Editor: In your experience, what have been the principal attributes clients are seeking in an e-discovery provider?
Robman: Our clients include some of the world's largest enterprises, law firms and government agencies, all of which face tightening regulations and increased exposure to litigation from the global economic downturn. That translates into a need to find information quicker, more precisely, under shorter deadlines, with more at stake than ever before. They can't afford a piecemeal approach involving multiple vendors, multiple hand-offs of data and multiple points of failure. So, they're looking for an e-discovery provider that can deliver the efficiency and cost savings of a true end-to-end solution. They're also looking for technology that helps them organize and manages data automatically, before the e-discovery process begins. They know that a proactive, predictive and defensible approach to e-discovery minimizes their final legal costs as well as their legal and regulatory risks. And above all, they need a provider who is committed to their success and responsive to their needs.
Sharp: This is evolving. It's evolving very much along the lines that you would expect in a standard product cycle. When e-discovery first burst onto the scene, all you needed to compete was a baseline processing and hosting capability. It wasn't technologically challenging, but, at the outset, there weren't that many people who had the basic know-how. Prices were exceptionally high because competition was scarce and customer options were minimal. As knowledge spread, barriers to entry came down, and there was a steady acceleration of market entrants. This eventually turned into a mad scramble as people sought to stake a claim before the market crowded out.
However, all this eventually changed as e-discovery expertise and processing became a commodity. As we would expect under standard product cycle models, the market began competing on price. Price was king, and efficiency became a key factor. Some players were unable to compete, and we saw lots of consolidation. Interestingly, consolidation and scale sometimes generated a boomerang effect, yielding inefficiencies and operational overheads that had negative impact on costs, and inhibiting the ability of some of the larger players to compete against more agile smaller players. Another source of competition was the in-house option. As e-discovery technical know-how penetrated law firms, the firms were able to bring processing and hosting technology in-house as well. Combined with the market slowdown through 2009, the only thing that clients seemed to care about was price.
But change is afoot again. The logic is simple. Low processing costs don't solve the client's key business problem. The client's key business problem is the cost of litigation, and the standout cost factor is the cost of review. You can process documents as fast as the speed of light, but you still have to review them, and that's where the pain is. The pain is massive. So with document volumes continuing to balloon, lower processing costs are a necessary but a relatively minor step along the way to a resolution to the problem.
Enter technology, act II. With the emergence in late 2009 and 2010 of predictive coding technology, we are facing a new round of change in the market. Predictive coding refers to a group of technologies in which the software can be trained to identify relevant documents. Predictive coding can dramatically reduce costs by eliminating the need to examine every record. Similarly, it facilitates stratified review by triaging the collection, with low-relevance documents being assigned to low cost review resources, and high scoring documents going to the higher-cost resources.
This technology is changing everything. The demands from e-discovery providers are not immune to this change either. First, we see that many law firms are looking to bring the technology in-house, where they have direct control over its use. This in itself is a dramatic change. Five years ago, when we introduced near-duplicate grouping to the industry, it was the innovative service providers who drove the technology to the law firms. Now it's precisely the opposite. Where the firms do choose to use the service providers for predictive coding, we see that they are being very selective with whom they work. This is not your standard text extract, indexing and hosting. Effective utilization of predictive coding technology requires an understanding of the business process of e-discovery and of the new options that the technology presents in early case assessment, culling strategies and review structure and flow. People are beating a new path, and they are looking for consultants to guide them down that path. Suddenly, just as product cycle models predict, we have a new round of differentiation. We see this with service providers now being differentiated by their predictive coding technology smarts and their early adopter experience in using this technology.
Svoboda: Today, clients are seeking e-discovery service providers who offer superior technology, possess top-notch project management skills, demonstrate the ability to control costs and have a well-established record supporting clients engaged in large complex litigation. During my 15 years as director of Litigation Support at Cravath, Swaine & Moore, I worked with hundreds of e-discovery service providers. (As further background, prior to joining Cravath, I held a similar management position in IBM's Corporate Litigation Department.) In order to insure that our clients were receiving value for their e-discovery dollar, we developed an evaluation criteria for e-discovery service providers. On a quarterly basis, we measured each provider against this tool, and then we ranked them from top to bottom. For example, we painstakingly checked the quality of each service provider's workproduct, and whether or not the images and data they provided met all of our project specifications.
The firms that provided regular project status reports scored higher than those who did not provide such reports. Meeting deadlines and being available for conference calls inside and outside normal business hours are other key measures. Without regard for time zones, we expected a high degree of project manager availability. Also, organization and management continuity is important, particularly in an industry where consolidations seem to be a regular event. Clients want to engage service providers who demonstrate they have a long term commitment to these critical applications. And, finally, we expected well-documented and timely invoices. At Evidence Exchange, our experience has shown that these are the principal attributes clients look for when selecting an e-discovery service provider.
Editor: Will cloud technology have an effect on the way in which e-discovery will be conducted?
Robman: Absolutely. One effect that's less often discussed is what you might call "cloud pollution." The price and ease of cloud storage makes it all too tempting to fall into the "archive everything' mentality. Whether a company is storing data in the cloud, on their own servers or both, it's critical to maintain a strict data management policy that ensures the deletion of old, irrelevant and potentially dangerous information. Even so, the sheer volume of data involved with cloud storage will inevitably drive companies to more automated solutions for e-discovery, litigation preparedness, records management and information governance. Simple e-discovery appliances and other legacy software simply aren't up to the task.
Sharp: The cloud model is compelling in the e-discovery context. In contrast to on-premise software, the cloud model gives the firm or corporation a very immediate and scalable solution. E-discovery needs are characterized by peaks and valleys. Under the on-premise model, in order to handle the peaks, you need scale, which is wasteful in the valleys. This is solved in the cloud model, where volume-based pricing means that the service provider is absorbing the volatility in your processing needs over time. When using a cloud provider with a standard e-discovery solution, the customer is also benefiting from a proven, reliable service delivery option.
The key question, though, relates to the customer's need to identify, adopt and assimilate new emerging technologies. This is an age-old debate for which there is no simple answer. Predictive coding is a case in point. We have seen that many service providers have been slow to adopt this technology, and that the market push is being led by the end users - law firms and corporations. It seems that some of the service providers have too much invested in their legacy platforms to risk the adoption of innovative game-changing capabilities. The conclusion is that while cloud computing can offer an attractive solution at the operational and cost levels, it's important to verify that your supplier has the awareness, know-how and knowledge to help you go down the path of adopting new capabilities, such as predictive coding, when and where required.
Editor: Does the expanded use of social media on the Internet compound the difficulty of conducting an e-discovery search? How do you propose limiting the search on such formats?
Robman: It does make e-discovery more difficult. But the case law around social media is still developing, and business adoption - while growing rapidly - is still relatively low, at least compared to other media like email. We expect that social media will certainly present major e-discovery issues but won't begin causing massive problems for at least another year or two, which means companies still have time to prepare. Technology will be part of the solution, but the most important thing companies can do right now is to develop, vet, implement and audit compliance with a sensible social media policy. For most businesses, there's no point in trying to ban social media as many users are going to embrace it irrespective of such a ban. But the more time companies take preparing and shaping employee compliance, the fewer worries they'll have in the coming years.
Published December 6, 2010.
