1 2 3 4 5 6 UNITED STATES DISTRICT COURT 7 NORTHERN DISTRICT OF CALIFORNIA 8
10 DREAMSTIME.COM, LLC, a Florida LLC, 11 Plaintiff, No. C 18-01910 WHA
12 v.
13 GOOGLE LLC, a Delaware LLC, and ORDER GRANTING MOTION FOR DOES 1–10, SUMMARY JUDGMENT 14 Defendants. 15
16 INTRODUCTION 17 Once an antitrust action, plaintiff now hangs its fraud- and contract-based claims on a 18 miscellany of alleged misrepresentations. Defendant moves for summary judgment. For the 19 reasons stated herein, the motion is GRANTED. 20 STATEMENT 21 Plaintiff Dreamstime.com LLC, a leading supplier of digital stock images based in 22 Romania, filed this lawsuit in response to an unexplained fall from its top-ranked position on 23 defendant Google LLC’s search engine. A prior order dismissed Dreamstime’s first 24 explanation — that Google deliberately torpedoed Dreamstime’s search rankings for its 25 competitive advantage, violating the Sherman Act (Dkt No. 72). A subsequent order further 26 trimmed the pleadings, entering judgment for Google on Dreamstime’s claims stemming from 27 one of the two contracts between the parties (Dkt. No. 85). Both orders, however, left 1 Dreamstime’s Section 17200 and remaining contract claims unrestricted, trusting discovery to 2 tease out what happened. 3 1. GOOGLE’S SERVICES. 4 This dispute involves two separate Google services: Google’s free organic search engine 5 and Google’s paid-for advertising service. 6 A. Organic Search. 7 Google Search, the world’s most used internet search engine, indexed trillions of 8 webpages, allowing a user to search the internet for particular information specified by the 9 user’s query. When a user typed a query into the search bar, the website generated a list of 10 responsive webpages ranked by Google according to factors such as the page’s relevance, 11 usability, and age, as well as user-specific factors, including the user’s past behavior and 12 browser settings. Under the hood, the search engine employed hundreds of proprietary 13 algorithms to interpret both the search query and Google’s index of webpages in order to 14 deliver a responsive list of results in seconds (Decl. Lehman ¶¶ 4, 6). 15 Google’s search engine generated organic search results, as opposed to sponsored search 16 results (advertisements). Still, like the alphabetical advantage a company named “AABACO” 17 could gain in a phonebook, webmasters could design websites with ranking algorithms in mind 18 to increase a webpage’s visibility on search engines — a practice known as search engine 19 optimization (SEO). But unlike the alphabetical ordering phonebooks rely on, the 20 confidentiality of Google’s ranking algorithms guarded against easy manipulation. Google 21 did, however, provide public information, through its Webmaster Guidelines, about its search 22 practices. The guidelines explained, in general and broadly applicable terms, what webmasters 23 could and could not do to help ensure that Google would index their websites and make them 24 eligible to rank in Google’s search results, as well as provided general guidance about how to 25 make pages that would be useful to users (Decl. Lehman ¶¶ 13–14). 26 (i) Google’s Honest Results Policy 27 To avoid any appearance of favoritism or impropriety, Google’s Honest Results policy 1 reasonable person might perceive an ulterior motive, such as fostering a current or potential 2 business relationship; delivering targeted benefits of any kind to Google employees, friends, or 3 family; and presenting Google products in a more favorable light. As Senior Search Engineer 4 Eric Lehman explains in his declaration (Decl. Lehman ¶ 9, Exh. C):
5 A fundamental principle for Google is the separation between organic search and Google’s advertising business. Google does not 6 allow websites to buy their way to the top of search results, it does not alter the ranking of its search results to benefit its commercial 7 partners or to induce companies to buy advertising, and it even avoids prioritizing fixing known issues with its search ranking when 8 they are reported through channels that are available only to advertisers. On the contrary, Google maintains a strict wall between 9 its organic search engine and its advertising business. The search team does not base ranking decisions on advertising considerations, 10 and the ads sales and support teams have no visibility into how ranking or other search decisions are made. 11 12 The search team is thus forbidden in most situations from giving special consideration or 13 favors to Google’s advertising or business partners on search-related issues (Decl. Lehman 14 ¶ 10). 