`
`
`
`
`
`
`
`
`
`UNITED STATES DISTRICT COURT
`NORTHERN DISTRICT OF CALIFORNIA
`
`CHASOM BROWN, ET AL.
`Plaintiffs,
`
`v.
`
`GOOGLE, LLC,
` Defendant.
`
`CASE NO. 20-cv-3664-YGR
`
`ORDER GRANTING IN PART MOTION FOR
`CLASS CERTIFICATION; GRANTING IN
`PART DAUBERT MOTIONS; AND DENYING
`MOTION TO STRIKE GOOGLE’S NON-
`RETAINED EXPERTS
`Re: Dkt. Nos. 609, 662, 663, 664, 703, 705
`Plaintiffs Chasom Brown, William Byatt, Jeremy Davis, Christopher Castillo, and
`Monique Trujillo bring this action against defendant Google, LLC, alleging seven counts based on
`Google’s alleged data collection practices: (1) violation of the Federal Wiretap Act, 18 U.S.C. §
`2510, et. seq., also known as the Electronic Communications Privacy Act (“ECPA”); (2) violation
`of the California Invasion of Privacy Act (“CIPA”), Cal. Penal Code §§ 631 and 632; (3) violation
`of the Comprehensive Computer Data Access and Fraud Act (“CDAFA”), Cal. Penal Code § 502
`et. seq.; (4) invasion of privacy; (5) intrusion upon seclusion; (6) breach of contract; (7) violation
`of California’s Unfair Competition Law (“UCL”).
`Pending before the Court are plaintiffs’ Motion for Class Certification, the parties’
`corresponding Daubert motions, plaintiffs’ Motion to Strike Google’s Non-Retained Experts, and
`several administrative motions to seal.1 Having carefully considered the parties’ briefing, the
`admissible evidence, the record in this case, and upon further consideration after oral argument
`which occurred on October 11, 2022, plaintiffs’ motion for class certification is GRANTED IN
`
`
`
`
`
`
`1 See Dkt. Nos. 609, 662, 663, 664, 703 and 705; see also Daubert v. Merrell Dow
`Pharms., Inc., 509 U.S. 579 (1993).
`
`As to the administrative motions to seal, these motions are DENIED to the extent the
`information is referenced and included in this Order. The specific motions will be addressed by
`separate court order.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`Northern District of California
`United States District Court
`
`
`
`Case 4:20-cv-03664-YGR Document 803 Filed 12/12/22 Page 2 of 34
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`PART. Many of Google’s arguments hinge on the general proposition that Google’s customer base
`is too big for class treatment. The notion that Google is too big to be held accountable does not
`persuade. The Court finds certification for injunctive relief only appropriate. Moreover, the
`Daubert motions are GRANTED IN PART and plaintiffs’ Motion to Strike is DENIED.
`BACKGROUND
`I.
`The Court incorporates the background provided in Judge Koh’s 31-page order denying
`Google’s motion to dismiss. (Dkt. No. 363, at 1-8.) In sum, plaintiffs allege that Google
`surreptitiously intercepts and collects users’ data even while users are in a private browsing mode.
`(Dkt. 395-2, Third Amended Complaint, (“TAC”) at ¶ 1.) The at-issue data includes: (i) “[t]he
`‘GET request’ sent from the user’s computer to the website”; (ii) “[t]he IP address of the user’s
`connection to the internet”; (iii) “[i]nformation identifying the browser software that the user is
`using, including any ‘fingerprint’ data”; (iv) “[a]ny ‘user-ID’ issued by the website to the user, if
`available”; (v) “[g]eolocation of the user, if available”; and (vi) “[i]nformation contained in
`‘Google cookies,’ which were saved by the user’s web browser on the user’s device at any time
`prior.” (TAC at ¶ 63(a)-(f).)
`LEGAL STANDARDS
`II.
`
`Daubert Motions
`A.
