Brown v. Google LLC
Brown v. Google LLC
Case 4:20-cv-03664-YGR (N.D. Cal. 2023)
April 4, 2023

van Keulen, Susan,  United States Magistrate Judge

Proportionality
Failure to Preserve
Scope of Preservation
Download PDF
To Cite List
Summary
Google's motion to exclude certain mapping and linking tables from the court's preservation directive was granted, as the court found that the burden of preserving these tables outweighed their potential benefit in the litigation. The court determined that the data in these tables could be derived from other preserved sources and that their preservation would be unduly burdensome.
Additional Decisions
CHASOM BROWN, et al., Plaintiffs,
v.
GOOGLE LLC, Defendant
PATRICK CALHOUN, et al., Plaintiffs,
v.
GOOGLE LLC, Defendant
Case No. 20-cv-03664-YGR (SVK), Case No. 20-cv-5146-YGR (SVK)
United States District Court, N.D. California
Filed April 04, 2023
van Keulen, Susan, United States Magistrate Judge

ORDER RE PRESERVATION OF CERTAIN MAPPING AND LINKING TABLES

I. INTRODUCTION

This Order addresses Google’s Motion for Relief Regarding Preservation. Brown Dkt. 782; Calhoun Dkt. 898 (the “Motion”). Specifically, Google requests that [REDACTED] tables be excluded from this Court’s directive that “all mapping and linking tables be preserved” on the grounds that the relevant data in these tables can either be derived from or is duplicative of data that is separately being preserved and that preservation of these specific tables is unduly burdensome. Id. Google’s request has been the subject of extensive briefing, two hearings, and Court-ordered meet and confer sessions among the Parties. The procedural history is more fully addressed in this Court’s Orders following the hearings. Brown Dkt. 830, 861; Calhoun Dkt. 960, 983. The Court has considered the Parties’ briefing, the video and transcription of meet and confer efforts and the presentation of counsel in oral argument. The Court determines that this matter is, at this stage, suitable for resolution without further oral argument. Civ. L.R. 7-1(b). For the reasons set forth below, Google’s motion is GRANTED.

II. OVERVIEW

At issue are [REDA CTED] tables, [REDACTED] tables and [REDACTED] tables. The [REDACTED]  tables are “linking” tables, named for a Google project known as [REDACTED]. These [REDACTED]tables contain links that are derived from other data sources. Brown Dkt. 781-3 at 10; Calhoun Dkt. 897-3 at 10. The [REDACTED] tables are “mapping” tables that contain mappings of identifiers relating to the Google Analytics product. Id. Google argues that it should not be obligated to preserve these [REDACTED] tables (hereinafter the “Subject Tables”) on two grounds: (1) The tables are not necessary to use the data that is being preserved; and (2) Preservation of these particular tables presents an extraordinary burden that was not previously presented to or considered by the Court. Brown Dkt. 781-3 at 5-6; Calhoun Dkt. 897-3 at 5-6. Plaintiffs in both cases raise numerous questions about the content of the Subject Tables and argue that without certainty as to the Subject Tables’ content, they should be preserved at Google’s expense. Brown Dkt. 800; Calhoun Dkt. 933-2. In the final round of supplemental briefing, post-hearing, Plaintiffs focus, among other issues, on whether failure to preserve the Subject Tables is a violation of earlier Court orders. Brown Dkt. 849; Calhoun Dkt. 974. Accordingly, the Court addresses the evidence and arguments in the briefing leading up to and including the hearing on February 14, 2023 (section III below) and then takes up the final round of supplemental briefing (section IV below).

