR Path. 5 As to the first contention, Pro Water fails to point to any representation or 6 combination of representations on Angie’s List’s part that plausibly suggest that 7 Angie’s List somehow communicated such a promise to service providers. The Court 8 thoroughly considered this issue in its Order granting Defendant’s motion to dismiss 9 the First Amended Complaint, (Order Mot. Dismiss FAC 13, ECF No. 45), and again 10 emphasized it in the Order on Defendants’ motion to dismiss the Second Amended 11 Complaint, (Order Mot. Dismiss SAC 8). 12 As to the second contention, generally, a business’s practices are not “unfair,” 13 nor do they “harm” clients and consumers, merely because that business offers its 14 products or services under certain terms to some but different terms to others. See 15 Buller v. Sutter Health, 160 Cal. App. 4th 981, 991–92 (2008) (finding nothing “unfair” 16 about a health insurer having a discretionary (i.e., only-if-asked) discount policy). 17 Applying this principle here, Angie’s List service providers received the benefit of their 18 bargain, and the mere fact that providers could appear in SR Path search results under 19 terms offered by one subsidiary of Angi (Angie’s List) and could also appear in SR Path 20 search results under different terms offered by a different subsidiary of Angi 21 (HomeAdvisor) does not, in and of itself, render Defendants’ conduct unfair under the 22 UCL. There must be something more, and here, Pro Water does not point to anything 23 more. See id.; cf. Spiegler v. Home Depot U.S.A., Inc., 552 F. Supp. 2d 1036, 1046 24 (C.D. Cal. 2008) (“[P]laintiffs[] claim that the UCL requires defendants to charge them 25 in accordance with . . . allegedly more accurate measurements . . . . However, that is 26 not the deal that plaintiffs struck, and the UCL cannot be used to rewrite their contracts 27 or to determine whether the terms of their contracts are fair.” (footnote omitted)); cf. 28 also Samura v. Kaiser Found. Health Plan, Inc., 17 Cal. App. 4th 1284, 1299 n.6 (1993)

20 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 21 of 28 Page ID #:2456

1 (observing that California UCL “does not give the courts a general license to review the 2 fairness of contracts but rather has been used to enjoin deceptive or sharp practices”). 3 Pro Water’s unique position as a dual advertiser further supports the conclusion 4 that its claim (as delimited by the only restitution model it offers) has no merit. Pro 5 Water paid HomeAdvisor for leads, and HomeAdvisor obtained those leads by, among 6 other things, getting its clients, including Subclass members, listed in the SR Path. To 7 show that Angie’s List’s conduct was unfair, Pro Water will have to show that the harm 8 it incurred outweighed the utility of Angie’s List’s allowing HomeAdvisor clients to be 9 listed in the SR Path. See Rubio, 613 F.3d at 1205. But in the case of a Subclass 10 member such as Pro Water, the harm engendered by this practice is very low, and the 11 utility is very high. The harm is low because Angie’s List’s inclusion of HomeAdvisor 12 service providers in the SR Path did not actually cause Pro Water to not appear in SR 13 Path. It continued to appear in SR Path, sometimes as an Angie’s List service provider 14 (based on its periodic payments to Angie’s List), and sometimes as a HomeAdvisor 15 service provider (based on its per-lead payments to HomeAdvisor). The utility of this 16 practice is high to both Angie’s List and HomeAdvisor: to Angie’s List, because the 17 practice increased the available pool of service providers and enabled Angie’s List to 18 better connect homeowners to providers; and to HomeAdvisor, because HomeAdvisor 19 was able to sell more leads to its own service providers—the very thing those service 20 providers were paying for and wanted HomeAdvisor to do. This utility also inured to 21 the benefit of both businesses’ clients, including Pro Water. 22 Under these facts, which are set forth in the operative Second Amended 23 Complaint and are otherwise undisputed, Pro Water’s claim for unfair business 24 practices lacks merit as a matter of law. The ability of the Court to dispose of Pro 25 Water’s claim with a judgment-on-the-pleadings-type analysis demonstrates that 26 commonality fails, as there is no need to generate any common answers that might drive 27 the resolution of the litigation, and, more simply, because there is no further litigation 28 to resolve.

