*1 This matter comes before the Court on Plaintiff's motion to compel (Dkt. No. 55). Having thoroughly considered the parties’ briefing and the relevant record, the Court finds oral argument unnecessary and hereby GRANTS in part and DENIES in part the motion for the reasons explained below.

Plaintiff moves to compel the production of documents responsive to its Request for Production (“RFP”) No. 66, which seeks “all IRIX data for applicants during the class period which fall within the group of 13% of data matches that are not matched based on the applicant's social security number.” (Dkt No. 55 at 5 (referencing Dkt. No. 56 at 2).) Among other objections, Defendant asserts that the request is unduly burdensome, explaining that, based on how Defendant stores its data, it would need to sort through 43 million reports prepared during the relevant period, rather than just the reports that Plaintiff seeks discovery on. (Dkt. No. 56-1 at 7–8.) Moreover, as Defendant further explains, the 13% of records sought by Plaintiff refer only to matches for a single data source. (Dkt. No. 82 at 2.) Absent sampling, a fully responsive production to RFP No. 66 would actually entail the production of at least 38 million reports. (Id.)

Parties may obtain discovery regarding any nonprivileged matter that is (1) relevant to any party's claim or defense, and (2) proportional to the needs of the case, considering the Rule 26(b)(1) factors. See Fed. R. Civ. P. 26(b)(1). Discovery, however, “has ultimate and necessary boundaries.” Oppenheimer Fund, Inc. v. Sanders, 437 U.S. 340, 351 (1978). Accordingly, the Court must limit discovery if it determines that the burden or expense of the proposed discovery outweighs its likely benefit, considering the needs of the case, the amount in controversy, the parties’ resources, the importance of the issues at stake in the action, and the importance of the discovery in resolving the issues. Fed. R. Civ. P. 26(b)(1), (2).

The Court agrees that production responsive to RFP No. 66 is necessary to provide Plaintiff the discovery it needs to support its class allegations. See Doninger v. P. N.W. Bell, Inc., 564 F.2d 1304, 1313 (9th Cir. 1977) (describing the need for discovery to determine “[t]he propriety of a class action” including “the existence of a class or set of subclasses.”). However, the production sought is unduly burdensome and not proportional to Plaintiff's pre-certification needs. The Court, therefore, endorses the sampling approach suggested by Defendant in its supplemental briefing but finds that a larger sample size is required. (See Dkt. No. 82 at 2–4.)

For the foregoing reasons, Plaintiff's motion to compel (Dkt. No. 55) is GRANTED in part and DENIED in part. By no later than December 10, 2021,[1] Defendant is hereby ORDERED to produce the requested data for the 23 fields identified by Plaintiff for one-million[2] randomly selected reports, without reference to any particular match type, i.e., not limited to those reports lacking social security number and/or birth date matching. Relatedly, Defendant must also produce the methodology it uses to select the sampled report data, to ensure the selection is, in fact, truly random.

*2 The Court declines to award attorney fees, as Defendant's position was substantially justified. Fed. R. Civ. P. 37(a)(5)(A)(ii).