Top Stories From Law.com

Printer-friendly Email this Article Reprints & Permissions

In the fall of 2005, a small Israeli technology startup came to San Francisco's Morrison & Foerster with a lawsuit -- and, soon enough, a problem.

The company had been mired in a contract dispute with one of its business partners, a huge American tech concern, and, unable to reach a settlement, was taking the matter to court.

The stakes weren't particularly high -- just a few million dollars. But after the case was filed, the defendant hit back with an electronic discovery request -- every relevant e-mail, Microsoft Word file, spreadsheet, you name it -- so onerous that its cost alone would take a fair chunk of any judgment.

"We saw that it was going to take several hundred thousand dollars to do this," says Oz Benamram, director of knowledge management and Israel practice counsel at MoFo. In fact, there was nothing terribly unique about this situation. As more correspondence and information is stored electronically, e-discovery is requiring more time, and more dollars, than ever before.

What was different was MoFo's solution. Realizing that the standard way of reviewing documents -- having teams of associates, or lower-priced contract attorneys, sift through anything that could be relevant, deciding what was responsive and had to be turned over, and what was privileged and needed to be kept -- wasn't going to cut it, the firm suggested a radical approach: automate almost everything.

By searching the data by specific keywords, anything that was potentially responsive would quickly be found. A second search, on this smaller batch, would look for lawyer names that might flag privileged materials, and for keywords that might indicate sensitive documents. Only those records would be reviewed by humans; everything else would be turned over without a glance. "In the end, about 18 percent of the material we turned up was reviewed," says Benamram. "We saved 80 percent of the cost." (MoFo also won the case for its small client.)

The technology that MoFo used wasn't particularly sophisticated; just some basic keyword searching. But the idea of turning over documents without even looking at them used to be heresy in the world of discovery. However, that universe is getting bigger at a breathtaking pace, and it's forcing firms to create more innovative, cost-effective approaches. Five years ago, a midsize case involved some 5 gigabytes (GB) of data. Today, anything less than 500GB is considered small.

All of this information needs to be processed -- converted from its raw form on e-mail servers, backup tapes and CD-ROMs into a format that can be more easily searched and analyzed. Then it must be reviewed. It all gets very expensive, very fast. The commercial e-discovery market (not counting what the government spends) was estimated at $1.295 billion in 2005, according to the Socha-Gelbmann Electronic Discovery Survey, up 56 percent from 2004. By 2008, the survey projects, revenues will hit $3.13 billion.

For law firms, the boom in e-discovery presents both an opportunity and a predicament. On one hand, there is money to be made here. By keeping processing and review in-house as much as possible, and leveraging technology to cut workloads and costs, Mountain View, Calif.'s Fenwick & West now reaps seven-figure annual revenues from e-discovery, according to Robert Brownstone, the firm's law and technology director. But the work is full of risks, too, and mistakes can be costly. No one wants to be the next Kirkland & Ellis, which lost a case, and a client, when its botched e-discovery helped trigger a $1.6 billion verdict in 2005 against Morgan Stanley & Co Inc.

Firms are taking different approaches to e-discovery, and so far, at least, no one has created the magic formula. But on one point there is agreement: Electronic discovery is here to stay, and firms need to make the process more efficient -- if not to make money, then to ensure that their clients stay put.

POSTER CHILDREN

For a long time, law firms had something of a "don't ask, don't do" policy on e-discovery. "Most firms just ignored the stuff," says George Socha, a St. Paul-based EDD consultant who co-authored the Socha-Gelbman survey. "Each side would look at the other and say, 'You don't really want to do this, right?' But now it's not going away." Part of the reason is the evolving case law. In 2003, five pre-trial decisions in Zubulake v. UBS Warburg laid out the duties of counsel, including an obligation to monitor electronic discovery and make sure that all sources of potentially relevant information were identified and searched. Gradually, even the old-line firms began ramping up their EDD technology. In 2005 Sullivan & Cromwell went out and hired Thomas Barnett, an outside e-discovery expert and former litigator, to build and run an in-house e-discovery department. "They wanted to be the poster child for doing things right, not screwing them up." says Barnett.

