In what circumstances will a court grant a request to limit the discovery of electronically stored information to certain keyword search terms?

(a) Prevention of abuse. The court may at any time on its own initiative, or on motion of any party or of any person from whom or about whom discovery is sought, make a protective order denying, limiting, conditioning or regulating the use of any disclosure device. Such order shall be designed to prevent unreasonable annoyance, expense, embarrassment, disadvantage, or other prejudice to any person or the courts.

(b) Suspension of disclosure pending application for protective order. Service of a notice of motion for a protective order shall suspend disclosure of the particular matter in dispute.

(c) Suppression of information improperly obtained. If any disclosure under this article has been improperly or irregularly obtained so that a substantial right of a party is prejudiced, the court, on motion, may make an appropriate order, including an order that the information be suppressed.

In this day and age the discovery of ESI is commonplace. Although the CPLR is silent on the topic, the Uniform Rules of the Trial Courts, several courts, as well as bar associations, have addressed the discovery of ESI and have provided working guidelines that are useful to judges and practitioners. Indeed, in 2006, the Conference of Chief Justices approved a report entitled “Guidelines for State Trial Courts Regarding Discovery of Electronically Stored Information.” New York's Uniform Rules For the Trial Courts specifically contemplate discovery of ESI. Rule 202.12(c)(3) allows a court, where appropriate, to establish the method and scope of electronic discovery (Uniform Rules for Trial Courts [22 NYCRR] § 202.12[c][3]).

[89 A.D.3d 79] The Uniform Rules addressing the discovery of ESI are fairly recent. They took effect in 2009. However, the Commercial Division Rules have addressed discovery of ESI for some time. Rule 8(b) of the rules contains requirements similar to those in the Uniform Rules. The Commercial Division for Supreme Court, Nassau County has built on Commercial Division rule 8(b) to develop the most sophisticated rules concerning discovery of ESI in the State of New York. That court also publishes in depth guidelines for the discovery of ESI (the Nassau Guidelines). While aimed at parties, the Nassau Guidelines are appropriate in cases, such as this, where a nonparty's data is at issue.

[...]

The Federal Rules of Civil Procedure take a similar, although slightly more restrictive, approach. Rule 45 provides specific protections to non-parties. A person responding to a subpoena “need not provide discovery of electronically stored information from sources that the person identifies as not reasonably accessible because of undue burden or cost” (Fed. Rules Civ. Pro. rule 45[d][1][D] ). Moreover, “non-party status is a significant factor in determining whether the burden imposed by a subpoena is undue” (Whitlow v. Martin, 263 F.R.D. 507, 512 [C.D.Ill.2009]). Nevertheless, a federal court may still “order discovery from such sources if the requesting party shows good cause, considering the limitations of Rule 26(b)(2)(C)” (Fed. Rules Civ. Pro. rule 45[d][1][D]). Rule 26(b)(2)(C)(i)-(iii) requires a court to limit any discovery: (1) “that is unreasonably cumulative or duplicative,” (2) “can be obtained from some other source that is more convenient, less burdensome or less expensive,” (3) “where the party has already had ample opportunity to obtain the information by discovery in the action” or (4) when “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 discovery in resolving the issues.” The Advisory Committee Notes explain that the costs of retrieving the information are properly part of this analysis.

[931 N.Y.S.2d 556]

Meanwhile, some federal courts have suggested strict limits on the discovery of specific types of data that are typically overwritten or ephemeral. For example, the Seventh Circuit Electronic Discovery Pilot Program has adopted several “principles” to guide litigants through the discovery of ESI. In particular, Principle 2.04 governing the scope of preservation states that certain categories of ESI “generally are not discoverable in most cases.” (Seventh Circuit Electronic Discovery Committee[89 A.D.3d 81], Seventh Circuit Electronic Discovery Pilot Program, 14–15, Oct. 1, 2009). [...]

(c) The matters to be considered at the preliminary conference shall include:

(1) simplification and limitation of factual and legal issues, where appropriate;

(2) establishment of a timetable for the completion of all disclosure proceedings, provided that all such procedures must be completed within the timeframes set forth in subdivision (b) of this section, unless otherwise shortened or extended by the court depending upon the circumstances of the case;

(3) where the court deems appropriate, it may establish the method and scope of any electronic discovery. In establishing the method and scope of electronic discovery, the court may consider the following non-exhaustive list, including but not limited to:

(i) identification of potentially relevant types or categories of ESI and the relevant time frame;

(ii) disclosure of the applications and manner in which the ESI is maintained;

(iii) identification of potentially relevant sources of ESI and whether the ESI is reasonably accessible;

(iv) implementation of a preservation plan for potentially relevant ESI;

(v) identification of the individual(s) responsible for preservation of ESI;

(vi) the scope, extent, order, and form of production;

(vii) identification, redaction, labeling, and logging of privileged or confidential ESI;

(viii) claw-back or other provisions for privileged or protected ESI;

(ix) the scope or method for searching and reviewing ESI; and

(x) the anticipated cost and burden of data recovery and proposed initial allocation of such cost.