21 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 22 of 28 Page ID #:2457

1 C. Predominance 2 Even if there are common questions, class certification would nevertheless fail 3 on the “predominance” requirement of Rule 23(b)(3) because the Class would not be 4 “sufficiently cohesive to warrant adjudication by representation.” Amchem Prods, 5 521 U.S. at 623. 6 Commonality “is about invoking common questions,” Allen, 2010 WL 7 11583099, at *5, whereas the predominance analysis requires district courts to 8 formulate “some prediction as to how specific issues will play out in order to determine 9 whether common or individual issues predominate in a given case,” In re New Motor 10 Vehicles Canadian Exp. Antitrust Litig., 522 F.3d 6, 20 (1st Cir. 2008). The 11 predominance analysis thus “focuses on ‘the relationship between the common and 12 individual issues’ in the case.” Wong v. Chinese Daily News, Inc., 737 F.3d 538, 545 13 (9th Cir. 2013) (quoting Hanlon v. Chrysler Corp., 150 F.3d 1011, 1022 (9th 14 Cir. 1998)). Courts must “identify[] the substantive issues that will control the outcome, 15 assess[] which issues will predominate, and then determin[e] whether the issues are 16 common to the class, a process that ultimately prevents the class from degenerating into 17 a series of individual trials.” Herskowitz, 301 F.R.D. at 469 (quoting Gene & Gene LLC 18 v. BioPay LLC, 541 F.3d 318, 326 (5th Cir. 2008)). The inquiry “is not whether 19 common questions predominate with respect to individual elements or affirmative 20 defenses; rather, the inquiry is a holistic one, in which the Court considers whether 21 overall, considering the issues to be litigated, common issues will predominate.” Id. at 22 470 (citing Amgen, 568 U.S. at 468). 23 Here, individual questions predominate over common questions with respect to 24 both liability and damages, making class certification inappropriate under 25 Rule 23(b)(3). Cf. Pulaski & Middleman, 802 F.3d at 985 (“Entitlement to restitution 26 is a separate inquiry from the amount of restitution owed under California’s 27 UCL . . . .”). 28

22 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 23 of 28 Page ID #:2458

1 1. Liability inquiry 2 First, for many of the reasons discussed in connection with commonality, 3 individual questions predominate over common ones in the liability inquiry. See Dukes, 4 564 U.S. at 375 (Ginsburg, J., concurring) (cautioning against eliding the Rule 23(a)(2) 5 “threshold criterion” of commonality with “the more demanding criteria” of 6 predominance under Rule 23(b)(3)). The first individualized inquiry, discussed above, 7 will relate to how the fact that Angie’s List provided certain Class members an 8 estimated number of contacts affects the unfairness inquiry. This would not be a mere 9 matter of sorting class members into those whose actual contacts exceeded or fell short 10 of the express estimates. The exact degree of excess or deficit is material to the “unfair” 11 inquiry under the UCL, such that the Court would need to conduct individualized factual 12 determinations for the class members who received express estimates (a material 13 portion of the Class by any measure) and then make individualized legal determinations 14 whether, given the accuracy of the estimate along with all other considerations, Angie’s 15 List’s conduct shakes out as “unfair” to that class member. 16 The rift between those who became Angie’s List advertisers before and after SR 17 Path was introduced further defeats predominance. Those who had been advertising for 18 a significant period of time before the introduction of SR Path had a distinct set of 19 expectations with respect to the return on their advertising investment. If the 20 introduction of SR Path caused a given service provider’s rate of contacts to 21 significantly rise, that fact will militate against a finding of unfairness with respect to 22 that provider. If, on the other hand, the introduction of SR Path caused the provider’s 23 rate of contacts to decrease, this would militate for a finding of unfairness. Again, this 24 rift is not something that could be addressed by merely carving the Class into further 25 subclasses, because the degree of the effect of the SR Path is material. The degree of 26 the effect varies from class member to class member and must be considered in light of 27 all the other factors. Individualized inquiries would be required to make the unfairness 28 determination.

