Legal Commentary

"In the beginning, there was brute force linear review of electronic data," where lawyers would start at one end of a collection and march right through it to make document-by-document decisions about relevance, privilege, etc., explain Anne Kershaw and Joseph Howie, in "Crash or Soar?," the October cover story in Law Technology News. Next came "clustering," similar documents are grouped to speed review.

But today's savvy litigators are embracing a new methodology, called "predictive coding" to control costs and increase the accuracy of results, they say. With it, a small subset of documents is "examined by lawyers, and decisions made on those records are then propagated through the document population." The result is transparent, saves time, enhances replicability, improves confidentiality, and shortens litigation timelines, say Kershaw and Howie, leaders of the nonprofit eDiscovery Institute.

The institute recently surveyed 11 brand name e-discovery vendors, including Catalyst Repository Systems, Equivio, Recommind, and Kroll Ontrack. Respondents cited a cost savings of 45% compared with traditional review, and seven reported individual cases where the savings were in excess of 70%.

Of course, there still is resistance from attorneys, the authors note. Many worry about whether judges will accept the methodology. But ironically, it's the judges themselves that are screaming the loudest for adoption, says Kershaw.

Read more about it in our October LTN. Kershaw and Howie also are my guests for the October Law Technology Now podcast, which will be available shortly.

Correction: In yesterday's Alert, Datacert was misspelled in the reference to its sponsorship of Wednesday's "Generals of the Revolution" conference.