The Court has previously addressed the complex challenges in managing the relevant data in these actions, challenges which arise out of the reality that Google’s operations capture a vast, if not precisely infinite, universe of data. Not all of that data is relevant to the claims in the Brown and Calhoun matters, and not all of the relevant data can be captured and stored. In light of these complexities, the Court appointed a Special Master to (1) manage issues of identification and production of relevant data; and (2) to develop a plan to preserve relevant data. Brown Dkt. 331; Calhoun Dkt. 377; Brown Dkt. 587; Calhoun Dkt. 766 (“Preservation Orders”). The Preservation Orders are largely based on sampling of data sources; they do not expressly address preservation of mapping and linking tables. Id. Plaintiffs raised the need for these tables in their objections to the Preservation Orders, citing the need for the tables to use the preserved data samples. Brown Dkt. 587 at 8; Calhoun Dkt. 766 at 8. At an August 4, 2022 hearing to clarify certain aspects of the Preservation Orders, the Court stated that “if you would need a mapping table to work with any of the data that is being exchanged in this case, that table has to be preserved.” Brown Dkt. 694 at 106:16-18; Calhoun Dkt. 841 at 106:16-18. The Court reiterated the scope of preservation of the tables at the hearing on January 10, 2023: If the table is needed to work with any of the data at issue, it has to be preserved. Brown Dkt. 828 at 8:2-8; Calhoun Dkt. 959 at 18. Pursuant to this Court’s Preservation Orders and Order re Clarification of Preservation Plan Orders (Brown Dkt. 657; Calhoun Dkt. 815 (“8/5/22 Order”)), Google is preserving [REDACTED] such tables. Brown Dkt. 781-3 at 3; Calhoun Dkt. 897-3 at 9. However, Google disputes its obligation to preserve the Subject Tables, which are not included in the [REDACTED] currently being preserved. Id.

III. ISSUES RAISED IN BRIEFING AND AT THE FEBRUARY 14, 2023 HEARING

A. The Subject Tables are not needed to use the Preserved Data

Google argues, with supporting evidence, that the data in the Subject Tables can either be derived from other preserved data sources or is duplicated in other preserved logs/sources. The Court finds this argument is well supported.

1. [REDACTED] Tables

In the first instance, the Motion is supported by the declaration of Daryl Seah, a Google Software Engineer with personal knowledge of the content and structure of the [REDACTED] tables at issue. Brown Dkt. 781-9; Calhoun Dkt. 897-11 (“Seah Decl.”) ¶ 1. Mr. Seah attests that [REDACTED] relevant [REDACTED] tables that contain a mapping or linking of user or device identifiers have been identified. Id. ¶ 4. Among these [REDACTED] tables are the [REDACTED]tables which comprise identifier linkages that are derived from other tables that are being separately preserved in accordance with the 8/5/22 Order. Id. ¶ 5. In light of this evidence, and in response to questions raised by Plaintiffs, at the January 10, 2023 hearing the Court ordered the Parties to meet and confer to provide Google an opportunity to demonstrate to Plaintiffs how data in the [REDACTED] tables can be derived from data in tables already being preserved. Brown Dkt. 830 at 2; Calhoun Dkt. 960 at 3. In other words, Google was to provide an actual demonstration of the facts attested to in the Seah declaration. The Court also cautioned the Parties that this demonstration was not intended as a doorway to re-open discovery; it was necessarily narrow in focus. Google was to “show both sides of the equation, which is here’s the data in the tables that we don’t think need to be preserved from the [REDACTED] tables, but here’s how and where you find in the material[.]” Brown Dkt. 828 at 30:2-5. The Court was clear that that Google would “not be required to provide the entire contents of the tables to Plaintiffs.” Brown Dkt. 830 at 2; Calhoun Dkt. 960 at 3.

In anticipation of the Court-ordered meet and confer, on January 26, 2023, Google sent Plaintiffs a letter that matched relevant data points, specifically linked Biscotti IDs, from one of the [REDACTED] tables to the source tables from which the Biscotti IDs were derived. Brown Dkt. 848-5 at 3-4; Calhoun Dkt. 973-5 at 3-4. Google further confirmed that the source tables from which the Biscotti IDs were derived were already being preserved. Id. at 5; see also Calhoun Dkt. 943-1 (“Trebicka Decl.”) Ex. 1 at -354, 355. At the meet and confer on January 28, 2023, Google walked the Parties through the identification of data from the exemplar [REDACTED] table to the source tables already being preserved. Meet & Confer Transcript, pp. 3-4.[1] Google walked through the evidence again at the February 14, 2023 hearing, for the benefit of the Court. Brown Dkt. 868 at 10-17 and 37-44; Calhoun Dkt. 985 at 10-17 and 37-44.