23 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 24 of 28 Page ID #:2459

1 2. Damages inquiry; restitution model 2 Moreover, Pro Water’s restitution model is not a fair proxy for restitution given 3 its theory of the case. Thus, individual questions predominate not only the liability 4 inquiry; they predominate the damages inquiry as well. 5 In a typical class action for money damages, parties seeking class certification 6 must provide the court with a proposal or model for calculating damages on a classwide 7 basis. Behrend, 802 F.3d at 1435 (instructing district courts to inquire whether the 8 proposed damages model translates “legal theory of the harmful event into an analysis 9 of the economic impact of that event.” (emphasis omitted)); Leyva v. Medline Indus., 10 Inc., 716 F.3d 510, 514 (9th Cir. 2013) (“[P]laintiffs must be able to show that their 11 damages stemmed from the defendant’s actions that created the legal liability.”). If no 12 such model is suggested, the court may conclude that individualized damages inquiries 13 will overwhelm any common questions, signaling a failure of the predominance 14 requirement. Behrend, 802 F.3d at 1433 (“If the model does not even attempt to do 15 that, it cannot possibly establish that damages are susceptible of measurement across 16 the entire class for purposes of Rule 23(b)(3).”). 17 The same principles apply to Pro Water’s proposed restitution model. As 18 discussed, the remedy at issue here is restitution under the California UCL, which “must 19 be of a measurable amount to restore to the plaintiff what has been acquired by 20 violations of the statutes, and that measurable amount must be supported by evidence.” 21 Colgan v. Leatherman Tool Grp., Inc., 135 Cal. App. 4th 663, 698 (2006). In 22 calculating restitution, California law “requires only that some reasonable basis of 23 computation of damages be used, and the damages may be computed even if the result 24 reached is an approximation.” Marsu, B.V. v. Walt Disney Co., 185 F.3d 932, 938–39 25 (9th Cir. 1999). 26 Applying these principles, the first problem with Pro Water’s restitution model 27 is that it uses appearances in SR Path search results as a proxy for contacts actually 28 received. As indicated by Pro Water’s restitution model example quoted above, a core

24 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 25 of 28 Page ID #:2460

1 assumption of Pro Water’s restitution model is that, if an Angie’s List service provider 2 received twenty contacts from SR Path, but 50% of the SR Path search results during 3 that time were HomeAdvisor service providers, then the Angie’s List service provider 4 should have received forty contacts from SR Path. The problem is that an appearance 5 in SR Path does not necessarily translate to a contact, and impermissible speculation is 6 required to conclude that a higher proportion of appearances directly corresponds to a 7 higher number of contacts. 8 Pro Water’s restitution model suffers from a related problem in its methodology. 9 In Pro Water’s initial example, the second input—the percentage of SR Path search 10 results that were Angie’s List service providers—was 0.5, or 50%. Pro Water would 11 have the Court believe that determining this input for a given service provider is a 12 simple matter of looking to Angie’s List’s data. Yet, Pro Water does not articulate with 13 specificity how it intends to parse out the data to derive this input (either to be used 14 across the class or calculated individually for each class member). 15 An example illustrates the problem. Suppose that an analysis of the entirety of 16 Angie’s List’s SR Path data in the relevant limitations period showed that, on average, 17 40% of the results displayed in SR Path search results went to HomeAdvisor service 18 providers, it would nevertheless be inappropriate to presume that this value for all Class 19 members is 0.4. As Defendants point out without dispute, some categories of Angie’s 20 List service providers had no corresponding category on HomeAdvisor, so no 21 HomeAdvisor providers would appear in SR Path searches in these categories. Angie’s 22 List service providers in such a category would receive an unfair windfall by being 23 assigned a y value of 0.4. 24 More generally, Defendants demonstrate that the balance of Angie’s List and 25 HomeAdvisor service providers in the SR Path search results was constantly in flux, 26 not only over time, but even from search to search. As discussed, if HomeAdvisor could 27 not provide any suitable service providers based on the search parameters the 28 homeowner entered into SR Path (or could find only one), Angie’s List would populate