15 Search Engineer Lehman appends a selection of internal emails demonstrating Google 16 employees refusing requests to modify its search results for the benefit of third-party clients, 17 including some with advertising budgets far in excess of Dreamstime. Lehman further swears 18 that he conducted a search of the messages sent to the email alias for organic search requests 19 from Google’s business partners and found no documents indicating that Google had been 20 asked to or did provide any private search-related advice to Dreamstime competitors Getty 21 Images or Shutterstock (Decl. Lehman ¶ 11–12). 22 B. Google Ads. 23 In sum, money could not buy a higher organic search ranking. Money could, however, 24 buy access to Google’s audience through its separated online advertising service called Google 25 Ads (formerly, AdWords). Google Ads allowed advertisers to bid for available advertisement 26 space across Google websites (including Google’s search results pages) and third-party 27 websites in Google’s “advertising network” through near-instantaneous, real-time auctions. 1 advantage of troves of user-generated data, Google Ads offered automated tools for advertisers ? 2 to target specific audiences. For example, advertisers chose keywords to describe their 3 advertisements and bid on them. When the selected keywords matched a user’s search query, 4 the advertisement could land as a sponsored result alongside the organic search results, as 5 pictured (Decl. Moser 3, 11, 15-16; Decl. Marian ¥ 44): 6 Google stock photography x § a 7 GE)images @ News ([G) Videos [J Books More Settings Tools About 1,730,000,000 results (0.58 seconds) 8 Ad - www.istockphoto.com/ + Stock Photography - iStock™ by Getty Images 9 Only $12! Images for Websites, Social Media, Brochures, Ads, and other Marketing Material. Discover Powerful Imagery That Helps You Stand Out Today. Types: Royalty Free Stock Photos, ™“< Video, Illustrations & Vectors, Backgrounds & Templates. 10 Ad □ Whutterstock.com/ Stock Stock Images - Find the Perfect Image for You 1 1 Che Fhotos, Vectors, Icons and Much More. Simple Licensing, Dedicated Customer Support. Get Instant Inspiration Exploring Eye-Catching Photos. Try, Share, Save or Buy Photos Now! Free Image Resizer. Signs & Symbols, Founded in 2003, a 12 1-Month Free Trial - $0.00/mo - Get 10 Free Images - More *
3 13 Ad + www.123rf.com/ » (888) 961-7475 Images From Only $0.19 Each - 123RF Stock Photos - 123RF.com ly$ h k Ph a Right Images For Your Business. Unlimited Commercial Print Runs. Get Started!
3 14 Ad - www.depositphotos.com/ + Depositphotos™ - Photostock - Buy Stock Photos . o 1] 5 Price meets quality. Advanced search filters. Find the high-resolution photos you need.
BZ www.shutterstock.com photos + a 16 Stock Photos, Stock Photography, and Royalty Free Images ... Organic Browse royalty-free stock photos, pictures and images. Find the ideal stock photo by browsing oO 17 Search our stock photography collection. Photos. All images; Photos Results Stock Photos, Images Stand Up for Pride - Travel Natural Backgrounds S kiped ki » Stock_phe hy Z, 18 en.wikipedia.org » wiki » Stoc! photography - Stock photography - Wikipedia Stock photography is the supply of photographs which are often licensed for specific uses. The 1] 9 stock photo industry, which began to gain hold in the 1920s, has ... 20 The advertising service also provided bidding tools to determine when and at what price 21 an advertiser would bid for available ad space. Manual bidding tools allowed an advertiser to 22 price its bids based on specific advertising goals, like generating web traffic or increasing 23 brand awareness. Google also provided automated bidding tools that used a variety of auction- 24 time signals, such as the viewer’s device, location, time of day, language, and operating system 25 to target specific actions users would take after interacting with an ad, such as making an 26 online purchase or calling a business phone. Dreamstime employed a variety of these manual 27 and automated bidding tools (Decl. Moser {J 18-24). 28
1 An advertiser’s bidding strategy did not determine its monthly bill from Google. Rather, 2 Google required advertisers to set average daily budgets, which Google then multiplied by the 3 average number of days in a month (30.4) to determine an advertiser’s monthly charging limit. 4 This system allowed Google to outspend the set daily budget on a given day (“overdelivery”) 5 to account for daily fluctuations in online traffic, while never exceeding the total monthly 6 charging limit (Decl. Moser ¶¶ 25–32). 7 2. DREAMSTIME BECOMES A GOOGLE CUSTOMER. 8 Dreamstime opened its advertising account with Google in 2004, quickly becoming one 9 of Google’s largest advertising clients in Romania. Google’s Advertising Program Terms 10 governed the parties’ advertising relationship. Dreamstime agreed to three successive versions 11 of the Ads Agreement throughout the relevant time period — March 2014 to September 2018, 12 according to Google — but neither side claims that any material difference between the three 13 versions exists. Each agreement expressly authorized Google to suspend or remove specific 14 advertisements from its network, to cancel advertising accounts, and to otherwise enforce 15 Google’s advertising policies. None of the three agreements made any guarantees about how 16 advertisements or advertisers would perform on Google Ads. And, importantly, none of the 17 agreements included any statement about Google’s separate organic search engine, let alone 18 any guarantees about how an advertiser’s website would appear in search results (Decl. 19 Moser ¶¶ 4–9, Exhs. A–C). 