`Federal Rule of Evidence 702 permits opinion testimony by an expert as long as the
`witness is qualified and based upon that qualification, the witness’s opinion is relevant and
`reliable. An expert witness may be qualified by “knowledge, skill, experience, training, or
`education” as to the subject matter of the opinion. Fed. R. Evid. 702. The proponent of expert
`testimony has the burden of proving admissibility in accordance with Rule 702. Id., Advisory
`Committee Notes (2000 amendments). At the class certification stage, “the relevant inquiry is a
`tailored Daubert analysis which scrutinizes the reliability of the expert testimony in light of the
`criteria for class certification and the current state of the evidence.” Rai v. Santa Clara Valley
`Transportation Auth., 308 F.R.D. 245, 264 (N.D. Cal. 2015); Grodzitsky v. Am. Honda Motor Co.,
`957 F.3d 979, 985–86 (9th Cir. 2020). For scientific opinions, they must be based on scientifically
`valid principles. Daubert, 509 U.S. at 589. Experts assist the fact finder in their own evaluation of
`2
`
`Northern District of California
`United States District Court
`
`
`
`Case 4:20-cv-03664-YGR Document 803 Filed 12/12/22 Page 3 of 34
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`the evidence by providing the fact finder with opinions based upon verifiable, scientific, or other
`objective analysis. Id. at 589–90.
`Class Certification
`B.
`A class action is “an exception to the usual rule that litigation is conducted by and on
`behalf of the individual named parties only.” Comcast Corp. v. Behrend, 569 U.S. 27, 33 (2013).
`“Before certifying a class, the trial court must conduct a rigorous analysis to determine whether
`the party seeking certification has met the prerequisites of Rule 23.” Mazza v. Am. Honda Motor
`Co., 666 F.3d 581, 588 (9th Cir. 2012) (internal quotation marks omitted). The rigorous analysis
`that a court must conduct requires “judging the persuasiveness of the evidence presented” for and
`against certification and “resolv[ing] any factual disputes necessary to determine whether” the
`requirements of Rule 23 have been satisfied. Ellis v. Costco Wholesale Corp., 657 F.3d 970, 982–
`83 (9th Cir. 2011). A “district court must consider the merits if they overlap with the Rule 23(a)
`requirements.” Id. (citations omitted).
`The party moving for certification first must show that the four requirements of Rule 23(a)
`are met. Specifically, Rule 23(a) requires a showing that: (1) the class is so numerous that joinder
`of all members is impracticable; (2) common questions of law or fact as to the class exist; (3) the
`claims or defenses of the representative parties are typical of those of the class; and (4) the
`representative parties will fairly and adequately protect the interests of the class. Fed. R. Civ. P.
`23(a). The moving party must then show that the class can be certified based on at least one of the
`grounds in Rule 23(b). See Fed. R. Civ. P. 23(b). Relevant here, certification under Rule 23(b)(3)
`is appropriate only if “the questions of law or fact common to class members predominate over
`any questions affecting only individual members” and “a class action is superior to other available
`methods for fairly and efficiently adjudicating the controversy.” Fed. R. Civ. P. 23(b)(3). Rule
`23(b)(2) permits certification of a class when “the party opposing the class has acted or refused to
`act on grounds that apply generally to the class, so that final injunctive relief or corresponding
`declaratory relief is appropriate respecting the class as a whole.” Fed. R. Civ. P. 23(b)(2).
`///
`///
`
`3
`
`Northern District of California
`United States District Court
`
`
`
`Case 4:20-cv-03664-YGR Document 803 Filed 12/12/22 Page 4 of 34
`
`
`
`III. DISCUSSION
`
`Daubert Motions
`A.
`The Court addresses the parties’ various Daubert motions as they inform the Court’s
`analysis of the plaintiffs’ motion for class certification. Four such motions are pending: (1)
`Google’s Motion to Exclude Opinions of Plaintiffs’ Damages Expert Michael J. Lasinski (Dkt.
`No. 661-3); (2) Google’s Motion to Exclude Opinions of Plaintiffs’ Expert David Nelson (Dkt.
`No. 663); (3) Google’s Motion to Exclude Opinions of Plaintiffs’ Expert Bruce Schneier (Dkt. No.
`661-4); and (4) Plaintiffs’ Motion to Exclude Portions of Rebuttal Expert Report of Konstantinos
`Psounis (Dkt. No. 702-1). The Court addresses each motion in turn.
`
`1.