In one of the earliest New York cases dealing with ESI the court noted that "electronic discovery raises a series of issues that were never envisioned by the drafters of the CPLR" (see, Lipco Electrical Corp., v. ASG Consulting Corp., 4 Misc3d 1019(A), 798 NYS2d 345 [Supreme Court Nassau County 2004]). Nevertheless, courts in New York, other states and the Federal courts have crafted guidelines and rules to govern discovery for

Page 3

ESI. Thus, Federal Rule of Civil Procedure §26(b)(2)(B) amended in 2006 states that "a party need not provide discovery of electronically stored information from sources that the party identifies as not reasonably accessible because of undue burden or cost" (id). In Chen-Oster v. Goldman, Sachs & Co., 285 F.R.D. 294 [S.D.N.Y. 2012] the court held, based upon the advisory committee notes to the amendments, that the undue burden or costs relate to the accessibility of the ESI not the cost of production. The court explained that "for example, the sheer volume of data may make its production expensive, but that alone does not bring it within the scope of Rule 26(b)(2)(B). Rather, the cost or burden must be associated with some technological feature that inhibits accessibility" (id). This comports with the presumption enunciated by the Supreme Court that "the responding party must bear the expense of complying with discovery requests" (Oppenheimer Fund Inc., v. Sanders, 437 US 340, 93 S.Ct. 2380, 57 L.Ed2d 253 [1978]). Thus, early cases rejected cost shifting on the basis of undue burden. As one court observed in a design defect case against General Motors "the mere fact that producing documents would be burdensome and expensive and would interfere with a party's normal operations is not inherently a reason to refuse an otherwise legitimate discovery request" (Baine v. General Motors Corp., 141 F.R.D. 328 [Middle District of Alabama 1991]).

[...]

In efforts to reduce costs and streamline ESI production, courts have endorsed the use of search terms which narrow the scope of the documents sought. There is an inherent problem

Page 5

utilizing search terms whereby such terms by their very nature may limit relevant discovery. The prevalence or richness of the yield is necessarily constrained by the search terms used. The yield rate which is defined as "the fraction of documents in a population that are relevant to an information need" (Maura R. Grossman & Gordon V. Cormack, The Grossman-Cormack Glossary of Technology Assisted Review, 7 Fed. Ct. L. Rev. 1, 26 [2013]) cannot fall below a minimum, yet undefined, standard. To alleviate these and other problems "where ESI is concerned, this typically means the parties are expected to meet and confer about custodians of relevant ESI, date ranges for searches of ESI, and other search parameters" (see, Brown v. Barnes & Noble, Inc., 474 F.Supp3d 637 [S.D.N.Y. 2019]). In Brown (supra) the court highlighted the burden and cost of preparing and reviewing ESI prior to furnishing it to the other side. Specifically, the court noted that "the negotiation of ESI protocols takes an inordinate amount of time and tends to delay, rather than expedite, discovery. Disputes frequently arise about which and how many custodians' records will be searched, search terms, the scope of so-called seed sets used in some active learning TAR processes, and appropriate sampling and quality control processes" (id). Furthermore, "the review process itself is time-consuming. The average rate of review is about 40-60 documents per hour, though the rate of review can vary considerably based on the complexity of the documents and the

Page 6

experience of the reviewers. Assuming a case involves review of 100,000 documents, it would take 2000 hours for an attorney to review these documents for production if reviewing them at a rate of 50 documents per hour" (id). Lastly, the court noted that collection and review can be very expensive. In the present case all those expenses are increased by the failure to agree on search terms and for protracted conferences and motions submitted. Consequently, where parties dispute the relevant search terms that should be employed the court must make an appropriate determination (Beaton v. Verizon New York Inc., 2020 WL 6449235 [E.D.N.Y. 2020]).

The parties each accuse the other of failing to diligently and faithfully engage in discovery. Numerous attempts to foster agreement between the parties concerning the proper scope of discovery proved fruitless necessitating this motion. The specific discovery issue in dispute concerns the scope and expense of production of electronically stored information ["ESI"]. The parties cannot agree upon search term protocols, the cost of production of ESI and the burden of such production.

[...]

Considering all of the above the court resolves the dispute as follows: the defendant shall produce all the ESI and utilize the three search terms proposed by the plaintiff, namely "Amity," "Amity Street" and "88-98 Amity. However, such ESI will be limited to the eleven custodians defendant proposed, namely, Tony Catanzaro, Gabriele Corazza, Michael Decker, Jamie Ferber, Sally Hu, Matthew Izzo, Kenneth Kruppa, Anthony Kowalski, Marco Martegiani, Rance McFarland, Joseph Roussine, Stefani Soncini, and Michael Wewiora. This protocol should yield the most relevant emails from the most relevant custodians at a modest expense. This incremental approach should resolve some of the more pressing issues raised by the parties (see, Haka v. Lincoln County, 246 F.R.D. 577 [Western District of Wisconsin, 2007]).