20 Over the course of the relationship, Google responded to Dreamstime’s substantial 21 business by offering it personalized advertising support services. By 2012, Google offered 22 Dreamstime a dedicated, European-based ad-support team (Decl. Enache ¶ 3–5, 7). 23 Dreamstime’s CEO Serban Enache swears that Google generally provided direct answers 24 to any search-related issues that came up in the first decade of Dreamstime’s relationship with 25 Google. In support, CEO Enache provides emails he exchanged with Google Ads personnel 26 discussing the possibility of meeting with the organic search team, as well as an email 27 exchange with Google Search Engineer Gary Illyes in December 2012. Although Engineer 1 Dreamstime provides no translation. Although the record is unclear due to the lack of 2 translation, it appears CEO Enache initiated the conversation himself, rather than through a 3 sales channel in violation of Google’s Honest Results Policy. Engineer Illyes responded in 4 English, answered a search-related issue, and ended the email, “I'll reply on the public thread 5 as well to hush away some myths” (Decl. Enache Exhs. A, AA). 6 3. DREAMSTIME’S SEARCH-RANKING DROP. 7 During Dreamstime’s first decade with Google, the stock-images website generally 8 ranked in the top three search results positions for key industry search terms, such as “stock 9 photos.” Beginning around October 2015, however, Dreamstime slipped to the bottom of the 10 first page of search results for these terms and eventually further. This drop coincided with a 11 drop in Dreamstime’s organic search traffic (Decl. Enache ¶¶ 79, 84, Exh. CC). 12 CEO Enache first discussed Dreamstime’s search-ranking drop with its ad-support team 13 in the first half of 2016. In the second half of 2016, CEO Enache repeatedly raised the issue to 14 Google representatives during meetings and by email. CEO Enache swears that during most of 15 the meetings, he raised his suspicion that an algorithmic update caused the drop and Google 16 representatives assured him that was not the case. Instead, the Ads representatives told CEO 17 Enache that the problem likely had to do with Dreamstime’s website (Decl. Enache ¶ 85–86, 18 Exh. DD). 19 When another significant search-ranking drop came in September 2016, CEO Enache 20 asked his ad-support team to put him in touch with Search Engineer Illyes, whom he had 21 spoken with previously, to signal a technical glitch Dreamstime found. Google Ads 22 Representative Ana Sipciu replied, telling CEO Enache that Engineer Illyes would not speak 23 directly with him and required that Dreamstime post its issues on Google’s public forum, citing 24 Google’s Honest Results Policy. Ads Rep Sipciu explained, “Internal policy forbids the 25 Search teams to intervene directly on any issues escalated through sales . . . unfortunately I 26 can’t put [Engineer Illyes] in direct contact with you, it’s against the policy and it would not 27 help the issue here at all (quite the contrary — as they wouldn’t be able to investigate the issue 1 In response, CEO Enache sent the following email (Decl. Enache ¶ 89, Exh. FF):
2 I wasn’t aware he is in the organic team, my message was meant as a reply to his comment. Ok, we will try to post through the link you 3 sent us, hopefully someone will read it this time.
4 I really can’t understand why someone would pay attention to an anonymous post, but ignores a legit complaint/ticket etc. I realize 5 this is the policy, but nevertheless it is a frustrating situation. We have reached the point where we ask ourselves whether it makes 6 sense to continue to invest in Adwords. So far we were making profit after 12 months. With the organic going downhill for 12 7 months in a row (last update on Penguin brought another major hit, we lost 5 % more in just 2 weeks), it's now unclear whether we have 8 any positive margin left. We have invested $500k in our main account (a total of $700k) in the last 30 days, but converted only 9 $144k. If we deduct the branded campaigns, we gained $70k in sales for $500k invested. This is simply unsustainable. The organic 10 traffic has a 20 % YOY decrease and its downhill trend accelerates. In Oct 2014 organic sales were above $300k vs Adwords $70k. In 11 Oct 2016 we will probably have $150k in organic sales and $80k in Adwords. Our decline for new customers in August was slowing 12 down at -19 % YOY after a disastrous -36 % YOY in July. Another Penguin hit and we're now at -25 % YOY. It's the level we had in 13 2013 and this has long term impact. The only source that is down is organic. 