`Daubert as to Michael J. Lasinski
`Google moves to exclude the testimony of plaintiffs’ damages expert, Michael Lasinski,
`under Federal Rule of Evidence 702 and Daubert. (Dkt. No. 661-3.) Lasinski is a Senior
`Managing Director at Ankura Consulting Group (“Ankura”) and head of the Intellectual Property
`Group. (Dkt. No. 608-9, Lasinski Report ¶ 2.) He has twenty-seven years of experience assisting
`clients in understanding and evaluating the financial aspects of intellectual property. (Id.) Lasinski
`received his Bachelor of Science degree in Electrical Engineering and a Master of Business
`Administration from the University of Michigan. (Id. ¶ 6.) As assigned, he assessed the feasibility
`of identifying and quantifying various measures of monetary relief tied to plaintiffs’ claims,
`including unjust enrichment, actual damages (restitution), and statutory damages. (Id. ¶ 12.) To do
`so, he relied on his review of discovery produced by Google, deposition testimony, publicly
`available materials, deposition testimony of Google personnel, plaintiffs’ experts’ reports, as well
`as other materials listed in Appendix B of his report. (Id. ¶ 17.)
`Lasinski offers an opinion on a methodology for determining three types of damages: (a)
`unjust enrichment, (b) restitution damages, and (c) statutory damages, and asserts eight opinions,
`summarized as follows: (1) discovery in this case can be used to quantify relief on a class-wide
`basis; (2) Google’s ChromeGuard analysis provides a reliable basis for quantifying certain relief;
`(3) Google’s internal analyses of the financial impact to Google because of third-party cookie
`blocking can be adjusted to reliably quantify Google’s unjust enrichment; (4) calculation of unjust
`4
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`Northern District of California
`United States District Court
`
`
`
`Case 4:20-cv-03664-YGR Document 803 Filed 12/12/22 Page 5 of 34
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`enrichment can be determined under a range of potential liability scenarios; (5) restitution
`damages can be determined as a function of the payments necessary to incentivize an individual to
`knowingly relinquish the choice to keep certain browsing private and allow a company to track all
`online activity; (6) an appropriate damages rate can be applied in calculating statutory damages;
`(7) these statutory rates can be readily updated to cover subsequent periods through the date of
`trial; and (8) that his analyses can be readily used as common proof in part because they can be
`adjusted to calculate and assess unjust enrichment, actual damages, and statutory damages for
`different time periods. (Id. ¶ 1.)
`The Court starts with a high-level summary of the disputed parts of Lasinski’s models
`before addressing the parties’ arguments.
`Overview of the Methods
`a.
`
`Unjust Enrichment Model
`i.
`In opining on the methodology for determining classwide unjust enrichment damages,
`Lasinski looks to Google’s internal analyses that describe the financial impact to Google of
`blocking third-party cookies by default in Chrome Incognito mode (“ChromeGuard”). (Id. ¶ 52.)
`He segments his analyses by product area (Display Ads, YouTube ads, and Search Ads), private
`browsing mode (incognito mode or other browsing modes), revenue source (personalization or
`conversion tracking), and the scope of conversion tracking (conversion tracking from traffic with
`third-party cookies or conversion tracking from all traffic). (Id. ¶ 60.) In doing so, he calculates
`unjust enrichment damages from Google’s U.S. revenues from Display Ads, Search Ads, and
`YouTube Ads in three scenarios. (Id. ¶¶ 133-36.)
`In Scenario One, Lasinski arrives at the unjust enrichment amount by calculating: (a) all of
`Google’s U.S. Display Ads shown to users in private browsing mode, (b) U.S. search revenue
`attributable to conversion from all private browsing traffic, and (c) U.S. YouTube Ads revenue
`attributable to personalization from third-party cookies and conversion from all private browsing
`traffic. (Id. ¶¶ 133-35.) By using this approach, Lasinski calculates a damages amount of $3.87
`billion. (Id. ¶ 136.)
`///
`
`5
`
`Northern District of California
`United States District Court
`
`
`
`Case 4:20-cv-03664-YGR Document 803 Filed 12/12/22 Page 6 of 34
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`In Scenario Two, he arrives at the unjust enrichment amount by calculating Google’s
`Display, Search, and YouTube revenue that is attributable to personalization from third-party
`cookies and conversion from all private browsing traffic. Here, he calculates a damages amount of
`$3.61 billion. (Id. ¶¶ 133-36.)
`In Scenario Three, he arrives at the unjust enrichment amount by calculating Google’s
`Display, Search, and YouTube that is attributable to personalization and conversion from third-
`party cookies. With this model, he calculates a damages amount of $567.4 million. (Id.)
