Rodriguez v. Google LLC

Background

Anibal Rodriguez, Sal Cataldo, Julian Santiago, and Susan Lynn Harvey brought a proposed class action against Google LLC. They alleged that Google collected, saved, and used data from third-party mobile apps through Google’s Firebase and Google Mobile Ads software development kits, even when users had turned off Google’s Web & App Activity (WAA) or supplemental Web & App Activity (sWAA) settings.

The complaint asserted three claims: invasion of privacy, intrusion upon seclusion, and violation of California’s Comprehensive Computer Data Access and Fraud Act (CDAFA). The plaintiffs sought certification of two classes covering individuals from July 1, 2016, through the present who had turned off their WAA and/or sWAA settings, whose activity on non-Google mobile apps was transmitted to Google through the specified software development kits, and whose devices were either Android or non-Android devices.

The plaintiffs’ damages expert, Michael J. Lasinski, offered models for class-wide damages. His models estimated unjust-enrichment damages of approximately $558.8 million under one scenario and $664.3 million under another, and actual damages of approximately $486.05 million based partly on a $3-per-device payment Google made to participants in its Screenwise consumer research panel. Google moved to exclude Lasinski’s opinions under Federal Rule of Evidence 702 and the standards for reliable expert testimony.

Class certification

The court held that both proposed classes satisfied Rule 23(a)’s requirements of numerosity, commonality, typicality, and adequacy. Each proposed class contained millions of users. The court found common questions about Google’s representations concerning the WAA and sWAA settings and the operation of Google’s software tools. The named plaintiffs’ claims were typical because Rodriguez, Cataldo, and Harvey used Android devices, while Santiago used an Apple device. The court also found no conflicts between the plaintiffs, their counsel, and the proposed class members.

The court further held that common questions predominated under Rule 23(b)(3) and that a class action was the superior method of resolving the dispute. It concluded that the objective questions of whether users had a reasonable expectation of privacy and whether Google’s alleged intrusion was highly offensive could be addressed with common evidence, including Google’s uniform disclosures and users’ common decision to turn off WAA or sWAA. The court rejected Google’s argument that continued use of Google services necessarily showed consent.

The court also found that the CDAFA issues could be addressed class-wide. It stated that California law recognizes disgorgement of profits resulting from unjust enrichment and that plaintiffs could use common evidence concerning Google’s data-collection and record-keeping practices. Individual issues concerning consent, third-party app disclosures, device settings, and damages did not outweigh the common questions.

The court separately certified a Rule 23(b)(2) class for injunctive relief. The requested relief included stopping Google from collecting, storing, and using WAA-off and sWAA-off app activity data; deleting already collected data and products, services, or algorithms built from it; and appointing an independent party to verify compliance. The court found that the plaintiffs had identified a common policy and the general contours of class-wide injunctive relief.

Daubert motion

Google asked the court to exclude Lasinski’s damages opinions as irrelevant or unreliable. The court denied that motion. It held that disgorgement—the recovery of profits allegedly retained through unjust enrichment—was potentially available at least for the CDAFA claim and was not categorically unavailable for the privacy claims. The court found that Lasinski’s models were sufficiently connected to the plaintiffs’ theory that Google was unjustly enriched by collecting their data.

The court rejected Google’s arguments that Lasinski’s models depended on fictitious scenarios, were unreliable because a similar model had been used in another case, or were improperly “cherry-picked.” The court found that Lasinski had explained his use of the $3 Screenwise payment and had considered other programs that paid users for their data. It also held that the model did not need to account for every individual variation because it used an objective market-based measure tied to Google’s own payments.

Disposition

The court granted the plaintiffs’ motion for class certification and denied Google’s motion to exclude Lasinski’s opinion. It certified the two specified classes: one for qualifying Android-device users and one for qualifying users of non-Android devices. The parties were ordered to appear for a case-management conference on February 1, 2024, and to submit a joint case-management statement by January 25, 2024. Judge Seeborg’s order addressed class treatment and expert admissibility, not the ultimate merits of the plaintiffs’ claims.