14 It baffles me that Google remains passive to our struggle and I can't 15 help but notice that Shutterstock remains untouched. Other sites go up or down, Shutterstock's rank is always preserved. It may be that 16 they are not favored, but it certainly looks that way. No matter how many times we reported them for being favored, other competitors 17 for using black-hat SEO tricks, no visible measures were taken against any of them. Even when the rules were clearly broken no 18 visible penalty was applied. They are definitely born under a lucky star. 19 Please don’t get me wrong, I know you can’t change these things or 20 the policy. But it is the only way to provide our feedback. 21 Dreamstime also followed Ads Rep Sipciu’s direction and posted on Google’s public 22 forum to receive a response from Engineer Illyes. While Engineer Illyes did respond, his full 23 response is not included in the exhibit CEO Enache provides to his declaration (Decl. Enache 24 ¶ 91, Exh. GG). 25 This order pauses to note that Dreamstime, at oral argument, highlighted CEO Enache’s 26 email as triggering Google’s duty to reveal information it allegedly concealed concerning 27 Dreamstime’s search performance. This email, according to Dreamstime, sent shockwaves 1 through Google’s ad-support team, worried that Dreamstime would begin cutting its 2 advertising spending if its search ranking did not improve (Decl. Marquart Exhs. O, P, Q, S). 3 Around the same time, Ads Rep Sipciu invited Dreamstime’s advertising team to Dublin 4 for a meeting with its Google ad-support team and specialists. CEO Enache responded “I don’t 5 know how to approach the organization of such a meeting because, as you know, we are 6 suffering a lot from the fall of the organic . . . I don’t want to refuse the invitation, but I don’t 7 believe that we can organize anything until we solve this problem.” Sipciu responded (Decl. 8 Enache ¶ 92, Exh. HH):
9 Yes, I know, the entire team is aware of the situation: ( - and we are doing everything we can to send all the signals that we can internally 10 about this issue. I hope that somebody from the organic team contacts you soon, I sent them the forum link. 11 Ads Rep Sipciu then emailed her team explaining that she doubted anything could be 12 done on the advertising side to make up for the organic issues and noted that she remained 13 open to any ideas the team had to somehow help address Dreamstime’s organic complaints 14 (Decl. Marquart Exh. O). 15 4. GOOGLE CONNECTS DREAMSTIME WITH SEO CONSULTANT. 16 A month later, Sipciu emailed CEO Enache explaining that she had been “trying to 17 understand if/where there is a SEO problem,” and sought out an external specialist from 18 Germany, Professor Mario Fischer, to prepare an initial analysis to explain Dreamstime’s 19 search-ranking decline. Professor Fischer, an academic and consultant specializing in search 20 engine optimization, came recommended by Google’s German team and provided the initial 21 assessment free of charge. Sipciu noted that “at a first glance [the analysis] seem[ed] to 22 confirm [Engineer Illyes’s] suggestion made on the forum — content [could] be improved.” 23 Sipciu also offered to give CEO Enache Professor Fischer’s “contact details so [Dreamstime 24 and Fischer could] work together, without any Google involvement” (Decl. Marquart Exh. R, 25 T, S; Decl. Enache ¶ 93, Exh. II) (emphasis added). 26 Professor Fischer summarized his findings as follows (Decl. Enache Exh. II): 27 1 [¶]
2 About the second assumption (error in the [core algorithm] Penguin update) I also think it doesn’t stand. The last core update, into which 3 Penguin was integrated, was well accepted and generally did not register any issues. As far as I know, it didn’t downgrade any 4 website. Many good sites were pushed up, including Dreamstime, as you can see in the slides. 5 The rollout of the update ran somewhat longer, it didn’t only take a 6 day. But we can see that during the period of the rollout the ranking went up significantly. The decrease came after this, with no relation 7 with the Google update.
8 It’s possible that [Dreamstime] tampered with the Disavow file without really knowing what they were doing and this could 9 possibly be a reason for the decline.
10 In my opinion, the real problem lies in the weak content of the site. The individual pages are not rich enough — only 30 descriptive 11 words for a picture.
12 This is enough for a user choosing a photo, but search engine robots/indexers need more text. The pages have an average of less 13 than 600 words, but 90% of them repeat on all pages. So Google naturally marks it as “doubled” content and stops taking it into 14 consideration. So there is not much left over to index.
