Hamzeh v. Pharmavite LLC, No. 4:24-cv-00472-HSG (PHK) (N.D. Cal. Sept. 1, 2026), full opinion (PDF)
A magistrate judge in the Northern District of California on September 1 kept sample product labels inside the time period of an agreed ESI protocol. Pharmavite, a defendant in a putative class action, had not explained why the labels should be treated differently from the rest of its ESI. The parties had agreed that ESI would be produced from January 1, 2017 forward. For two categories, compiled sales spreadsheets and sample product labels, Pharmavite wanted production to start instead at the limitations period, January 25, 2020. The court denied the plaintiff's motion to compel as to the spreadsheets without prejudice. The plaintiff, in the court's view, had not shown why earlier sales figures were relevant or proportional. By contrast, the labels must be produced from 2017.
What happened
The plaintiff complained that Pharmavite was withholding ESI from non-custodial sources for the agreed period. Pharmavite answered that it had produced ESI from both custodial and non-custodial sources under the protocol's search terms. The court accepted that representation, pending a status report on the non-custodial production. The real dispute, Pharmavite said, concerned two categories, which it called Sales Data and Exemplars. Pharmavite has been producing the Sales Data as spreadsheets that compile figures from internal databases. The Exemplars, in Pharmavite's definition, are the labels the plaintiff and the putative class could have relied on when buying the products.
Pharmavite argued that neither category is produced through the ESI protocol process. In its view both bear only on damages. That makes them relevant, Pharmavite said, only from the start of the limitations period. The plaintiff answered that a statute of limitations is not a proper limit on the scope of discovery.
The court's analysis
The party seeking discovery, the court stated, bears the burden of establishing relevance. In turn, the resisting party must show why the discovery should be denied. The spreadsheet ruling went against the plaintiff because it had not established relevance. The label ruling went against Pharmavite because it had not explained why the labels should be exempt from production.
The court observed that the spreadsheets are compiled from internal databases. That is somewhat different, it said, from a standard protocol, which runs agreed search terms across identified sources. Pharmavite's argument that the spreadsheets fall outside the protocol therefore had some merit, in the court's view. In any event, the court continued, the plaintiff had not shown why sales figures from before the limitations period are relevant. Nor had the plaintiff shown that producing them would be proportional to the needs of the case. The court denied the motion as to the spreadsheets without prejudice. The plaintiff may return to the issue if new facts give it a good-faith basis to show that the earlier figures are relevant and proportional.
The labels must be produced from 2017. The court stated that Pharmavite had not explained why the labels are exempt from the agreement covering non-custodial material from that date forward. As far as the record showed, the labels are ESI, the court noted. An agreed ESI protocol, the court said, should by definition reflect what the parties have agreed about relevance and proportionality. Pharmavite cited no case law for exempting sample labels from one. Nor did it argue that three more years of labels would be a burden. That burden was limited regardless, the court added, because the parties had apparently agreed that Pharmavite need only produce sample labels rather than every duplicate.
Why it matters
The magistrate judge treated the protocol as the parties' own agreement on relevance and proportionality for the ESI it covers. A producing party that wants one category treated differently therefore has to explain why the protocol does not reach that material. Failing that, it should expect to show that the extra production would be unduly burdensome. For the labels Pharmavite did neither, so the agreed period governed. For the spreadsheets Pharmavite argued that they are produced outside the protocol process. The court credited that argument in part, then rested its ruling on the plaintiff's failure to show relevance and proportionality.
A court may accept that a category is produced by a process the protocol does not govern. The requesting party still has to say what the earlier data would show. Otherwise production of the earlier data starts where the producing party proposes.
For a producing party, the time to take a category out of the protocol is when the protocol is negotiated. Structured data, compiled reports, and exemplars that will be produced outside the search-term process belong in the protocol by name. Each should carry its own time period. Absent a carve-out, a category takes the protocol's period.
The full opinion is available as a PDF.