Plaintiffs’ argument for preservation of the Subject Tables generally and of the [REDACTED]  tables in particular is fairly summarized as an interest in knowing the full content of the tables and the algorithms behind the linkages. “Although Google suggests the mapping and linking values contained in these [REDACTED] tables are derived from other tables (and therefore duplicative), Google is silent as to what other information is contained in these [REDACTED] tables.” Brown Dkt. 800 at 7; see also Calhoun Dkt. 933-2 at 7; Meet & Confer Transcript at 17; Brown Dkt. 889-4 at 4; Calhoun Dkt. 995-4. Examples of the scope of both Brown and Calhoun inquiries include:

  • What other identifiers are stored in the tables?
  • What other data are stored in the tables and keyed to those identifiers?
  • What information is stored in the [REDACTED] tables that accounts for the petabytes of data Google seeks relief from preserving?
  • From what other sources do the [REDACTED] tables draw data? Id.[2]
In support of these inquiries, at the February 14 hearing the Brown Plaintiffs[3] raised an issue regarding the scope of potential liability that has been the subject of previous disputes, and remains in dispute, between the Parties in Brown. “Our point is, if you recall, Judge, way back, I think in July of 2021, we were arguing about signed-in versus signed-out, and there was a question on what happens when somebody signs in, not on Google, but on a third-party web property in incognito or in private browsing . . . .” Brown Dkt. 868 at 22:11-15. More particularly to the issue at hand, Plaintiffs argued that when a user is not signed into a Google account (hence a potential class member) but signs in on a third-party web property in private browsing mode, there may be a linkage to a user identification. Id. at 22:2-22. “There is no substitute for that [REDACTED] .”[4] Id. at 23:16. Plaintiffs further argue, “We have never been given any logs or sources as to how the [REDACTED] itself, is actually used.” Id. at 25:17.

In addition to their liability argument, Plaintiffs argued the unique [REDACTED] is needed for injunctive relief. “And you can appreciate, your Honor . . . that is certainly relevant to the injunctive relief.” Id. at 25:20. “If the Judge and jury find for the Plaintiffs, right, and believe that all sources that had used [REDACTED] including in signed-in, in Incognito Activity, right, process or products that have been developed that need to be deleted, right, there needs to be some, right, like we need to have a trail, right, to [REDACTED] and if Google is destroying that associating from the unique source, how are the parties going to discuss that?” Id. at 26:4-11.

Google responded to these arguments with its demonstration in the hearing illustrating that: (1) The linked GAIA IDs in the [REDACTED] tables are pulled from source tables that are already being preserved (id. at 13-15); and (2) For this linkage the [REDACTED] tables use a “[REDACTED]” that is selected from the identified GAIA IDs. There is no independent “[REDACTED] as argued by Plaintiffs. Id. at 10, 27. In the briefing, at the hearing and in the meet and confer process, Google has suggested, but not demonstrated, that Plaintiffs could construct custom pipelines from the preserved source tables to reconstruct the linkages.[5] Id. at 16; Brown Dkt. 848- 6 at 3; Calhoun Dkt. 973-6 at 2.

The Court distills Plaintiffs’ concerns raised in the briefing and at the hearing down to two issues: (1) That the [REDACTED] in the [REDACTED] Tables is unique and cannot be recreated; and (2) How Google has used the [REDACTED] in its operations. As to the first issue, Google has established to the Court’s satisfaction that the [REDACTED] tables select a from among the linked Biscotti identifiers. In reconstructing linked Biscotti identifiers from underlying data, Plaintiffs can similarly select a [REDACTED]. While it is true that the Plaintiffs’ reconstructed [REDACTED] will not be the same as Google’s [REDACTED] in the [REDACTED] table, it will show that linkages exist which Plaintiffs can then use for purposes of arguing liability and seeking injunctive relief.

As for the second issue, while preservation of the [REDACTED] tables came up late in the process of clarifying this Court’s Preservation Orders, the existence of [REDACTED] tables was not new information to the Plaintiffs. Brown Dkt. 889-3 at 8; Calhoun Dkt. 995-3 at 8. And while Plaintiffs now raise concerns about Google’s obligation to disclose and preserve [REDACTED] tables during the discovery process (see section IV below), the fact remains that discovery is now closed. To the extent the answer as to how Google uses the [REDACTED] from the [REDACTED] tables is unanswered, it will remain so. Compelling Google to preserve the [REDACTED]  tables, including the [REDACTED], will not answer Plaintiffs’ question as to how Google used those IDs.