15 The new websites ranked above (startupstockphotos.com is very weak, butpexels.com is better than Dreamstime) are not a measure 16 of an error in the algorithm. It would be too easy if just a few key figures could explain the different rankings of domains. []. 17 Getting good ranking is naturally more difficult for an image 18 platform in text searches. []. But we need to investigate further and to know exactly what the client did, we can’t assess this from the 19 exterior by making a few simple investigations. 20 Discussing Professor Fischer’s analysis, CEO Enache swears: “I reasonably assumed 21 and took the analysis, coming from Google, to mean that no Google algorithmic change or 22 glitch was the source of the devastating problem we were experiencing. To the contrary, I was 23 led to believe there must be some problem on Dreamstime’s end as the Fischer analysis 24 purportedly suggested. Given that I had no access or insight into the inner workings of 25 Google’s proprietary and secret algorithm, I took the analysis Google provided at face value 26 and had no basis at the time to doubt it” (Decl. Enache ¶ 94). 27 5. CEO ENACHE CONTACTS GOOGLE EXECUTIVES. 1 In November 2016, around the same time Dreamstime received Professor Fischer’s initial 2 analysis, CEO Enache contacted executives on Google’s Partnerships team, seeking to be put 3 in touch with Google’s legal team to discuss a host of issues that eventually materialized into 4 this lawsuit. Initially, Google Partnerships Executive Brian Lam met for a “business-to- 5 business call (no legal)” with CEO Enache and another Google advertising executive. 6 Following the call, CEO Enache provided a list of the issues Dreamstime then had with 7 Google, including those related to its organic search decline. Eventually, Executive Lam 8 responded in part (Decl. Enache ¶ 95, Exh. KK): 9 The Partnerships team isn’t the right organization to address 10 concerns about search ranking. As [Ads Rep Sipciu] has communicated in the past, we cannot provide guidance to our 11 clients/partners around search ranking. I’m aware that [Sipciu’s] team has referred Dreamstime to outside SEO evaluators in the past, 12 so I would continue to work with those parties and consider implementing their findings and recommendations. 13 CEO Enache responded with confusion, noting that its initial request had been for a contact 14 from Google’s legal team, and again asked for a legal contact, which Executive Lam provided. 15 Internally, Executive Lam had worked with Ads Rep Sipciu and another Partnerships 16 executive, Josh Capilouto. Executive Capilouto, before Executive Lam sent a generic response 17 to CEO Enache explaining Google’s honest results policy, wanted to review Dreamstime’s 18 concerns closely, speak with the “relevant Search product folks, and connect with Search p- 19 counsel” to see if Google could and would provide a more substantive response. Due to 20 redactions for privilege, it is unclear what Executive Capilouto learned from his investigation. 21 Ultimately, the record reflects that after meeting with “legal,” Executive Capilouto provided 22 the draft email Executive Lam sent to CEO Enache (Decl. Marquart Exhs. K–M) 23 Relying on discussions with Google representatives in Europe and Mountain View, as 24 well as the third-party analysis of Dr. Fischer, Dreamstime CEO Enache swears he started to 25 doubt his suspicions of an algorithm change and instead invested millions in Ads spending to 26 offset the loss in organic search traffic while hiring additional SEO consulting firms to assess 27 1 its website’s issues. Using the consultant reports, Dreamstime began investing in website 2 upgrades to resolve its content issues (Decl. Enache ¶¶ 94–97). 3 6. UPDATE TO GOOGLE’S SALIENT TERMS ALGORITHM. 4 Although both sides now agree that Google did not deliberately demote Dreamstime’s 5 search ranking, Dreamstime insists that a late 2015 update to one of Google’s algorithms did 6 cause Dreamstime’s search-ranking drop. For its part, Google insists that the update is 7 extremely unlikely to be the explanation for any ranking changes Dreamstime may have 8 experienced (Decl. Lehman ¶ 29). 9 The update in question tweaked one part of a multipart algorithmic signal known as the 10 salient terms signal. For each webpage in Google’s search index, the salient terms signal 11 generated terms that should have been associated with that webpage, helping Google’s search 12 engine find an appropriate list of webpages that might be responsive to a search query. The 13 change was straightforward: The update would cause the portion of the algorithm in question 14 to “give more weight to (and thereby make more “salient”)” certain words based on how the 15 webpage displayed them (Decl. Lehman ¶¶ 30–32, Exh. M). 16 Google’s search team conducted several experiments to test the proposed change. The 17 primary experiment, one long employed by Google to test whether proposed changes made 18 search results better or worse, involved showing before and after results to outside raters based 19 on various randomly generated search queries. The raters there found the proposed update to 20 the salient terms algorithm to be positive. The experiment included no mention of Dreamstime 21 or any of its webpages (Decl. Lehman ¶ 33). 22 In addition to the primary experiment, the testing process involved a “novel side-by-side . 