`Lasinski explains that such segmentation is intended to assist the trier of fact in
`determining Google’s unjust enrichment under a range of liability scenarios. (Id. ¶ 133.)
`Restitution Model
`ii.
`Next, Lasinski opines on a methodology for calculating classwide restitution damages.
`Here, he uses the payment amount necessary to incentivize individuals to knowingly relinquish
`their online privacy as a base rate. He then takes that rate and multiplies it by the number of
`unique monthly private browsing instances (“UMPBI”). (Id. ¶ 184.)2 In doing so, Lasinski
`calculates a restitution damages amount of approximately $9.1 billion. (Id. at Fig. 72.)
`In arriving at the base rate, Lasinski analyzed payments that Google made or considered
`making to users for their data, users’ willingness to pay to prevent data collection, and research
`organizations’ willingness to pay for data collection. (Id. ¶ 140.) Lasinski explains that the most
`reliable input comes from Google itself. (Id. ¶ 165.) Since 2012, Google has used a market
`research company called Ipsos to collect information on how users use the Internet. (Id. ¶ 143.)
`Through Ipsos, Google conducts what are known as Screenwise Panel studies. (Id.) These studies
`allow selected participants to get paid if they voluntarily link their devices and allow Google to
`collect certain information about the users’ online activities. (Id.) Some examples of the data
`collected through the Screenwise Panel studies include: the content and advertising shown on
`devices, interactions with that content and advertising, webpages, information you type into your
`devices, cookies, and device information. (Id. ¶ 144, Fig. 57.)
`
`
`2 “[A] single UMPBI represents one or more pageloads in Incognito mode or an Other
`Private Browsing Mode on a single device during a one-month period.” (Lasinski Report ¶ 139.)
`6
`
`Northern District of California
`United States District Court
`
`
`
`Case 4:20-cv-03664-YGR Document 803 Filed 12/12/22 Page 7 of 34
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`As part of the Screenwise Panel, participants are paid a baseline minimum of $3 per month
`per device. This base rate does not decrease based on one’s browsing activity, but it can increase
`based on other factors. (Id. ¶¶ 149-50.) To calculate damages, Lasinski uses this $3 rate that
`Google pays survey participants as an input. (Id. ¶ 183.) He then multiplies that by the number of
`UMPBI during the class period for both classes. (Id. ¶ 184.)
`Statutory Damages Model
`iii.
`Third, in opining on a methodology for calculating statutory damages on a classwide basis,
`Lasinski used data and internal documents produced by Google which quantified incognito traffic.
`This quantified traffic was then used to calculate the four bases that Lasinski offers as common
`proof of calculating statutory damages: (1) the total number of individual pageloads in the relevant
`browsers; (2) the total UMPBI with one UMPBI referring to a specific browse that used a private
`browsing mode at least once a month; (3) calculation of UMPBI based on the peak UMPBI for
`one month; and (4) the total number of class members for each of the classes. (Id. ¶¶ 185-195.)
`Lasinski opines that all four bases would allow the Court to calculate statutory damages and can
`be used as common proof because they can be adjusted to account for different time periods and
`subclasses. (Id. ¶ 198.)
`
`Damages Allocation Method
`iv.
`Finally, Lasinski opines on two methods to allocate individual damages. The first method
`includes taking the aggregate damages number and dividing it by the total number of UMPBI
`during the class period. Awards would be distributed to class members based on the number of
`UMPBI attributable to each class member.3 The second method includes calculating individual
`damages on a per capita basis based on the number of class members. (Id. ¶ 197.)
`
`Overview of Parties’ Contentions
`b.
`Google moves to exclude Lasinski’s opinions and methodologies for computing damages
`on the grounds that: (1) Lasinski fails to account for uninjured users who consented to the at-issue
`
`
`3 Based on Lasinski’s calculations, the Court calculates an average UMPBI of about .725
`per class member. (See id. at Figs. 74 and 75.)
`
`
`7
`
`Northern District of California
`United States District Court
`
`
`
`Case 4:20-cv-03664-YGR Document 803 Filed 12/12/22 Page 8 of 34
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`data collection; (2) his primary input for calculating class-wide restitution damages is speculative
`and unreliable; (3) his restitution opinions fail to account for variability in user benefit from
`personalization and valuation of privacy; (4) his failure to deduct costs from the unjust enrichment
`analysis is contrary to the law and has no basis in the facts; (5) all of his unjust enrichment
`scenarios are flawed and unreliable; (6) his proposed methods for apportioning damages ignore
`individual differences and are not efficient or feasible for distributing damages; and (7) his
`opinions on calculating statutory damages are unreliable.
