• Cullen and Dykman LLP Blogs

  • Archives

  • Is Predictive Coding the Right Litigation Strategy for Your Company?

    Many companies associate the discovery phase of a lawsuit with numerous attorneys manually reviewing each one of its hundreds, if not thousands or millions, of documents. While such manual review certainly has its advantages, modern technology utilizing predictive coding may assist in expediting the document review process.

    Predictive coding is becoming increasingly popular due to the voluminous amounts of electronic data that companies are storing throughout their day-to-day operations. However, some companies have refrained from implementing this technology out of fear of the unknown – ­specifically, that the predictive coding algorithms may retrieve unwanted information during discovery. It is partly due to such fears that many of these companies will opt for the customary manual review of its electronically stored information (“ESI”).

    Predictive coding utilizes sophisticated algorithms to sort through ESI. In order to create the algorithms, an attorney that is most familiar with the case will input information into the predictive coding program that will train the computer how to find each document. The attorney will then assign a score to a sample set of relevant documents. The computer will utilize the score and sample set to project a much more accurate algorithm, which is then used in the retrieval of additional documents that fit the sample set. As the discovery process develops, adding or retracting certain words into the software can modify the algorithms. The predictive coding software will then integrate these modifications throughout the discovery process without disrupting the search.[1]

    As predictive coding technology has advanced, so has its accuracy. Some data indicates that predictive coding technology yields a 75% accuracy rate, which is relatively high when compared with the 60% accuracy rate achieved by manual human review, and the 20% accuracy rate of keyword searches.

    Despite the slow adoption of predictive coding, in 2012 the software program gained judicial validation in the Southern District of New York.  In Da Silva Moore v. Publicis Groupe,[2] Judge Andrew Peck expressed his approval for the use of predictive coding and impliedly commented that such strategies should be imposed to achieve just, speedy, and inexpensive discovery costs. To further promote the use of predictive coding, Judge Peck wrote, “[w]hat the Bar should take away from this Opinion is that computer-assisted review is an available tool and should be seriously considered for use in large-data-volume cases where it may save the producing party (or both parties) significant amounts of legal fees in document review.”[3] Judge Peck even opined that while keyword searches alone “are not overly useful,” the use of keywords, “along with predictive coding and other methodology, can be very instructive.”[4]

    Judge Peck’s opinion also sets forth factors that courts may consider when deciding whether it is appropriate for a party to use predictive coding during the discovery process. Such factors include:

    (1) the parties’ agreement, (2) the vast amount of ESI to be reviewed (over three million documents [in this case]), (3) the superiority of computer-assisted review to the available alternatives (i.e., linear manual review or keyword searches), (4) the need for cost effectiveness and proportionality under [Federal Rule of Civil Procedure] 26(b)(2)(C), and (5) the transparent process proposed by [the parties].[5]

    Judge Peck’s decision should not be taken lightly.  As parties, especially larger institutions, approach discovery, predictive coding is likely to result in monetary and productive benefits.  It is for these reasons that legal counsel for corporate clients should keep in mind Judge Peck’s sentiments regarding this technology as part of their litigation strategy.

    If you or your company would like more information regarding predictive coding or other e-discovery related issues, please email Cynthia A. Augello at caugello@cullenanddykman.com or call her at (516) 357-3753.

    Special thanks to Melissa Cefalu, a law student at Maurice A. Deane School of Law, and Scott Brenner, a law clerk at Cullen and Dykman LLP, for their assistance with this post.


    [1] Edward Sohn, Predictive Coding Today: Before You Jump In, What Should You Consider?, The Metropolitan Corporate Counsel, June 2013, at 28, available at http://www.metrocorpcounsel.com/pdf/2013/June/28.pdf.

    [2] 287 F.R.D. 182 (S.D.N.Y. 2012). To read a summary of the decision, please refer to our earlier blog post available at the following link: http://ediscovery.cdllpblogs.com/2012/02/southern-district-york-addresses-predictive-coding-e-discovery/.

    [3] Id. at 193.

    [4] Id. at 185.

    [5] Id. at 192.