23 . . experiment in which outside search raters compared the top 25 salient terms generated by 24 the . . . algorithm both before and after the proposed algorithmic change for various randomly 25 selected test webpages.” Raters evaluated 400 English language webpages and 2,300 26 webpages in total. One of these webpages belonged to Dreamstime, a page selling an image of 27 two flamingos. The raters determined that the flamingo webpage constituted a “loss” because 1 words. More specifically, out of the top twenty-five terms, six became numbers, rather than 2 words — without the change, only three of the top terms had been numbers. 3 Dreamstime’s webpage was not the only loss identified by the raters in the side-by-side 4 experiment. Among the various other webpages found to be “losses” by the raters included 5 webpages belonging to a Dreamstime competitor and Google’s own properties, including 6 YouTube and Google+ webpages (Decl. Lehman ¶ 35). 7 Google’s Search Engineer Eric Lehman opined that the change to the salient terms signal 8 remains “extremely unlikely to be the explanation for any ranking changes that Dreamstime 9 may have experienced in Google Search” referencing contemporaneous notes accompanying 10 the launch report that stated there had been no “correlation between [the change] and any effect 11 on ranking” (Decl. Lehman ¶ 29, Exh. M). 12 Dreamstime’s expert, on the other hand, opined “that a Google algorithmic change, such 13 as the change to the . . . Salient Terms signal . . . is the most likely cause for Dreamstime’s 14 organic traffic loss beginning in November of 2015.” Specifically identifying “what Google 15 knew, or could have known, but failed to disclose,” Dreamstime Expert Stricchiola opined that 16 “[a]s of at least [November 2015], Google had in its possession documentation that showed 17 that a search algorithm change it had pushed live to its search engine had a direct, and 18 negative, impact on the Dreamstime website” — referring to the launch report (Stricchiola 19 Dec. ¶¶ 70–73, 101, 123). 20 7. INCREASED ADVERTISING SPENDING. 21 According to Dreamstime, Google concealed information about the salient terms update 22 to induce Dreamstime to spend more on Google Ads. But Dreamstime produces no evidence 23 showing that any of Google’s advertising employees knew of the salient terms update. Instead, 24 the inducement half of Dreamstime’s theory is based on general assertions that Google 25 provided an ad-support team to Dreamstime and the Google representatives on that team held 26 themselves out as having superior knowledge they did not in fact possess. Google purportedly 27 hid the fact that these ad-support personnel remained members of Google’s sales team, secretly 1 Still, CEO Enache swears that “Dreamstime would have budgeted millions less for [its 2 Google advertising campaigns] were it not for Google’s statements and omissions about the 3 causes for [Dreamstime’s] diminished organic search performance.” For proof, CEO Enache 4 notes that “in November and December of 2016, Dreamstime spent approximately $1.4 5 [m]illion on AdWords, compared with approximately $640,000 in those same months in 2019” 6 (Decl. Enache ¶¶ 96, 98). 7 * * * 8 Dreamstime filed this action in March 2018. As discussed, extensive Rule 12 practice 9 ensued, eventually trimming away Dreamstime’s antitrust claims and one set of contract 10 claims. Dreamstime’s unilateral recasting of its claims also bears mentioning. Dreamstime 11 opens its opposition to the pending motion as follows (Opp. at 1):
12 This is no longer an antitrust case. There is no need to show harm to competition or targeting of Dreamstime. This case centers on 13 what Google told (and did not tell) Dreamstime about its organic search performance and about how specific types of ads worked, not 14 on any anticompetitive motives. . . . Google misrepresented material facts to Dreamstime to induce it to spend more on AdWords than it 15 would have if it had known the true facts. . . . It turns out Google’s representatives did not have the expertise they claimed, did not 16 understand how the technology worked, and most importantly told many specific lies about how certain ads worked and why 17 Dreamstime’s organic search performance was suffering. Those lies lured Dreamstime to overspend on AdWords. 18 In line with its opening statement, Dreamstime now foregoes any argument that Google 19 deliberately torpedoed Dreamstime’s search ranking. Dreamstime also foregoes any argument 20 that Google breached any express contract terms, either by charging Dreamstime more than 21 allowed by its advertising budgets or that Google unfairly disapproved certain Dreamstime 22 advertisements. 23 Dreamstime is left with two claims, one for violation of Section 17200 under all three of 24 its prongs, and the other for breach of the covenant of good faith and fair dealing implied in the 25 parties’ advertising agreement. Each of these claims are based on both search-related and ad- 26 related alleged misrepresentations. 27 Google now moves for summary judgment. Dreamstime opposes. 1 This order follows full briefing and a telephonic hearing. 2 ANALYSIS 3 Summary judgment is appropriate when there is no genuine dispute as to any material 4 fact. A genuine dispute of material fact is one that “might affect the outcome of the suit under 5 the governing law.” Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 247–48 (1986). 6 1. ORGANIC SEARCH CLAIMS. 7 As stated, Dreamstime claims that Google fraudulently concealed the reasons for 8 Dreamstime’s search ranking drop in order to boost its advertising revenue. Whether dressed 9 as a breach of the covenant of good faith and fair dealing implied in the parties’ Ads 10 Agreement or as an unfair, unlawful, or fraudulent business practice in violation of Section 11 17200, Dreamstime’s concealment theory has no merit. 12 A. Dreamstime’s Organic Search Issues Fell Beyond the Scope of the Advertising Agreement’s Implied Covenant. 