OPINION AND ORDER

DEBRA FREEMAN, Magistrate Judge.

In the above-captioned action, referred to this Court for general pre-trial supervision, Plaintiffs allege that defendant Deutsche Bank National Trust Company ("Defendant"), which serves as trustee for 21 mortgage-backed securitization trusts, breached contractual and fiduciary obligations owed to Plaintiffs in connection with those trusts. Currently before this Court is Plaintiffs' letter motion (Letter to the Court from Christopher J. Lucht, Esq., dated May 12, 2017 ("Pl. Mtn.") (Dkt. 144)) to compel Defendant to run a search for electronically stored information ("ESI"), particularly emails and attachments, using approximately 245,000 individual loan numbers as discrete search terms — or, at least, to run a sample search, using 16,000 select loan numbers as search terms — and to produce documents responsive to Plaintiffs' document requests that the search returns. This Court initially heard argument on this issue during a telephonic discovery conference held on May 8, 2017, and has now considered the parties' supplemental written submissions, including Plaintiffs' letter motion, Defendant's opposition (see Letter to the Court from Rollin B. Chippey, II, Esq., dated May 19, 2017 ("Def. Opp.") (Dkt. 145)), and Plaintiffs' reply (see Letter to the Court from Christopher J. Lucht, Esq., dated May 23, 2017 ("Pl. Reply") (Dkt. 146)). This Court finds persuasive Plaintiffs' arguments that the requested search for individual loan numbers would likely result in the discovery of relevant documents not previously produced, but also takes note of the burden that the requested search for 245,000 or so loan numbers would impose and the fact that the "yield" of that search may be relatively modest. Under the circumstances, Plaintiffs' motion to compel is granted, but, at this time, only as to 16,000 loan numbers proposed by Plaintiffs as a test sample.

DISCUSSION

It is this Court's understanding that the searches of ESI that Defendant has conducted to date have used, as search terms, relevant trust names and Defendant's own "trust identifiers." (Def. Opp., at 2 n.2.) Defendant explains that its "routine" internal practice is to use these trust identifiers — not individual loan numbers — in communicating by email, both internally and externally, about loans within the trusts. (See id., at 3.) Thus, Defendant argues that, for the most part, emails regarding the loans relevant to this case would already have been located through the searches that it has conducted, and that a new search for the individual loan numbers will yield little, if anything, new. From this, Defendant posits that the burden of the proposed search, even as to a subset of the loans in question, would outweigh its benefit and not be proportional to the needs of the case.

Plaintiffs, however, have provided the Court with four examples of email strings that have been produced in discovery by Defendant, each of which include emails referencing individual loan numbers and not the trust names or identifiers on which Defendant has relied. (See Pl. Mtn., Exs. 2-5.) While it is true that these four email strings have been located and produced by Defendant as a result of the searches it conducted, it appears that these documents were located only because at least one of the emails in each string included the trust name or identifier, or because an attachment to one of the emails in the string included such information. Plaintiffs have fairly demonstrated that, but for the happenstance that particular emails or attachments in each string included the trust name or identifier, the remainder of the relevant emails would not have been located by Defendant's prior searches, and therefore would not have been produced. Further, Plaintiffs' submitted evidence shows that Defendant's assertion that it was not its practice to reference individual loan numbers in emails, without also referencing a trust name or identifier, is incorrect — or, at least, that Defendant's employees did not utilize that practice consistently.

In arguing that the additional burden imposed by the electronic search requested by Plaintiff would be disproportionate to its likely yield, Defendant points to the results of a similar search for individual loan numbers, performed in another case, Royal Park Investments, SA/NV v. Deutsche Bank National Trust Co., No. 14cv4394 (AJN) (BCM) (S.D.N.Y.). (See Def. Opp., at 2.) According to Defendant, its search for 72,000 individual loan numbers identified by the plaintiff in Royal Park (see Dkt. 356, in 14cv4394, at p. 2), returned 733,000 documents not previously produced, of which — based on its review of a sample of those documents — Defendant deemed 1.38 percent, at most, to be responsive to the plaintiffs discovery demands (id.). Based on this, Defendant argues that a similar search here, regardless of whether it is performed using the entire universe of individual loan numbers at issue or only a subset of 16,000 of them, "would not provide a statistically significant responsiveness rate." (Def. Opp., at 2; see also PI. Mtn., Ex. 6 (Letter to Rollin Chippy, Esq., from Christopher J. Lucht, Esq., dated Apr. 6, 2017).)

This Court notes, though, that, in Royal Park, the plaintiff apparently did not concur with Defendant that the responsiveness rate for the search performed was as low as Defendant represented. (See Dkt. 381, in 14cv4394, at p. 25-26.) Moreover, even if Defendant has accurately described the statistical yield for the search it performed in Royal Park, this Court has no way to gauge the potential importance of the non-duplicative documents that were located through that search. In this Court's view, Plaintiffs should be given the opportunity to review at least the additional documents that a sample loan-number search would uncover and that Defendant would produce as responsive to document requests, and then to return to this Court if, in their view, the volume and/or significance of any newly produced documents warrants a further, more extensive search.

Finally, while Defendant seemed to take the position in Royal Park that the loan-number search being requested in that case would cost $180,000 in vendor fees alone (see Dkt. 381, in 14cv4394, at p. 22), Defendant here states that its anticipated vendor costs would be only about $22,000 for the 16,000 loan-number search proposed by Plaintiffs (see Def. Opp., Ex. 1 (Declaration of Andrew Goodman Submitted In Opposition To Plaintiffs' Motion To Compel the Trustee To Use Numerals Representing Loan Numbers As Stand-Alone ESI Search Terms, dated May 19, 2017 ("Goodman Decl.") (Dkt. 145-1) ¶ 61)). As in Royal Park, this Court finds it reasonable for Plaintiffs to share one-half of these vendor costs, in the approximate amount of $11,000. (See Dkt. 381, in 14cv4394, at p. 25.)

CONCLUSION

For the foregoing reasons, Plaintiffs' motion to compel (Dkt. 144) is granted, to the extent that Defendant is directed to run a search for ESI (specifically emails), using, as search terms, the 16,000 individual loan numbers listed in Defendant's Repurchase Tracking Log (DBNTC PHOENIX LIGHT 2731279), and to produce to Plaintiff all responsive, non-privileged documents discovered through that search, unless already produced in discovery in identical form. If Defendant believes it would be helpful to add qualifiers to this search, to try to avoid "false hits" (in other words, to try to exclude documents where the searched-for series of digits (constituting the loan numbers) are found, instead, in telephone numbers, Social Security numbers, or zip codes), then counsel should confer in good faith regarding the use of such qualifiers, and attempt to reach agreement on methodology.

This Court's ordering of only a sample loan-number search at this time is without prejudice to Plaintiffs' right to renew its motion for a more extensive search, should Plaintiffs believe that the results of the sample search would justify a search for more loan numbers.

The Clerk of Court is directed to close Dkts. 144 and 146 on the Docket of this action, consistent with this Opinion and Order.

SO ORDERED.