In eDiscovery, merely finding data isn’t sufficient; it needs to be used strategically for the greatest business benefit. Recent years have prioritized lowering eDiscovery costs and minimizing the total document population, but according to Elizabeth Erickson, an engagement manager and eDiscovery consultant with UBIC, “at a certain point you begin to hit a wall. You’ve gone as far as you can to drive down costs, then you just can’t go any further.” As a solution, Erickson and her colleagues see a way to adapt the linear EDRM progression of a single data set into a more fluid model that retains or even increases in value from case to case.
The majority of eDiscovery costs, some 70 percent, reside in the review stage. While technology-assisted review (TAR) and predictive coding are exceedingly valuable tools, at some point, labor-intensive human coding is necessary. Traditionally, after litigation is completed, all the relevant material is eliminated. “When the case is over, you purge everything,” Erickson describes. “Purge the metadata. Purge the attorney coding, all of that document review you paid for, it’s all gone. We thought, why are we doing that? Certain types of coding performed during review are very valuable and can be used for other litigation or business matters.”
Erickson cites documents coded as privileged as a prime example: “Privilege doesn’t change from one case to another. That information would be very valuable to retain and use for other litigation. Why spend money for another human reviewer to perform a privilege review on the same set of documents? It’s almost kind of insane – even if you’ve done it electronically, you’ve already done it once. But in some cases, this could be happening over and over and over again. We wanted to help our clients take advantage of this valuable information that they’ve already paid for and eliminate a redundancy.”
What was fundamentally lacking was the framework to carry over data from review to review. “We asked, what if a repository was available where all of that information could reside, but not have it exist in multiple instances? You don’t need numerous copies when you can have one that maintains a multitude of metadata associated with it, including coding decisions made by attorneys during litigation in an associated matter,” notes Jay Stromberg, director of marketing. UBIC’s software and workflow can accommodate this approach.
“The way we conceptualized it is with the creation of a Big Data Case Manager application, a repository of all of the information that is associated with a case – all of the data that’s been collected, processed, reviewed and coded,” Stromberg describes. “Then when you have litigation and you know that a specific custodian is involved, you go to that repository for access to the information. It’s a combination of preservation and reuse.”
Though the product bears the results of data review, the investment is essentially one in information governance, which, Erickson concedes, is not always a corporate priority. Yet, she says, the tendency for companies to focus spending on immediate needs and not the potential efficiencies a more robust information governance approach can offer is ultimately a disservice to their preparedness: “Then when you have litigation, you’re less prepared for it, and your litigation efforts may be more costly. If you spend a little money at the outset to bolster your organization, your workflows, and the way that you collect and store your data, when it does come to litigation you’re ahead of the game. You can take advantage of those efficiencies and move forward in a very cost-effective way. You’re still going to perform all the investigations, analysis and review of documents during discovery, but you're starting out with some of that information already in place.”
The team at UBIC likens the relationship between information governance and eDiscovery to an infinity symbol, with data freely flowing between the two loops, becoming ever more refined with each pass through the central knowledge center. “We are trying to think about eDiscovery and information governance as something cyclical,” says Stromberg. “You have a pre-litigation world, where data can be analyzed, audited and previewed in advance of litigation. The information in the metadata and the decisions that were made previously about that data give you a head start on litigation.”
“Beyond that, what if you can get out ahead of a case even further, before it becomes litigation,” adds Erickson. “Perhaps even prevent it from reaching the point where it is an incident that results in litigation. That would be the ultimate way to cut litigation spend.”
“By way of example, we have a tool called Email Auditor,” Erickson goes on to say. “It uses artificial intelligence and predictive coding technology to audit employee emails. Let’s say you have a sales organization, and you’re concerned about price-fixing with your sales team. Maybe there has been litigation in the past, so it is a known problem for you. Instead of waiting for an incident to happen, what if you use software to audit your employees’ emails? It looks at their emails and is able to say, ‘Take a look at this. Let me know if this is what you’re worried about?’ If it is it, the system then tries to find more emails like that. Now you have a heads up that there’s some problematic activity in your organization, and you can intervene before it becomes litigation.”
Instead of an eDiscovery cycle progressing on a single matter from left to right, with UBIC’s discovery data continuum, the two equal halves of information governance and eDiscovery are sharing all the work that is performed in order to be as proactive and efficient as possible. Erickson summarizes it as establishing “an overarching concept to extend the life of your data beyond just a one-time use.”
Published November 3, 2015.