Subscribe to the
Corporate Counsel Advisory Newsletter
to stay up-to-date on the latest news and best practices.
Ideas and suggestions are always welcome. Please let us know how we can improve your newsletter! We welcome your feedback.
Editor: Vanessa Lloyd, Corporate Counsel Marketing, LexisNexis
Contributing Editor: Tom Hagy, HB Litigation Conferences LLC, former VP LexisNexis and Publisher of Mealey’s™ Litigation Reports
LexisNexis® for Corporate Counsel
LexisNexis® Webinar Center
LexisNexis® Legal Newsroom
Live CLE Webinars | OnDemand Webinars
Big Data: What It Is and What It’s Good For
Are you or your outside law firms using data analytics to identify profitability of individuals or specific matters? Or to support or refute claims and allegations? Or to spot and exploit trends and opportunities? While definitions may vary, these are some of the things made possible by “big data,” according to legal and technical experts who shared their insights during a one-hour panel discussion hosted by LexisNexis® Litigation Solutions at LegalTech® New York on Jan. 30, 2013. Speaking to a packed conference room, the panel comprised four thought leaders who were moderated by nationally recognized e-discovery expert George Socha, co-founder of EDRM.
Orrie Dinstein, Chief Privacy Leader and Senior IT & IP Counsel for GE Capital, started off by saying that “big data” is not just a large amount of data, although that is part of it. He also believes there needs to be a problem to solve. Dinstein defines big data with the “Three Vs.” First, is volume. That means terabytes or petabytes―once unimaginable volumes of data.
Second is velocity. Dinstein defined this with a hypothetical. A power plant has a thousand sensors all connected to a central system and each sensor is reporting back status information at constant intervals. Taking Dinstein’s example, at a rate of a thousand bits of information perhaps every second, each day would generate over 1.4 million bits of data, or 511 million bits of data a year―and that is only one source of data. Capturing all that data poses a challenge.
The third “V,” Dinstein said, is variety. In the e-discovery context, the parties may be gathering email, documents, Web activity, and voice mail―both structured and unstructured data. When all of this goes into a single bucket―that variety creates a big data type of issue, he said.
Neiditz said big data also helps solve problems and tell stories. With different modes of analysis brought to bear on varieties of data it can be used for “sense making” and pattern detection, he said.
Browning Marean of DLA Piper gave the example of Google™ Flu Trends (www.google.org/flutrends), in which the company analyzes peoples’ search terms as indicators of, as Google puts it, “current flu activity around the world in near-real-time.”
Other examples of big data’s uses were the examination of information coming into call centers with details of product problems or HR complaints, data that can be mined to spot potential problems.
How Your Outside Firms Might Use Big Data
Chris Emerson, Director of Practice Economics at Bryan Cave, brought his technology background to the firm, applying collection, coding and analysis of data to determine, for example, whether a case involving overtime violations in an employment case was really as big an issue as a plaintiff attorney claimed it was. “We were able to show that the magnitude of the violations was much smaller than they thought, then we were able to push the case to a settlement. This also saved the client money by not having to get outside analysis,” he said.
Emerson said he is using data, including time entries at the firm―with a fair amount of coding, cleanup, human work and “training” the system―to analyze the outputs of attorney work in several task categories. He said the system filters data much in the same way an email system filters junk mail―just one type of filter an email system puts on incoming messages. Once in place they were able to understand how much time it took attorneys to complete certain tasks. From that, the firm was able to produce better budgeting information, he said. In this way, Emerson replaced what at one point in time was a man standing next to workers with a clipboard and stopwatch.
Emerson said he uses this kind of data analysis to determine the profitability of a particular client or matter―or even a particular attorney. “We deliver a lot of financial reporting to the management and compensation committees about attorneys. We found that some matters, although they looked good originally, were not that profitable,” he said. With his processes in place they are able to look at the top 10% a bottom 10% of profitable attorneys and explain the factors that play into why they are different.
But something had to be done with how information was relayed. “We knew that the traditionally dense spreadsheets needed to be replaced by narratives that would explain why an attorney’s number was what it was. We then wrote a real piece of software that can go out anywhere at the firm and connect different pieces of software, databases and Web―then determine different story angles. The report includes factors such as who is new to the firm and who has business relationships―just two factors that play into an attorney’s profitability. “Using this kind of analysis we also can generate next opportunities and dashboards. For example, there might be an opportunity to collect from a certain client or identify a matter that is buried among others.”
Responding to the types of data Emerson said was being collected on individuals, Dinstein said “the privacy lawyer in me is sitting on the edge of his seat.” He raised questions about disclosures, consents by the individuals and the potential for international privacy issues for global firms. But, in the context of the discussion whether this posed a big data challenge, he agreed that this is a big data issue because of the variety of data being analyzed from different sources.
High-Trust Culture Required
Neiditz said the privacy issue is a critical one, and can only take place in a “high-trust culture” which, he said, apparently is present at Emerson’s firm. “You have to allow for the laying of different kinds of data over other kinds of data,” he said. “You need good information security and rules.” Neiditz sees the upside, but also wonders if this information becomes to your legal career what your credit score has become to online dating. “Will your profitability score at one firm follow you to your next job interview?” he asked.
Moving on, Emerson provided another example in which data analysis helped Bryan Cave on the business front. There was an assumption at the firm that there was a correlation between the size of a book of business and an attorney’s productivity. Because his group was able to analyze the right data, they were able to say there was no such correlation. He said there is a culture of trust at Bryan Cave which permits him to dig as far as needed to get the right answer to a business question, even if it isn’t a popular answer.
An audience member asked about―in the employment context―information such as who has a relationship with whom at a company and whether that information can be added to a database and analyzed.
Telling a Positive Story
Neiditz said that if a company or firm wants to demonstrate that it does not have a hostile work environment, for example, then they “need to work with this data on a privileged basis and demonstrate improvement over time. Then you have a story of improvement. That is what big data can do for you. If it is within a framework of a structured program it can be immensely powerful. These things have to take place in a high trust environment with a truly legitimate authority that is using that information very carefully and with consent.” Later, Marean added that use of social media data will be most robust in the employment law arena.
Marean said that in the context of litigation holds, which he manages, he believes you have to start with human interviews so you get the highest quality of information initially. From there you gather information from various silos and databases and create data maps to get your arms around the data. You must start with interviewing four or five people who have the information you need and find out where the data is, what it looks like and what you should be looking at. Neiditz said organizations need to be careful about what records they are creating, partly because of privacy concerns and partly because information may be discoverable later on. He also cautioned people that the law allows Facebook® and others to block you from looking at private data, but that the government does not have such roadblocks. “The government has broad access to big data watering holes,” he said.
Emerson suggested that if you can find a way to audit email and other documents, this will enable you to identify the departments with the highest risk profiles. “Executive managers will not want to be in the highest risk group, so you will set up a race to the bottom among competitive managers,” he said.
Asked for final thoughts for the audience, Marean said you must study and understand your organization’s information governance. Neiditz added that such a program has to enable you to keep useful information for the purpose you retained it and eliminate the information you don’t need. Not an easy task. Corporations have to look at their infrastructure, including employee devices used for business, or BYOD, cell phone clouds and data lakes that have been allowed to remain because they have value to your company. But the information can be taken out of context and become very dangerous.
Dinstein said companies and firms need to focus on privacy. You must know where the data is and where it came from. Next is security. “The bigger the repository, the more the bad guys will be interested in it,” he said, noting that hackers from China who found large U.S. companies too difficult to attack, turned to M&A law firms because they know attorneys have volumes of valuable information, like trade secrets and other intellectual property.