a. Conclusion: [REDACTED] Tables

The Court takes no position on the liability issue outlined by the Brown Plaintiffs above but rather remains focused on the issue before it: whether preservation of the [REDACTED] tables at this stage of the litigation is proportional to the needs of the case. In sum, the Subject Tables were to be preserved pursuant to a narrow directive of this Court: if you would need a mapping table to work with any of the data that is being exchanged in this case, that table has to be preserved. Brown Dkt. 694 at 106:16-18; Calhoun Dkt. 841 at 106:16-18. Then, in response to the Motion, the Court instructed the Parties to meet and confer under the following guideline: [T]he Court anticipates that this meet and confer will require Google to provide to Plaintiffs information about the contents of the tables at issue sufficient to demonstrate that pertinent information in the tables is already being preserved or may be derived from preserved data. However, Google will not be required to provide the entire contents of the tables to Plaintiffs.” Brown Dkt. 830 (emphasis added). The Court recognizes that unanswered questions remain, but they are outweighed by Google’s evidence of the derived data and the burden of preservation of the [REDACTED] Tables as discussed below.

The Court is satisfied that in the Seah declaration, the subsequent demonstration and written explanations cited above, Google has met the burden assigned by the Court to demonstrate that the data being preserved in the [REDACTED] tables can be derived from data sources already being preserved in this action.

2. [REDACTED] Tables

Google’s motion is supported by the declaration of Srilakshmi Pothana (Brown Dkt. 781- 10; Calhoun Dkt. 897-12 (“Pothana Decl.”)) a Google Software Engineer with personal knowledge of the content and structure of the [REDACTED] tables at issue. Pothana Decl. ¶ 1. Ms. Pothana attests that [REDACTED] tables that “contain a mapping or linking of user or device identifiers that may be included in data received by Google Analytics when users visit Google Analytics customer websites” have been identified. Id. ¶ 3. The subject [REDACTED] Tables are [REDACTED] of these tables, and they are [REDACTED] the and [REDACTED] . Id. ¶ 5. These [REDACTED] tables contain mapping between Google Analytics User ID (UID) or client ID (CID) and Biscotti. Id.

Further, and critically, the mappings between the UID/CID and Biscotti contained in these [REDACTED] tables comes from data stored in Google [REDACTED] logs sources which are currently subject to preservation requirements in the Brown action. Id. ¶ 6. In other words, the mappings in these tables are duplicative of mappings in already preserved data.[6] Id.; see also Brown Dkt. 868 at 7:23-8:3; Calhoun Dkt. 985 at 7:23-8:3.

As with the [REDACTED] tables, Google detailed this information in the January 26, 2023 letter to Plaintiffs in advance of the Court-ordered meet and confer. Brown Dkt. 848-5 at 5-7; Calhoun Dkt. 973-5 at 4-6. Specifically, Google provided sampled mapping entries from the [REDACTED] mapping CID/UID to Biscotti ID. For the Brown matter, Google explained that for a given sampled entry in the [REDACTED] logs preserved under the Preservation Plan, any mapping of CID/UID to Biscotti will be self-contained within the sampled data. More particularly, the CID and UID values can be found in the “[REDACTED]“ field and the Biscotti values are in the “[REDACTED]” field. Brown Dkt. 848-5 at 7. For the Calhoun matter, Google explained that because these [REDACTED] Tables include no mappings to GAIA IDs, they are of no use in interpreting any of the preserved Data in the Calhoun action. Calhoun Dkt. 973-5 at 6.

As noted above, Plaintiffs generally question the larger content of the [REDACTED] tables, to which Google responded in written communications on January 28, January 30 and February 17, 2023. In response to questions specific to the Brown Plaintiffs, Google offered a further demonstration based upon Plaintiffs’ experts’ data. Unfortunately the Parties were able to agree as to the parameters of that demonstration. See section IV below.

a. Conclusion: [REDACTED] Tables

Nevertheless, the Court is satisfied that in the Pothana declaration, the subsequent demonstration and written explanations cited above, Google has met the burden assigned by the Court to demonstrate that the data being preserved in the [REDACTED] Tables comes from data stored in Google [REDACTED] logs sources which are currently subject to preservation requirements in the Brown action. Pothana Decl. ¶ 6. Further, the Calhoun Plaintiffs have not suggested that in fact this data is needed to address the data preserved in that action, but rather raise other issues which are addressed in section IV below.