Last December, changes to the Federal Rules of Civil Procedure brought e-discovery front and center, requiring litigants to determine the scope of e-discovery, what data was accessible, what was not and who should bear the costs, as part of their pretrial planning conference. "You need to be ready to address this stuff practically from the first time you're in front of a judge; there is no avoiding it," says Socha.

The more e-discovery work that a firm takes on, the bigger its infrastructure -- and investment -- needs to be, turning law firms into small-scale (and not-so-small-scale) technology companies. Morrison & Foerster now has more than 30 terabytes (or 30,000 GB) of disk storage across the firm just for litigation and discovery. Sullivan & Cromwell has hired 16 computer forensics experts, project managers and technical specialists in the past two years to handle its in-house e-discovery work. Fenwick & West dedicates 20 dual- and quad-processor computers to the work, and relies on 35 different software packages.

IN OR OUT?

One key issue for many firms is to what extent -- or not -- they should be involved in the technical nitty-gritty of e-discovery, the processing work, where data is extracted from all the places it may reside and converted to a format more accessible to lawyers.

For starters, it can be difficult to make money at processing. Vendors flooding the market have driven down prices sharply, forcing firms to lower their margins -- or eliminate them entirely. Foley & Lardner used to charge $1,000-$1,200 per gigabyte of data, but now charges $250, because that's what vendors currently charge. "At $1,200 per gigabyte, the economics of doing it in-house made sense," says Bruce Blank, the firm's director of litigation services and support. "We covered our costs and made miscellaneous revenue." But at $250 per gigabyte, he says, in-house processing is essentially done at cost. Foley now sends out about 80 percent of its processing work, according to Blank, and is looking at dropping its in-house price further, because vendors are starting to go even lower: "If you don't, how do you explain that to clients?"

Yet relying on vendors carries its own risks: namely, relying on vendors. Firms can quickly find themselves losing control of the work, and having a hard time making sure it's done right -- and on time. Paul, Weiss, Rifkind, Wharton & Garrison, which currently does not do processing in-house, is considering a change in strategy -- not to make money (it will be charged at cost), but to boost efficiency. "A lot of our e-discovery happens in regulatory matters, where we need it yesterday," says H. Christopher Boehning, a litigation partner at Paul Weiss. "With a vendor, you're in a queue, and you have only so much control."

But for Fenwick & West, which does almost all of its processing in-house, the Foley & Lardners and Paul Weisses all have it wrong: Firms need to do the work, and they can make money at it. By keeping it under one roof, and automating as much of the process as possible, the job can be done faster and cheaper than if handed off to a vendor. "Our processing work ends up [costing clients] at least 20 percent less than the services of any outside vendor," says Brownstone. Even still, Fenwick comes out in the black. "We don't use processing as a loss leader for our review efforts or vice versa," says Brownstone. "Both kinds of work are profitable."

CONTRACT ATTORNEYS

By all accounts, however, the real money in e-discovery is in the review stage, where all the material culled during processing has to be analyzed for responsiveness and privilege. "About 70 percent of [e-discovery] costs are right there," says Browning Marean, a litigation partner at DLA Piper. It's little surprise: This is the stage, after all, where lawyers traditionally roll up their sleeves -- and bill away. But the burgeoning volume of e-data has put pressure on firms to reel in costs. "The rule of thumb is that 1GB of data equals 75,000 pages," says Marean. "If a review team can go through 200 pages an hour, do the math. Clients are doing the math. They're appalled."

To appease clients, firms have taken three approaches: keep staffing matters with associates, but leverage technology to cull the document sets; hire lower-cost contract (a.k.a. "project") attorneys to do review; or outsource review to a vendor (and its own team of lawyers).

It's hard to say which of the first two techniques is more profitable. An associate at a large firm might be billed out at $250 per hour, while a contract attorney, if billed as a pass-through, might run $50. Yet contract attorneys are often billed above cost, and the surcharge can vary wildly, even from client to client. Firms tend to put large numbers of contract attorneys on cases, too, so while associates may be more profitable per head, contract attorneys may be more profitable as a group.