13 Every contract imposes on each of the parties a duty of good faith and fair dealing in the 14 performance of the contract such that neither party shall take any action that effectively 15 destroys or injures the right of the other party to receive the benefits of the contract. “[T]he 16 scope of conduct prohibited by the covenant of good faith is circumscribed by the purposes and 17 express terms of the contract.” Carma Developers, Inc. v. Marathon Development California, 18 Inc., 2 Cal.4th 342, 373 (1992). 19 The advertising agreement governed the parties’ advertising relationship, providing 20 Dreamstime access to Google’s network and advertising products for a price. Unsurprisingly, 21 as stated, the agreement made no promises related to Google’s separate organic search engine, 22 nor did Google affirmatively disclaim or limit its liability concerning any potential promises 23 related to organic search. Indeed, Google repeatedly informed Dreamstime of its policy 24 against giving special consideration or favors to Google’s advertising partners on search- 25 related issues. The implied covenant cannot write into the advertising agreement a duty or 26 promise related to organic search (Decl. Moser ¶ 9; see Hearing Tr. at 33; Decl. Enache 27 Exhs. L, FF). 1 Dreamstime nevertheless attempts to hinge its implied covenant claim on Google’s 2 provision of free, “premium ‘support’ in performance of the AdWords contract[.]” Under this 3 theory, Dreamstime argues that although “Google may not initially have been obligated to 4 explain Dreamstime’s drop in search performance . . . once it set out to do so it had to do so in 5 good faith, to deal fairly.” For support, Dreamstime quotes Badie v. Bank of America, 67 Cal. 6 App. 4th 779, 795 (1998), which noted that “the exercise of discretionary powers conferred on 7 a party by contract must also be evaluated under the implied covenant[.]” Badie and the 8 decisions it relied on involved contracts that expressly provided one side with discretion to 9 perform (or not). This is not our situation. The advertising agreement neither required nor 10 conferred upon Google the discretion to provide a dedicated ad-support team to Dreamstime, 11 let alone support on search issues beyond the scope of the agreement. Even if Google did 12 knowingly misrepresent the cause of Dreamstime’s search-ranking drop, the ads agreement 13 does not provide Dreamstime’s remedy. 14 This order finds that the alleged misinformation about Dreamstime’s organic search 15 performance fell outside the scope of the advertising agreement and thus could not (and does 16 not) breach the covenant of good faith and fair dealing implied in that agreement. Summary 17 judgment on the search-related implied covenant claim is GRANTED. 18 B. Section 17200: Google Had No Duty to Disclose What it Did Not Know or Believe to be True. 19 Section 17200 proscribes practices which are unlawful, unfair, or fraudulent. 20 Dreamstime claims that Google engaged in all three. The claim under the unlawful prong is 21 derivative of Dreamstime’s breach of the implied covenant of good faith and fair dealing and 22 falls with it. Dreamstime likewise tethered the alleged unfair conduct to its implied covenant 23 claim, so it falls as well. 24 Turning to the fraud prong, to establish a Section 17200 violation for “fraudulent” 25 business practices, “it is necessary only to show that members of the public are likely to be 26 deceived.” In re Tobacco II Cases, 46 Cal.4th 298, 312 (2009). Unlike common law fraud, 27 1 prove that a fraudulent deception under Section 17200 was actually false, known to be false by 2 the perpetrator, or reasonably relied upon by a victim who incurs damages. Ibid.; see also 3 Clemens v. DaimlerChrysler Corp., 534 F.3d 1017, 1025–26 (9th Cir.2008). Rather, “[t]he 4 determination as to whether a business practice is deceptive is based on the likely effect such 5 [a] practice would have on a reasonable consumer.” Morgan v. AT & T Wireless Services, Inc., 6 177 Cal.App.4th 1235, 1256 (2009). 7 For an omission to be actionable under Section 17200, “the omission must be contrary to 8 a representation actually made by the defendant, or an omission of a fact the defendant was 9 obliged to disclose.” Daugherty v. Am. Honda Motor Co., 144 Cal.App.4th 824, 835 (2006). 10 A defendant has a duty to disclose information in four circumstances: (1) When the defendant 11 is the plaintiff's fiduciary; (2) when the defendant has exclusive knowledge of material facts 12 not known or reasonably accessible to the plaintiff; (3) when the defendant actively conceals a 13 material fact from the plaintiff; or (4) when the defendant makes partial representations that are 14 misleading because some other material fact has not been disclosed. Hodsdon v. Mars, Inc., 15 891 F.3d 857 (9th Cir. 2018) (relying on Collins v. eMachines, Inc., 202 Cal.App.4th 249, 255, 16 (2011)). 17 Dreamstime’s fraud claims fail for one simple reason. Nobody at Google, either on the 18 advertising or the organic side, ever believed that Dreamstime’s ranking dropped as a result of 19 the update to Google’s salient terms algorithm. While fraudulent practices under Section 20 17200 do not require proof of scienter, an omission theory does require that the defendant 21 possess actual knowledge of the concealed fact. 22 The best evidence Dreamstime has to show that Google knew the update affected 23 Dreamstime’s search ranking is the flamingo webpage and corresponding launch report. But 24 the evidence still falls short. The launch report indeed demonstrates that at least some 25 members of the organic search team knew that some testing of the salient terms update 26 indicated a possible negative effect on one of Dreamstime’s product pages (of which, 27 Dreamstime has millions) as to the salient terms algorithm. What the launch-report evidence 1 the update had as to the salient terms algorithm also translated to a drop in Dreamstime’s 2 search ranking. To the contrary, contemporaneous notes to the launch report stated that there 3 existed no “correlation between [the change] and any effect on ranking” (Decl. Lehman ¶ 29, 4 Exh. M). 