B. The burden of preserving the Subject Tables outweighs their utility in these actions.

Google argues that preservation of the [REDACTED] Subject Tables is unduly burdensome both in terms of storage capacity and financial costs. Brown Dkt. 781-3 at 4-11; Calhoun Dkt. 897-3 at 17-19. Google Software Engineer Lawrence Greenfield first cites the overall shortage of storage space arising out of supply-chain issues in recent years. Brown Dkt. 781-4 ¶ 3; Calhoun Dkt. 897- 4 ¶ 3 (“Greenfield Decl.”). In this context of an overall shortage of storage capacity, Mr. Greenfield further avers to the practical implementation requirements of data storage: hardware, such as servers; electricity, required to safely maintain the hardware; and labor for employees to support and maintain the storage infrastructure. Id. ¶ 5.

To assign a dollar figure to these implementation requirements, Google proffers the rates utilized in its Google Cloud Storage (GCS) program. Id. ¶ 6. GCS allows businesses or individuals to store and retrieve data for a fee, and the pricing schedule is published at https://cloud.googl.coom/storage/pricing. Id. Google justifies its use of the GCS rates in its calculations on the basis that if it is using space to store preserved data, it is unable to offer that space to the public. Id. ¶ 7. The cost estimates set forth in the tables below utilize the lowest-cost storage option available under GCS, at the time of filing Google’s motion. Id. ¶ 6.

With these calculations in hand, Google then demonstrates the relative burden of maintaining the Subject Tables in the context of its overall preservation obligations as ordered by the Court. Google first calculates estimated quantities of data to be stored under the Preservation Plans, excluding the Subject Tables, and the costs associated with those quantities over a three year period. Id. ¶¶ 8, 9. Google then estimates the quantities of data to be preserved in just the [REDACTED] Subject Tables over three years and applies the same cost calculation to demonstrate the following:

Table 1

Seah Decl. ¶5; Pothana Decl. ¶¶ 4, 5, 7; Greenfield Decl. ¶8; Brown Dkt. 781-3 at 11l Calhoun Dkt. 897-3 at 10.

Table 2


Greenfield Decl. ¶¶9,12, 14; Brown Dkt. 781-3 at 12; Calhoun Dkt. 897-3 at 10. The impact is noteworthy. The preservation burden of the Subject Tables, in terms [REDACTED], approaches [REDACTED] the rest of Google's preservation obligations at approximateley [REDACTED] the cost.

Plaintiffs generally object that Google's evidence of burden as to the Subject Tables was not subject to discovery and is therefore intested. Calhoun Dkt. 933-2 at 1-2. More particularly, Plaintiffs point out that Google's financial analysis is premised on a "retail cloud storage cost" model, not its actual “out of pocket” costs. Calhoun Dkt. 933-2 at 1. Accordingly, the Brown Plaintiffs urge the Court to turn this issue back to the Special Master. Brown Dkt. 800 at 5-6.

The Court determines that such a path is neither necessary nor a judicious use of the Parties’ or the Court’s resources at this stage of the litigation. Discovery has closed. In the Calhoun matter, summary judgment has issued against Plaintiffs and is now on appeal.[7] In Brown, a class certification order certifying a class for injunctive relief but not for damages has issued, and interlocutory review of that ruling by the Ninth Circuit has been denied. Further, it is not disputed that there are vast amounts of data at issue in this case and that, accordingly, the Preservation Orders have broad reach and scope.[8] Nor can it be realistically disputed that data preservation on the magnitude ordered in this case imposes practical and financial burdens on Google. As for the disputed financial model, in this context the question is academic in light of basic truths: storage space is not unlimited; there are costs to storing data; here, where there are indisputably large amounts of data, it is reasonable to assume that the associated costs are significant. Further, as explained below, the Court is no longer considering cost-allocation resolution to this dispute, so the precise cost to Google is not so important as to warrant further discovery at this late stage of the respective cases.

The relevant question is whether that burden is unreasonable. More particularly, in the parlance of Rule 26, the issue is whether preservation of the data in the Subject Tables in particular is proportional to the needs of the litigation. At this stage of the litigation, in both the Brown and Calhoun matters, the Court does not need further discovery to make this evaluation. Both sides identify the factors to be considered which the Court addresses below.

Importance of Issues at Stake in the Litigation: The issues of privacy and the unauthorized collection of personal data is of importance not just to the litigants, but the general population. However, the importance of the data at issue in the Subject Tables as it relates to the issues in these actions, where, as here, that data has been demonstrated, largely if not entirely, to be derived from or duplicative of data already being preserved, is far less.