25 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 26 of 28 Page ID #:2461

1 the remaining results with its own service providers. Thus, the proportion of search 2 results that went to Angie’s List service providers is intimately and immediately 3 dependent upon the parameters each individual homeowner input into the SR Path. 4 These observations cause the Court to ask what subset of SR Path search results 5 Pro Water might propose to use to calculate this input for any given service provider. 6 Pro Water suggests that “[d]efendants . . . can provide information for any given time 7 period regarding . . . (3) the number of times the Class members appeared in SR Path 8 results, [and] (4) the corresponding number of times HA Advertisers appeared in SR 9 Path results.” (Reply 14.) In essence, the argument stumbles on the word 10 “corresponding.” If a Class member appeared in a given SR Path result, then it makes 11 no sense to look at the number of times HomeAdvisor service providers appeared in 12 that same—i.e., the “corresponding”—SR Path result, because that Class member was 13 already appearing in that SR Path search result. Instead, for Pro Water’s model to work, 14 for each Class member, there must be some subset of SR Path search results in which 15 the Class member did not appear, and in which the Class member asserts it should have 16 appeared. 17 This is precisely the aspect of the restitution model Pro Water fails to address: 18 the subset of SR Path search results to be used to derive the second input for each class 19 member. The answer is not simply “all SR Path searches during the Class period”; 20 some Angie’s List service provider categories had no corresponding HomeAdvisor 21 category, so those service providers were not harmed by the appearance of 22 HomeAdvisor service providers in the SR Path, and they are therefore not entitled to 23 any restitution. Nor would the answer be (for, say, Pro Water), “all searches for water 24 treatment service providers during the Class period”; for such searches, it is possible 25 that Pro Water might not have appeared in the SR Path results because another 26 parameter the homeowner provided (such as location or timeframe) excluded Pro Water 27 as a possibility. Thus, it is erroneous to assume that a result on a given SR Path search 28 that went to a HomeAdvisor water treatment service provider would have necessarily

26 Case 2:19-cv-08704-ODW-PLA Document 103 Filed 10/17/22 Page 27 of 28 Page ID #:2462

1 gone to Pro Water instead. For the restitution model to work, Pro Water must be able 2 to articulate a subset of SR Path search results in which it did not appear and would 3 have appeared, and it must articulate a classwide method for analyzing Angie’s List’s 4 data to arrive at the proper input value for each class member. Pro Water fails in this 5 burden. 6 For these reasons, Pro Water’s restitution model fails. When a restitution or 7 damages model fails and the moving party presents no alternative method for 8 calculating damages on a classwide basis, individual issues predominate the damages 9 inquiry, and class certification is properly denied. See, e.g., Behrend, 569 U.S. at 34. 10 D. The HomeAdvisor Subclass 11 Having determined that certification of the Class is not appropriate, the Court 12 considers whether certifying the HomeAdvisor Subclass may be appropriate. Although 13 Pro Water defines the HomeAdvisor Subclass, its Motion lacks a clear indication of 14 how the harm the Subclass suffered is different from the harm the Class suffered and, 15 more generally, does not address any class certification questions or points of analysis 16 that are unique to the Subclass. Moreover, Pro Water provides no restitution model 17 based on the harm specific to the Subclass. On this basis, the Court sees no reason to 18 define a Subclass separate from the Class, and certification of the Subclass is further 19 denied on this basis. 20 /// 21 /// 22 /// 23 /// 24 /// 25 /// 26 /// 27 /// 28 ///

27 Case 2:19-cv-08704-ODW-PLA ee ea Filed 10/17/22 Page 28 of 28 Page

1 Vil. CONCLUSION 2 For the foregoing reasons, the Court DENIES Pro Water’s Motion to Certify 3 | Class with prejudice. (ECF No. 81.) The Court STRIKES Defendants’ Objections 4|| and Response to Objections. (ECF Nos. 95, 96.) 6 IT IS SO ORDERED. 8 October 17, 2022 9 \ 10 SZ: □□

OTIS D. GHT, II

3 UNITED STATESDISTRICT JUDG

14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

28