`Plaintiffs oppose arguing: (1) Lasinski’s damages calculations are based on assumptions
`consistent with their theory of the case; (2) his actual damages calculations are based on the same
`key input Google used in the Screenwise Panel study; (3) Google documents and testimony
`confirm that incremental costs do not exist for private browsing traffic; (4) Lasinski’s method for
`calculating statutory damages are sound and simply lead to penalties Google does not like; and (5)
`Lasinski is not required to adopt Google’s preferred method of apportioning damages.
`Assumption of Harm to Every Class Member4
`i.
`First, Google argues that all of Lasinski’s opinions should be excluded because he fails to
`account for uninjured class members.
`A plaintiff seeking certification under Rule 23(b)(3) must show that damages are capable
`of measurement on a class-wide basis, and such calculations “need not be exact.” Comcast, 569
`U.S. at 35. Under Comcast, a plaintiff must show that its proposed damages model is consistent
`with its theory of liability in the case. Id.
`As an initial matter, Google’s arguments “reflect[] a merits dispute about the scope of [its]
`liability, and is not appropriate for resolution at the class certification stage of this proceeding.”
`See Ruiz Torres v. Mercer Canyons Inc., 835 F.3d 1125, 1137 (9th Cir. 2016). Relatedly,
`Lasinski’s methodology is consistent with plaintiffs’ class definition and theory of liability.
`Google’s disagreement with Lasinski’s assumption is no reason to exclude his opinions. Perez v.
`
`
`4 The Court notes that both parties used part of their Daubert briefing to argue the merits
`of their consent arguments. Counsel is aware of the legal standard under Daubert and should know
`that such argument is wholly inappropriate. The parties are warned that future conduct will result
`in the briefing being stricken and/or sanctions.
`
`8
`
`Northern District of California
`United States District Court
`
`
`
`Case 4:20-cv-03664-YGR Document 803 Filed 12/12/22 Page 9 of 34
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`Rash Curtis & Assocs., 2019 WL 1491694, at *3 (N.D. Cal. Apr. 4, 2019) (“any arguments that
`[an expert] based his opinions on an improper assumption go to the weight, not the admissibility
`of the testimony.”)
`Accordingly, the motion is DENIED on this ground.
`$3 Input for Calculating Restitution Damages
`ii.
`Second, Google argues that Lasinski’s restitution opinion should be excluded because the
`$3 input that he uses is speculative and unreliable. Google does not persuade. The $3 rate is
`derived from looking at what Google actually pays Screenwise participants for agreeing to allow
`Google to collect their browsing data. In arriving at this number, Lasinski acknowledged that the
`full scope of the data collected in the Screenwise studies differs in some respects from the at-issue
`data and accounted for that in using the baseline of $3 rather than some higher amount that some
`participants receive through the Screenwise program. (Lasinski Report ¶¶ 147-50.) Lasinski opines
`that this rate is conservative because that is the rate willing people would accept to give up their
`data and that unwilling users, hence users in private browsing mode, will likely require more
`incentive to give up their data. (Dkt. No. 661-3, Lasinski Dep. 101:13-102:22.) The Court finds
`such considerations and analysis logical and reliable. The question is one of weight, not
`admissibility.
`Accordingly, the motion is DENIED on this ground.
`Variability in User Benefit and Valuation of Privacy
`iii.
`Third, Google avers that Lasinski’s restitution opinions should be excluded as unreliable
`because he ignores: (i) the strong evidence of variability in the benefits that users may have
`received from ad personalization and (ii) the value users place on their private browsing data and
`their privacy. In response, plaintiffs argue that under their theory of restitution such subjective
`valuations are not necessary.
`Where there is “enrichment from the receipt of nonreturnable benefits,” restitution can be
`measured by either: (a) the value of benefit to the defendant; (b) the cost to plaintiff of conferring
`such benefit; (c) the market value of the benefit; or (d) the price the defendant has expressed a
`willingness to pay, if the defendant’s asset may be treated valid on the question of price.
