Corporate Law Departments Keep Hands Off eDiscovery

  • Corporate Law Departments Keep Hands Off eDiscovery
  • Corporate Law Departments Keep Hands Off eDiscovery
  • In his introductory statement to a recent survey of in-house lawyers, Archer Daniels Midland’s global director of operations David Cambria asked a provocative question: Who’s to blame for the slow pace of innovation in the legal industry?

    “It’s not longer enough to complain about outside counsel and other service providers,” Cambria wrote, adding, “law departments are failing to innovate in game-changing ways, and that is making us our own worst enemy.”

    The eighth annual “law department operations survey,” released last week by the Blickstein Group in cooperation with Huron Legal, is based on responses from more than 100 different legal departments. It provides new data on the use of technology in the eDiscovery sphere.

    Below, the chart from the survey shows that 59.6 percent of respondents said they were “not interested” in bringing predictive coding in-house. Another 50 percent of respondents were “not interested” in hosting data on internal servers.

    Both measures could lead to cost savings, according to the report and lawyers familiar with the technology, who say that using predictive coding during eDiscovery can trim down the universe of information that needs review, and thus reduce the amount of data that needs to be hosted.

    Chart from the Blickstein Group/Huron Legal Eighth Annual Law Department Operations Survey.

    CLICK to enlarge. Chart from the Blickstein Group/Huron Legal Eighth Annual Law Department Operations Survey.

    Jon Osgood, assistant general counsel & director at Ford, is quoted in the report: “Ford can process and store data on Ford’s own network servers, resulting in significant cost savings in monthly hosting fees which would otherwise be paid to an external vendor.”

    At Ford Motor Co., the law department reportedly saves $2 million on an annual basis just on hosting and processing charges by bringing both predictive coding and data hosting in-house.

    According to the widely quoted 2012 Rand Institute for Civil Justice report on eDiscovery costs, document review constitutes the largest cost of eDiscovery. Data hosting is classified as a component of processing, which is the second most expensive part of the process.

    In an interview, Brad Blickstein, a principal at the Blickstein Group who helped develop the survey, said that 64 of the survey respondents — which was conducted online — answered the question about predictive coding.

    Blickstein said it was important to note that many of the respondents may be outsourcing their eDiscovery and relying on an outside vendor who uses predictive coding.

    But other data suggests predictive coding simply hasn’t caught on.

    For instance, kCura said its Relativity software has been used in a total of 82,198 cases, of which 8,513 used analytics that includes predictive coding. That works out to 9.7 percent, although not everyone who used analytics chose to use the predictive coding feature — it is merely an option.

    The number of users who employed analytics has consistently hovered around 10 percent during the time that kCura has operated, according to Shawn Gaines, the company’s director of marketing.

    Gaines also said that cases in which users employed predictive coding contain more than twice as many documents on average as cases where predictive coding was not used, and require one-third fewer reviewers. The predictive coding cases take approximately one month longer to resolve, however.

    So why aren’t more corporate law departments willing to use these tools?

    According to Tom Barnett, special counsel of eDiscovery and data science at Paul Hastings, there’s no single reason that explains corporate law department’s resistance to adopting these measures, but there are a number of legitimate ones.

    Predictive coding, Barnett explained, is a complex technology that requires a large capital and resource investment. Employing someone who has the expertise to use predictive coding efficiently — which often requires a close understanding of statistical sampling techniques and machine learning algorithms — only makes sense for companies with sufficient volume of litigation to keep that person busy, he said. 

    Data hosting requires far less expertise, but entails putting a slice of the company’s data on internal servers, which requires an infrastructure investment and entails the company taking on the risk in the event of a data breach.

    As a result, many corporate law departments have started purchasing “managed services” where they provide the corporate law department with the resources to do these tasks, and have priced it an attractive point that in-house counsel are not yet motivated to bring it in house, Barnett said.

    “It’s fraught with risk,” he added. “It’s about risk management.”

    *

    Top