Keyword searches have become increasingly essential as the need to produce email and other electronically stored information (“ESI”) has become the norm in most discovery phases of litigation. But, as one district court has noted, “all keyword searches are not created equal.” Victor Stanley, Inc., v. Creative Pipe, Inc., 250 F.R.D. 251, 256-57 (D. Md. 2008). A recent decision penned by Magistrate Judge Andrew Peck of the Southern District of New York should serve as a “wake-up call” to lawyers engaged in electronic discovery.
Background
In Gross Construction Associates, Inc. v. American Manufacturers Mutual Ins. Co., 2009 WL 724954, at *1 (S.D.N.Y. Mar. 19, 2009), Judge Peck found himself in “the uncomfortable position of having to craft a keyword search methodology for the parties” after a multi-million dollar dispute arose over alleged defects and delays in the construction of the Bronx County Hall of Justice (the “Bronx Courthouse”). The project was developed and owned by the Dormitory Authority of the State of New York (“DASNY”), with non-party Hill International (“Hill”) as DASNY’s construction manager. After DASNY agreed to produce Hill’s project-related ESI, an issue arose as to how to separate Hill’s project-related emails from its unrelated emails. Id.
Each side proposed search terms to be used to extract the relevant emails from Hill’s New Jersey server. DASNY suggested a limited universe of keywords, including “DASNY,” “Dormitory Authority,” “Hall of Justice,” and the parties’ names. In stark contrast to DASNY’s minimalist approach, the other parties “requested the use of thousands of additional search terms” related to general construction issues, such as “driveway,” “build,” “delay,” and the like. DASNY objected, and the court agreed, that such a broad keyword search would require Hill to produce its entire email database because, as a construction management business, virtually all of Hill’s projects relate to those general terms. Id.
The Court’s Decision
What ultimately brought the court to its “uncomfortable position” was the fact that Hill supplied no information as to what terms its employees used in emails regarding the Bronx Courthouse project. Rather, “Hill’s only contribution to the discussion” was to agree that DASNY’s search terms were too narrow and the other parties’ terms were too broad. Thus, the court was forced to craft a keyword search “without adequate information from the parties (and Hill).” At a conference held the day before the opinion was filed, the court ruled that DASNY’s search terms would be used, in addition to the names of the parties’ personnel involved in the project. Id.
The court’s written opinion expressed frustration with this “latest example of lawyers designing keyword searches in the dark, by the seat of the pants, without adequate . . . discussion with those who wrote the emails.” Id. at *2. Although decisions from other courts had previously warned counsel about this problem and proposed remedies, Judge Peck lamented that “the message has not gotten through to the Bar in this District.” Id.
In warning against the danger of lay lawyers attempting to design effective search methodology, the court cited Victor Stanley, Inc., 250 F.R.D. at 262, which instructed that keyword search methodology “should be tested for quality assurance; and the party selecting the methodology must be prepared to explain the rationale for the method chosen to the court, demonstrate that it is appropriate for the task, and show that it was properly implemented.” Additionally, United States v. O’Keefe, 537 F. Supp. 2d 14, 24 (D.D.C. 2008), was cited for the observation that “for lawyers and judges to dare opine that a certain search term or terms would be more likely to produce information than the terms that were used is truly to go where angels fear to tread.”
Thus, under Judge Peck’s view, lawyers should not rely on their own non-scientific guesses about what keyword searches will yield the most appropriate universe of ESI. Instead, lawyers should consult with the individuals who wrote the emails, as well as qualified search designers. There are several ways that lawyers can enlist a non-party’s help in crafting a keyword search. As Judge Peck noted, in this case “Hill was in the best position to explain to the parties and the Court what nomenclature its employees used in emails.” Gross Constr. Assocs., 2009 WL 724954, at *1. Lawyers therefore should take the obvious step of asking non-parties what terms they used when emailing about the subject matter of the lawsuit. The court also suggested an ex-ante approach for non-parties like Hill, namely using “a standard ‘Re’ line in its Bronx Courthouse emails to distinguish that project from its other work.” Id.
Conclusion
Ultimately, the court held that the best solution to crafting an effective keyword search is cooperation among counsel, and “strongly endorse[d]” the Sedona Conference Cooperation Proclamation. Id. at *2. That proclamation provides, in part, that parties will engage in cooperative and proportionate discovery, in order to trend-down the escalating costs of electronic discovery. See www.TheSedonaConference.org. At a minimum, the court instructed counsel to “carefully craft the appropriate keywords, with input from the ESI’s custodians as to the words and abbreviations they used, and the proposed methodology must be quality control tested to assure accuracy in retrieval and elimination of ‘false positives.’” Counsel should understand — “even those lawyers who did not come of age in the computer era” — that cooperation, open communication, and transparency are vital to creating the best keyword searches and convincing the court that the search used was reasonable. Id. at *3.