`9
`
`Northern District of California
`United States District Court
`
`
`
`Case 4:20-cv-03664-YGR Document 803 Filed 12/12/22 Page 10 of 34
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`Restatement (Third) of Restitution and Unjust Enrichment § 49.
`The Court finds that under any of these formulations, this is precisely what Lasinski’s
`method measures. He measures the value that Google actually pays to users who voluntarily give
`up their right to privately browse the Internet. Lasinski’s method does not turn on the subjective
`feelings and beliefs of class members. Rather, he employs an objective standard focused on paying
`users to relinquish their right to browse privately.
`Indeed, Google employs a similarly objective standard in compensating Screenwise
`participants. The amount Google pays participants does not fluctuate depending on the
`participants’ subjective beliefs. Rather, Google sets a base rate and then compensates based on
`objective measures such as how many devices are registered and what kinds of devices are used.
`These same objective criteria that Google employs in its Screenwise panel study is what Lasinski
`uses as a base for his calculation of restitution damages. Therefore, Google’s disagreement with
`the restitution method applied and the outcome of that method does not make Lasinski’s method
`unreliable. Google’s concerns go to weight, not admissibility.
`Accordingly, the motion on this ground is DENIED.
`Unjust Enrichment Model and Costs
`iv.
`Fourth, Google seeks to exclude Lasinski’s unjust enrichment model on the grounds that it
`looks at Google’s revenue and does not take into account Google’s costs. Plaintiffs argue that
`Lasinski did not deduct costs from his calculation because there are no incremental costs to
`Google.
`“Disgorgement is a remedy in which a court orders a wrongdoer to turn over all profits
`obtained by violating the law.” Consumer Fin. Prot. Bureau v. Gordon, 819 F.3d 1179, 1195 (9th
`Cir. 2016). A party that is wronged and seeks unjust enrichment “may present evidence of the total
`or gross amount of the benefit, or a reasonable approximation thereof, and then [the defendant]
`may present evidence of costs, expenses, and other deductions to show the accrual or net benefit
`the defendant received.” Meister v. Mensinger, 230 Cal. App. 4th 381, 399 (2014). Importantly,
`“the residual risks of uncertainty in calculating net profit is assigned to the wrongdoer.” Id.
`
`
`10
`
`Northern District of California
`United States District Court
`
`
`
`Case 4:20-cv-03664-YGR Document 803 Filed 12/12/22 Page 11 of 34
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`In proffering his unjust enrichment model, Lasinski did not account for costs in calculating
`Google’s profits. He testified that he did not include costs because there was no proof of
`incremental costs to Google included in the record. Lasinski relied on plaintiffs’ other expert,
`Hochman, who explained to Lasinski that there are no costs to Google for maintaining the
`advertising infrastructure that Google has already built. (Lasinski Dep. 165:20-166:8.) Hochman
`also explained to Lasinski that many of the costs associated with the at-issue data collection is
`passed down to users by way of Google’s tracking beacons. (Id. at 163:15-23.) In response,
`Google argues that Strombom’s Report identifies many components of Google’s costs including
`those associated with generating the at-issue data.
`Google misstates the evidence. Strombom testified during his deposition that Google does
`not account for costs associated with private data, and that he had no evidence of actual costs
`saved because of Google launching ChromeGuard. (Dkt. No. 698-8, Strombom Dep. 93:1-13,
`111:21-119:10, 119:11-17.) Google’s 30(b)(6) witness, Sonal Singhal, did not contradict him. Ms.
`Singhal was designated to talk about an Ads Impact document that looked at the impact of
`ChromeGuard on Google’s finances. She testified that, while questions about costs are outside of
`her expertise, she “[didn’t] think that there is any cost that would vary.” (Dkt. No. 698-7, Singhal
`Dep. 23:25-24:9.) Plaintiffs may rely on those admissions.
`Thus, the only evidence of any potential costs is in the form of a declaration by George
`Levitte, a Google employee, who testifies generally about form-sharing agreements with
`publishers. (Dkt. No. 659-8, Levitt Decl., ¶¶ 11-13.)5 Levitte explains that there are costs
`
`
`5 Pending before the Court is plaintiffs’ Motion to Strike Google’s Non-Retained Expert
`Declarations for Glenn Berntson, Steve Ganem, George Levitte, and Jonathan McPhie. (Dkt. No.