Plaintiff contends that the raw data set of documents potentially responsive is enormous and has proposed limiting the search by using search terms to reduce the size of the data set. According to Plaintiff, all documents and information possibly responsive to Defendants' requests resulted in 19 gigabytes of compressed ESI and related attachments. Plaintiff contends that uncompressed the size is 36 gigabytes and is comprised of 280,000 email messages. Plaintiff argues that to review all of the information would take approximately 1400 hours, if the emails are only one page long each. Assuming the emails are an average of two pages long, Plaintiff contends that review time would be approximately 2800 hours. Given Plaintiff's counsel billing rates, review of the data would result in an expense to Plaintiff of approximately $500,000. Defendants rejected this proposal, stating that Plaintiff would not be unduly burdened. Defendants contend that keyword searches will exclude many documents from the scope of the search, including documents attached to emails in .pdf format.

[...]

ESI is clearly discoverable under the CPLR's broad discovery guidelines. See, e.g., Mosley v. Conte, 2010 WL 3536810 (Sup.Ct. NY Co. 2010). "As no specific State statute addresses ESI, courts have interpreted the CPLR so as to be virtually parallel to the Federal provision' set forth in Rule 34 of the Federal Rules of Civil Procedure." Mosley, 2010 WL 3536810, at *6, quoting Delta Fin. Corp. v. Morrison, 13 Misc 3d 608 (Sup.Ct. Nassau Co. 2006).

Here, the information sought is discoverable, and Plaintiff does not dispute that. Rather, the conflict arises over the allegedly enormous amount of data that Plaintiff and counsel would have to sift through to find information relevant to the discovery requests. CPLR 3103 states:

(a) Prevention of abuse. The court may at any time on its own initiative, or on motion of any party or of any person from whom discovery is sought, make a protective order denying, lifting, conditioning or regulating the use of any disclosure device. Such order shall be designed to prevent unreasonable annoyance, expense, embarrassment, disadvantage, or other prejudice to any person or the courts.

Insofar as Plaintiff seeks an order directing that, in searching for and reviewing ESI responsive to Defendants' discovery requests, Plaintiff may locate documents using keyword search terms designed to cull out non-responsive documents, the motion for a protective order is granted as directed infra.

[...]

The Court notes, where keyword searches are ordered, they must not be "designed . . . in the dark, by the seat of the pants, without adequate . . . discussion of those who wrote the emails." William A. Gross Construction Assocs., Inc. v. Amer. Manuf. Mut. Ins. Co., 256 F.R.D. 134, 135 (S.D.NY 2009). It has been observed:

"While keyword searches have long been recognized as appropriate and helpful for ESI search and retrieval, there are well-known limitations and risks associated with them, and proper selection and implementation obviously involves technical, if not scientific knowledge. . . ."

Id., quoting Victor Stanley, Inc. v. Creative Pipe, Inc., 250 F.R.D. 251, 260 (D.Md. 2008). The Court strongly encourages counsel for both parties to work cooperatively in the area of electronic discovery and endorses The Sedona Conference Cooperation Proclamation. www.TheSedonaConference.com. See William A. Gross Constr. Assocs., Inc., 256 F.R.D. at 134. It has been well stated:

Electronic discovery requires cooperation between opposing counsel and transparency in all aspects of preservation and production of ESI. Moreover, where counsel are using keyword searches for retrieval of ESI, they at a minimum must carefully craft the appropriate keywords, with input from the ESI's custodians as to the words and abbreviations they use, and the

Page 3

proposed methodology must be quality control tested to assure accuracy in retrieval and elimination of false positives.' It is time that the Bar- even those lawyers who did not come of age in the computer era- understand this.

Id. at 136. See also, Moore v. Publicis Groupe & MSL Group, ___ F.R.D. ___, 2012 WL 607412 (S.D.NY 2012). This production should be completed within sixty days of the order signed on these motions. If there is difficulty complying with that time frame, it should be brought to the Court's attention promptly.

Applying these standards to the instant motion for a protective order, we find that the motion by defendant was premature. The more prudent course of action would have been for defendant to first make a motion to limit or strike the discovery requests initiated by plaintiff that it found to be overbroad, irrelevant, or unduly burdensome. If, following the resolution of that motion, defendant still believed the costs associated with searching for, retrieving, and producing ESI to be prohibitive, defendant could then file a motion for the costs to be shifted to plaintiff.

There is no occasion, however, for us to opine on the propriety of shifting costs in this matter. There is simply no evidence in the record supporting the fee structure proposed by defendant. There is no indication, for example, as to what experts, if any, will be used to restore deleted or missing documents, or even if there are deleted or missing documents that need to be restored. Indeed, it is unclear how defendant arrived at the retrieval costs it cited. While it may be, as the motion court noted, that defendant demonstrates a reason for either limiting or narrowing plaintiff's discovery requests and shifting some or all of the cost to plaintiff, it has simply not done so on the record to date.