Electronic discovery (e-discovery) is viewed by most businesses as a form of litigation insurance: a means of satisfying the information demands of litigators and government regulators in a timely and cost-effective fashion. This only makes sense, as no company wants to suffer potential multimillion-dollar judgments or regulatory penalties.
More and more corporate IT executives, however, are beginning to use e-discovery internally for a different purpose: to fill the increasingly complex and costly internal information needs of their own end users.
E-discovery products use data classification and search techniques that vendors claim are far more sophisticated than simple keyword searches used in most archiving systems, and even powerful Web search engines like Google Inc.'s. The results are better organized corporate data and more accurate search results.
"If you do a search on, say, Java, a Web engine will come back with material that has to do with coffee, Sun's programming language, and an Indonesian island," said Craig Carpenter, vice president of marketing at Recommind Inc., an enterprise search vendor. In contrast, Carpenter asserts that Recommind's e-discovery platform, MindServer, performs "conceptual searches" that find and categorize related information. "If you do a search on shipyards," Carpenter said, "boats also come up."
Tag and release
About two years ago, St. Louis-based Bryan Cave LLP began evaluating various search technologies, not just for litigation purposes, but for "something more fundamental, that would let us flow information to where it needed to go; to assimilate and integrate related information together for some purpose," said John Alber, the law firm's technology partner. "I've come to believe that it's wrong-headed to look at e-discovery, or any search technology, in an application-specific way."
Bryan Cave now uses MindServer to classify information associated with a particular legal matter or project, and place it in a repository that's "accessible to the team on an ongoing basis and easy to find with the search tool," Alber said. Without e-discovery, "you have to manually assemble the information out of a bunch of emails and document attachments," often a time-consuming task, he added. One major payback, according to Alber, has been the ability to bring new team members up to speed more quickly.
Another crucial eDiscovery component is the ability to classify information that was generated outside structured database and document management systems, noted Barry Murphy, a principal analyst at Forrester Research Inc. in Cambridge, Mass. Now that approximately 95% of all business data is generated and stored electronically, "classification is not only a necessity, but an inevitability," he said.
Once classified, preset policies can be used to determine how and how long information is stored, and which security and recovery parameters are applied. Information can also be associated with and pushed out to different projects and groups of knowledge workers, for instance, via a Web portal.
Cut costs and save time
In addition to enabling IT organizations to better manage information across the enterprise, e-discovery also leads to major savings in hardware costs and IT man-hours.
Kazeon Systems Inc.'s e-discovery platform, for example, has enabled one energy firm to manage some 100 terabytes of information distributed across 25 networked storage devices, according to Steve D'Alencon, director of product marketing at the Mountain View, Calif.-based enterprise search vendor. Information is now classified by criteria such as last access time, file size and owner. As a result, Kazeon's customer reclaimed 40% of its networked storage capacity and was able to "slash storage spending," D'Alencon said.
Another Kazeon customer is using the vendor's e-discovery offerings to automatically identify credit card numbers and other personal information and then classify it as confidential, according to D'Alencon. The information can then be stored securely and retained for seven years, as regulations require.
For Bryan Cave, e-discovery's major payback has been knowledge management. "We are getting a much higher level of performance in our engagement and case teams," Alber said. "The information they need to stay current and act is concentrated, and thus easy to use."
E-discovery products can also help companies rein in storage costs through deduplication and compression. Recommind's recently announced Decisiv Email, for example, automatically classifies emails and attachments as they are generated, and "can resolve multiple emails to a single instance," Alber reported. "That's not easy or trivial," given that the amount of stored email at the law firm has been growing 60% per year, and lately has risen to 80%, he said.
According to Recommind, Decisiv Email enabled one early adopter to reduce electronic storage costs by 30%, the storage of hard copy materials by 70%, and secretarial workload by 14%.
Like most businesses, the legal firm has found that more and more business-critical information winds up in emails, Alber said. Decisiv Email, he hopes, will "help us get control of that wild information" that lies beyond the scope of document management systems, he said.
"Once an email or attachment is classified, we can associate it with documents in our document management system that relate to the same matter; and when the retention period is up, we can confidently delete the data or, more likely, put it into off-offline storage," Alber said.
Implementation takes time
Though the benefits are many, implementing an e-discovery infrastructure is often time-consuming and difficult for many businesses, industry sources warn.
For one thing, fully automated data classification is still a ways off. Products such as Decisiv Email typically categorize some information automatically by sensitivity and importance, then refer the rest to a human authority. This is not surprising, given the difficulty of setting standard, universal policies for classifying information.
Alber said he expects Bryan Cave's Decisiv Email installation to take about 18 months. "Getting people to classify data is culturally extremely difficult, so we want to make it as easy as possible and, frankly, that needs a lot of thought," he noted. "We want to make sure we get it right."
Another difficulty is that the e-discovery market currently consists primarily of point products. While vendors are beginning to integrate the different technologies, "in the short term, you will need a number of point products -- message archiving, file archiving, document management/review application, information classification, search/indexing -- to enable e-discovery within your organization," Forrester's Murphy said.
But corporate decision makers ought not wait for the perfect e-discovery product, experts agreed. "Making data findable [and] increasing productivity of knowledge workers is a soft and fuzzy ROI, but very real," Murphy said.
And, of course, it doesn't hurt to have e-discovery in place if a litigator or government regulator happens to come knocking on your door, either.
Elisabeth Horwitt is a freelance writer based in Waban, Mass.