5 Dreamstime makes a number of claims based on its belief that Google and Professor 6 Fischer, the third-party SEO consultant Google suggested to Dreamstime, constituted the same 7 entity. But the record reflects that Google did not pay for Professor Fischer’s analysis and Ads 8 Rep Sipciu expressly noted that Dreamstime could contact Fischer without Google. Nor is 9 their evidence demonstrating that Google weighed in on the results of the analysis or otherwise 10 determined Professor Fischer’s findings. Dreamstime attempts to claim that Google vetted the 11 report, but Ads Rep Sipciu merely noted that the Fischer analysis seemed to draw the same 12 conclusion Search Engineer Illyes reached responding to Dreamstime’s post on the public 13 forum. 14 At oral argument, Dreamstime argued that CEO Enache’s October 2016 email to the ad- 15 support team and the team’s response thereto imposed a duty upon Google to disclose details 16 concerning the salient terms update. In other words, through threats and incessant demands, 17 Dreamstime cajoled Google’s ad support team into trying to help Dreamstime with its search 18 issues. By suggesting Professor Fischer to Dreamstime to help address its organic search 19 issues, Dreamstime asserts that Google then gained a duty to disclose the launch report. Ad 20 Rep Sipciu’s gesture, however, did not change the fact that Sipciu had no knowledge of the 21 launch report, nor did it demonstrate that anybody at Google believed an algorithmic update 22 caused Dreamstime’s search-ranking drop. 23 In any event, the alleged fraudulent business practices here are each variations on the 24 same theme: Google refused to hand over, to one of its advertiser clients, confidential 25 information concerning a global update to its salient terms algorithm. Dreamstime had no right 26 to it. This order finds that Google’s refusal to relinquish confidential data to its advertisers was 27 not likely to deceive members of the public. 2. ADS-BASED CLAIMS. 1 Google further moves for summary judgment on Dreamstime’s ads-based claims 2 asserting that the AdWords Agreement disavows any guarantees about advertisement results, 3 and assigns to advertisers the risks of disappointing results. Section 9 of the AdWords 4 Agreement provided (Dkt. No. 50-1) (emphasis added): 5 Disclaimers. To the fullest extent permitted by law, google, on 6 behalf of itself and its partners and affiliates, disclaims all warranties, whether implied, statutory or otherwise, including for 7 non-infringement, satisfactory quality, merchantability and fitness for any purpose, as well as any warranties arising out of any course 8 of dealing or usage of trade. To the fullest extent permitted by law, the programs and Google and partner properties are provided “as 9 is,” “as available” and “with all faults,” and customer uses them at its own risk. Google, its affiliates, and its partners do not make 10 any guarantee in connection with the programs or program results. Google makes no promise to inform customer of defects or errors. 11 The disclaimer reflected what Google sees as the fundamental unpredictability inherent 12 in online advertising, where many different factors and players have a role in determining an 13 ad’s success. On Google’s platform, advertisers, not Google, make the ultimate decisions 14 about campaigns, further necessitating this disclaimer. Dreamstime does not directly challenge 15 the enforceability of Google’s disclaimer. During oral argument, Dreamstime even noted 16 “[w]e take Google’s point. There’s a lot out of their control” (Hearing Tr. at 28). 17 Instead, Dreamstime focuses on two primary points. First, Dreamstime argues that the 18 advertising misrepresentations it alleges, and that form the basis for its breach of the implied 19 covenant of good faith and fair dealing claim, were not guarantees of a specific result. Yet 20 Dreamstime’s own CEO swears that Dreamstime’s claims are in fact about “dozens [of] 21 specific ad campaigns that did not perform as represented by Google and for which the 22 [average automated bid amounts] were so high that Dreamstime seeks a full refund in the form 23 of restitution” (Decl. Enache ¶ 15). Implying a promise or term into the agreement to cover 24 Dreamstime’s claims would be directly at odds with the express disclaimer in the advertising 25 agreement. 26 Second, as it did in connection with the search-related claims, Dreamstime calls attention 27 to a line of decisions holding that where a discretionary provision gives one party absolute and 1 unfettered discretion to perform or not, the implied covenant must be read into the term. See, 2 e.g., Third Story Music, Inc. v. Waits, 41 Cal. App. 4th 798, 805 (1995). These decisions are 3 inapplicable. The no-guarantee provision provides neither party with any discretion to perform 4 its obligations under the contract. 5 CONCLUSION 6 For the reasons stated herein, summary judgment in favor of Google on all of 7 Dreamstime’s remaining claims is GRANTED. The clerk shall please CLOSE THE file. 8 9 IT ISSO ORDERED. 10 11 Dated: July 3, 2020. 12 WIL ALSU 13 UNITED STATES DISTRICT JUDGE
il
Z 18 19 20 21 22 23 24 25 26 27 28