The amount in controversy: As a result of the recent rulings on class certification (Brown) and summary judgment (Calhoun), this factor is indeterminate at this juncture. There is a strong argument that the amount in controversy in Calhoun is now zero, though it is unknown how the appeal to the to the Ninth Circuit factors into the determination. Brown is limited to injunctive relief, hence from the perspective of dollars available as damages, there is no amount in controversy. However, from the perspective of scope of relief sought, the “amount” is large. Nor can the precise financial costs of the preservation of the Subject Tables as it relates to the amount in controversy be determined. What is known is that the costs of preservation of large amounts of data in the Subject Tables is not insignificant.

Relative access to relevant information: It is the nature of this dispute over data collected by Google that Google has far greater access to relevant information. However, there are elaborate Preservation Orders in place in both cases to ensure Plaintiffs’ access to relevant information. The Subject Tables are but one piece of those Orders and, as demonstrated, a piece that is accessible through data that is otherwise being preserved.

The Parties’ Resources: Again, this factor generally weighs against the larger corporate defendant. However, the Preservation Orders in these cases are taxing on Google’s resources – both practical and financial, and as discussed above Google’s resources are not without limits, particularly with regards to storage capacity.

Importance of Discovery in Resolving the Issues. The Court addresses this factor above. At this stage of the litigation, and for all of the reasons identified in this Order, the importance of discovery as to the full content and structure of the Subject Tables is not sufficiently important to re-open discovery.

Taking all of the above factors into consideration, as reflected in the reasoning set forth in this Order, preservation of the Subject Tables is not proportional to the needs of the respective actions. Finally, the Court had indicated it may consider an allocation of costs. However, in light of the weight of evidence, preservation of the Subject Tables is simply not warranted. And the complexities of identifying the costs both in terms of dollars and storage capacity is similarly not warranted at this stage of the respective cases.

IV. ISSUES RAISED IN POST-HEARING BRIEFING

At the February 14 hearing, the Court ordered Google to provide a written response, directed to both the Brown and Calhoun Plaintiffs, to the Calhoun Plaintiffs’ February 6, 2023 letter. Brown Dkt. 861; Calhoun Dkt. 983. The Court further ordered a Joint Submission on any unresolved issues, which the Court has now considered. Brown Dkt. 890; Calhoun Dkt. 996 (“Joint Submission”).

A. Brown Litigation

In the Joint Submission, the Brown Plaintiffs do not challenge Google’s demonstration that Biscotti IDs in the [REDACTED] tables are being preserved in underlying source tables. Plaintiffs do complain that the underlying tables reflect only samples of data, as required by the Preservation Order, and as a result, the entire [REDACTED] tables cannot be reconstructed. Brown Dkt. 890 at 10. Plaintiffs’ complaint is not well taken. The Subject Tables were only subject to preservation as needed to use the preserved data; the Subject Tables were never subject to preservation in their entirety.

Which brings the Court to the Brown Plaintiffs’ remaining arguments in the Joint Submission, that the Subject Tables should have been disclosed during the discovery process pursuant to various Court discovery orders and the Special Master process and that Google’s initial and continuing failure to preserve the Subject Tables is a violation of the Preservation Order. Brown Dkt. 890 at 6-10. Plaintiffs argue that the appropriate response of the Court to the Motion is to “reinforce its prior Preservation Order, which prevents Google from challenging class certification and class membership because of the absence of data.” Id. at 6-7.

As to whether the Subject Tables should have been identified as data sources during discovery or in the process before the Special Master, the Parties are well aware that the identification process was laborious, with much give and take between Plaintiffs, Google and hours of input from the Special Master. Unpacking that process is far afield from the issues before the Court today and the Court declines to do so. As for the place of the Subject Tables in the context of preservation of data, the fact remains that preservation of mapping and linking tables was first brought to the Court’s attention by the Brown Plaintiffs’ objections to the Preservation Order. See section II above. As noted above, the Court ordered the preservation of tables needed to work with the preserved data. The gravamen of Google’s pending request is that the Subject Tables are not needed, therefore they are outside the scope of the Preservation Order. To evaluate Google’s position, the Court ordered the process detailed in this Order. Because the Court finds that Google has satisfied its obligation to this Court in demonstrating that the pertinent information in the [REDACTED] tables and [REDACTED] tables is available in other preserved tables, preservation of the Subject Tables is not needed and therefore the issue of whether the Subject Tables should have been preserved all along is moot.

