Bridgestone v. IBM Approves Predictive Coding Use, Rejects Progressive

Adam Kuhn
August 12, 2014

The combined use of keywords and predictive coding is consistently being upheld as defensible. Trying to argue against this method is about as easy as stopping a train. What is less clear is whether a party – after Progressive v. Delaneycan adopt predictive coding in the face of a restrictive case management order. However, just as the controversial Progressive opinion was issued, another court came down with an opposite holding, allowing a party to switch horses in midstream and adopt predictive coding in the face of a conflicting case management order. In Bridgestone v. IBM, U.S. Magistrate Judge Joe B. Brown of the Middle District of Tennessee upheld the switch to predictive coding. In addition, the Bridgestone court approved the combined use of keyword and predictive coding search methodologies. eDiscovery practitioners should read and understand Bridgestone to learn the predictive coding lessons taught by this case. There are two key issues presented in Bridgestone that attorneys face on a routine basis. The first is whether a party can use keyword searches to narrow the universe of potentially responsive information before adopting a predictive coding workflow. And the other is whether a party may adopt predictive coding technology in the middle of discovery and in the face of a conflicting case management order. In short, the court answered both affirmatively. Bridgestone filed the $600 million suit over alleged defects in a sales and distribution program that IBM was hired to build that apparently threw Bridgestone’s “entire business operation into chaos.” After using keyword searches and deduplication, Bridgestone culled 16.6 million records down to 2.3 million. To support its use of predictive coding, Bridgestone argued that it would take 50 attorneys approximately three months to review the remaining documents and that predictive coding offered faster, cheaper, and more accurate results. In response, IBM argued that keywords and predictive coding were “incompatible” and that the parties were “already pregnant, if you will, on the search term process.” IBM ultimately could not support this argument with case law and it was disregarded by the court (the same argument was also made and rejected in In re Biomet). IBM also asserted that the parties had agreed to a case management order requiring “attorney review,” which arguably required manual review and precluded the adoption of predictive coding. Citing to Progressive, IBM asked the court to enforce the existing order regarding the parties’ initial choice to rely exclusively on keywords and manual review. The court also rejected this argument, recognizing that the scale of the review and Rule 26’s proportionality standard required a more reasonable result:

In the final analysis, the uses of predictive coding is a judgment call, hopefully keeping in mind the exhortation of Rule 26 that discovery be tailored by the court to be as efficient and cost-effective as possible. In this case, we are talking about millions of documents to be reviewed with costs likewise in the millions. There is no single, simple, correct solution possible under these circumstances.

The Bridgestone case is significant as it reinforces the notion that the responding party is in the best position to determine how to produce its responsive documents and to select search methodologies to accomplish this objective. Attorneys can now rely on Bridgestone and In re Biomet to defend the combined use of keywords and predictive coding as a defensible search method. Equally important, Bridgestone limits the Progressive holding that parties could be foreclosed from using predictive coding if that search methodology is not memorialized in a case management order. As other cases have shown, it may be a good practice to do so. However, as Bridgestone observes, notions of proportionality and reasonableness may require a different result, particularly since discovery should “be as efficient and cost-effective as possible.” Arbitrarily limiting a party’s choice of search methodologies does little to advance those objectives, much less the mandate from Federal Rule of Civil Procedure 1 that “every action and proceeding” be resolved in “a just, speedy, and inexpensive” manner.