Indeed, two firms that are seeing profits from e-discovery-Fenwick and Kirkpatrick & Lockhart Preston Gates Ellis-have adopted polar opposite models. Fenwick uses associates -- from a pool of up to 150 -- for almost all cases, using sophisticated software tools to limit the number of associates needed. Cases in which hundreds of gigabytes of raw data were collected require six to 12 lawyers for review, according to Fenwick's Brownstone: "We apply radical de-duplication and statistical analysis to keep what the reviewers see to the absolute minimum of data." Kirkpatrick, on the other hand, staffs no associates on review. Indeed, its document analysis technology group doesn't even have associates. Instead, the firm draws on a pool of some 200 in-house project attorneys who do nothing but review documents.

It's the third approach -- using a review vendor -- where firms have the greatest difficulty. For one thing, there isn't money to be made here because it's typically the client, looking to cut its e-discovery costs, who has contracted with the vendor (who can charge anywhere between $30 and $80 per hour for its reviewers, according to Socha). Pfizer Inc, for example, has seen savings of 40 to 50 percent by using its own vendor -- Huron Consulting Group -- instead of a traditional law firm review, according to Laura Kibbe, senior corporate counsel and managing director of the company's discovery response team.

Using an outside vendor also puts a strain on the firm's ability to oversee review and to make sure it's proceeding as quickly and accurately as possible. That's a problem, because mistakes -- as Kirkland & Ellis discovered -- can cost cases, and clients. But a novel solution, at least in legal circles, can be found at Sullivan & Cromwell, where Barnett hired six project managers, nonlawyers who steer the firm's cases through every phase of e-discovery. "Project managers are not a concept normally thought about at a law firm," he says, "but discovery is like a complex engineering problem -- it's detailed, repetitive -- and has to be handled as such." Barnett's managers all have a technical background, but they also have experience with user support and e-data. "They manage vendors and see problems," he says. "They keep documentation to make sure that everyone is doing the job, and keep the wheels turning."

Of course, given the speed with which e-data is growing, it's just a matter of time before any human review is going to be too costly. The only long-term solution is to embrace technology, and find better ways to whittle down huge document sets -- or, as in Morrison & Foerster's case, replace people with machines.

Keyword searching is just the beginning. Newer tools like Attenex and Recommind, which can categorize, or cluster, documents by type, are enormous time- and money-savers because they can quickly let firms see which batches can be reviewed by paralegals and which need to be examined by senior attorneys. Other programs eliminate not just duplicate records, but near-duplicate ones, too. And systems that can learn as they analyze data, figuring out when a keyword is being used in a significant way and when its appearance is meaningless, are in development.

But not every firm has a tech-savvy client base, ready to unleash the fancy software -- or to pay for it. "The reluctance is understandable; it's an up-front expense," says Mark Brennan, counsel at Bryan Cave. "But these are sophisticated technologies that ultimately reduce costs."

PROFIT CENTER?

Finally, there is arguably the thorniest question of all: Once firms figure out how to do e-discovery efficiently, then what? Run it as a business, generating revenue for the firm? Or treat it like photocopying, a pass-through to clients? Here firms have adopted not just different strategies, but different philosophies.

At Kirkpatrick & Lockhart, the firm's document analysis technology group -- now numbering seven partners and 17 staff attorneys who act as liaisons with the project lawyers -- was, "right from the beginning, a revenue stream for the firm," says Helen Bergman Moure, a partner in the group. Moure won't disclose revenues, but notes that it is "not a money-losing operation" and that the bulk of work comes from helping other law firms with e-discovery work. Meanwhile, Fenwick & West is considering spinning off its own e-discovery group. "It's on the table," says Brownstone. "If we were our own entity, we wouldn't be bound by the firm's rigid conflict rules, and we would be able to take on more clients."

But other firms, including some of the most proactive players in the e-discovery space, think that a spin-off is unwise. "Yes, it can be a huge income stream, in the short term," says Morrison & Foerster's Benamram. "But clients are happier if you keep costs to a minimum, and happy clients are what give a law firm long-term success." Neither Bryan Cave nor Sullivan & Cromwell is looking to turn e-discovery into a profit center, and while DLA Piper once toyed with the idea of a spin-off, it has since reconsidered, wanting to keep any tips and tricks it has learned close to the vest. "Other firms are our competition," says DLA's Marean, "and [if keeping this knowledge in-house] gives us a competitive advantage, to heck with them."