As to the [REDACTED] tables specifically, at the February 14 hearing there was a discussion around a further demonstration by Google based upon the Plaintiffs’ experts’ data. Unfortunately, the Parties were unable to agree as to parameters that for that demonstration. However, on the showing by Google to the Court as set forth in Google’s letter of January 26, 2023 (Brown Dkt. 848-5) and at the February 14 hearing, the Court is satisfied that the material in the [REDACTED] Tables is duplicative of data otherwise being preserved. For the reasons expressed above, all other complaints raised by Brown in the Joint Submission are overruled.

B. Calhoun Litigation

In the Joint Submission, the Calhoun Plaintiffs raise five issues, numbered A-D(1) and D(2). Calhoun Dkt. 890 at 11. For Issues A and B, the Plaintiffs request supplemental declarations from Google personnel, not counsel, to verify certain representations made during the meet and confer process. Id. at 10-11. In response, Google points to a declaration and deposition testimony already in the record on these issues. Id. at 4. The Court has reviewed Google’s proffer and is satisfied that the testimony under oath by Google representatives fulfills Plaintiffs’ requests. The issues raised in D(1) and D(2) regarding disclosure of the Subject Tables in discovery and the Special Master process are addressed in the Brown discussion above.

Finally, in issue (C), the Plaintiffs complain that Google will not respond to the question of whether the Subject Tables are being preserved for any other matter, arguing that the answer informs the burden on Google to preserve the Subject Tables in this case. Id. at 12. Google declines to respond on relevance grounds and suggests that the answer is privileged or “otherwise protected.” Id. at 4. The Court does not need an answer. Although it sounds like a simple question, nothing in this case relating to data management is simple. Is the question directed at the Subject Tables in their entirety? What if only portions of the Subject Tables are subject to other preservation obligations? What constitutes “preservation” in other cases? What is the “preservation” time period? How responses to these more detailed questions would actually impact the burden of preserving the Subject Tables in this case is unknown and, certainly at this juncture, unknowable. Additionally, and not insignificantly, are the responses subject to privilege or attorney work product protection? The burdens of storage costs and limitations on storage space are what they are, as established by Google. See section III.B. above. And the burden of preserving the Subject Tables is in addition to the overall burden of the Preservation Orders. It is in this context that Court weighs proportionality and concludes that the burden of preserving the Subject Tables outweighs the needs of the litigation.

V. CONCLUSION

For all the reasons set forth herein, the Court finds that the burden of preserving the Subject Tables outweighs the needs of the Brown and Calhoun litigations and therefore the Motion is GRANTED.

SO ORDERED.

Dated: March 20, 2023

Footnotes

At the Court’s request, the Parties submitted the video recording and transcript of the meet and confer session. See Brown Dkt. 851; Calhoun Dkt. 975. The Parties are requested to file the transcript of the meet and confer session under seal on ECF.
Google provided some additional information in response to Plaintiffs’ questions on January 28 (Brown Dkt. 848-6; Calhoun Dkt. 973-6); January 30 (Brown Dkt. 848-7; Calhoun Dkt. 973-7); and February 17, 2023 (Brown Dkt. 889-5; Calhoun Dkt. 995-5).
At the February 14 hearing, the Calhoun Plaintiffs deferred substantive argument regarding the Subject Tables pending a response their outstanding letter to Google for additional information. Calhoun Dkt. 985 at 31-37; 49-51. Pursuant to this discussion, the Court ordered further meet and confer and supplemental briefing, which is addressed in section IV below.
The identifier for the linked Biscotti IDs is called a “Group ID” by Google and referred to as a “[REDACTED] by Plaintiffs. Because the tables originate with Google, the Court will use Google’s “[REDACTED]” term, except when quoting Plaintiffs.
Most recently Google also suggested that source code for [REDACTED] which is being preserved, “could potentially” be used to re-generate the [REDACTED] tables. Brown Dkt. 889-5; Calhoun Dkt. 995-3. Google does not sound confident in this process, and the Court does not factor this suggestion into its reasoning.
The Calhoun Plaintiffs deferred substantive argument. See footnote 3 above.
As explained in the Court’s January 17, 2023, order in Calhoun, the Court retains subject matter jurisdiction to rule on Google’s request seeking to clarify its obligation to preserve mapping and linking tables in that case. Calhoun Dkt. 960 at 2.
Although not broad enough, in Plaintiffs’ view; Plaintiffs in both cases continue to assert their objections to the approach of sampling data rather than preserving it in its entirety.