`704-2.) Having considered the parties’ briefing, and the record in this case, the Court DENIES the
`motion because plaintiffs have not shown harm. Plaintiffs admit that Google disclosed three of the
`four witnesses (Ganem, Levitte, and Bernston) in its initial disclosures before the close of
`discovery, and plaintiffs were able to depose these witnesses. While Google did not disclose
`McPhie in its amended disclosures, Google explains that it did disclose Gregory Fair, who was
`McPhie’s former supervisor. Fair has since left the company and Google substituted McPhie as a
`declarant in lieu of Fair. (Dkt. No. 746-3, Opposition to Motion to Strike, at 5.) McPhie’s
`declaration covers the topics Google disclosed for Fair in its amended disclosures. (See Dkt. No.
`705-6.) Accordingly, the Court finds that any harm to plaintiffs is minimal, and the motion is
`DENIED on that ground. However, the Court will allow plaintiffs to depose the witnesses on the
`topics contained in their declarations to the extent relevant. The parties shall meet and confer
`regarding a schedule for depositions. A joint proposed schedule shall be filed within seven (7)
`11
`
`Northern District of California
`United States District Court
`
`
`
`Case 4:20-cv-03664-YGR Document 803 Filed 12/12/22 Page 12 of 34
`
`
`
`associated with Google generating revenue but that the profits (which would reflect costs) cannot
`be calculated for “individuals who browsed while in Incognito mode, or those individuals who use
`any other browser’s private browsing mode.” (Id. ¶ 20.) While possible, this conflicting evidence
`reflects on the weight of the testimony, not admissibility.
`Accordingly, the motion is DENIED on this ground.
`Method for Apportionment
`v.
`Next, Google’s attempt to exclude Lasinski’s method for apportioning the aggregate
`damages does not persuade. With respect to Lasinski’s method of apportionment, Google does not
`have standing to complain. The Ninth Circuit counsels that when “the only question is how to
`distribute the damages, the interests affected are not the defendant’s but rather those of the silent
`class members.” Six (6) Mexican Workers v. Az. Citrus Growers, 904 F.2d 1301, 1307 (9th Cir.
`1990); see also Story Parchment Co. v. Paterson Parchment Paper Co., 282 U.S. 555, 563 (1931)
`(“[w]here the tort itself is of such a nature as to preclude the ascertainment of the amount of
`damages with certainty, it would be a perversion of fundamental principles of justice to deny all
`relief to the injured person, and thereby relieve the wrongdoer from making any amend for his
`acts.”) Further, the argument is premature.
`Accordingly, the motion is DENIED on this ground.
`
`Statutory damages
`vi.
`Lastly, Google moves to exclude Lasinski’s methods for calculating statutory damages on
`the basis that the proposed methods fail: (1) to account for instances where Google would not have
`received the at-issue data—like when a user starts a private browsing session but closes it before
`navigating to a third party website; (2) to provide a way to calculate statutory claims for the
`California CIPA class; and (3) to opine on a method for apportioning statutory damages.
`Google does not persuade. As to the first argument, Google offers only a hypothetical
`situation with no proof or data to support the idea that Lasinski should have accounted for
`instances where Google did not receive data. That Google can think of different ways in which
`
`
`days from the date of this order.
`
`12
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`Northern District of California
`United States District Court
`
`
`
`Case 4:20-cv-03664-YGR Document 803 Filed 12/12/22 Page 13 of 34
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`
`Lasinski ought to have created his model does not make it unreliable and is thus no reason to
`exclude. Google’s second argument is merely a rehash of its arguments about Lasinski’s
`apportionment method. As previously explained, Google does not have standing to raise such
`arguments. Moreover, none of Google’s concerns about apportionment render Lasinski’s methods
`and opinions unreliable. As to the third argument regarding the California CIPA class, Lasinski
`explained both in his report, and during the hearing, that his model could be adapted to account for
`just California residents by extrapolating the percentage of California users based on publicly
`available information from the general United States population. (Lasinski Report ¶ 197; see also
`Dkt. No. 775, Hearing Transcript, at 36:24-39:9.) Plaintiffs also explained that this information
`can also be collected by using Geolocation data and users’ IP addresses. (Hearing Transcript, at
`36:24-39:9.) The Court finds the proffered explanations reasonable.
`Thu