Time For Document Review 2.0

In an incredibly short amount of time—roughly half a decade—e-discovery has risen from a niche field within the legal and forensics spaces to become a major disruptive force within not just the legal space but amongst large enterprises as well. Equally axiomatic is the skyrocketing cost of e-discovery: while the cost per gigabyte (GB) of data to be collected, processed, reviewed, analyzed and produced has dropped from more than $5,000/GB to less than $2,500 in many cases, the amount of data subject to this process has exploded, more than making up for this reduction in variable costs. The inevitable result is exasperated clients who are looking to reduce costs by deploying more technology in-house while simultaneously forcing their outside counsel to become far more efficient by also using more modern tools.

But rather than burying their heads in the sand, proactive clients and outside counsel are effectively addressing this issue by focusing their attention, efforts and budgets on the most egregious cost drivers. An often overlooked element of e-discovery is that fully 70% of all e-discovery costs are generated at a single stage—document review—which requires the time of expensive attorneys (which is unavoidable) almost always using outdated tools (which is avoidable). Even a small improvement in the efficiency of reviewers can have a tremendous impact on all areas of e-discovery, including cost, speed, accuracy and ultimately the odds of winning a particular case. These same proactive players are finding that a compelling tool in the fight against ever-increasing e-discovery costs is one which helps lawyers review documents accurately and consistently. In the fast-moving world of e-discovery, these players find themselves enjoying a huge strategic advantage.

Brief Overview of the E-Discovery Process
Best illustrated by the EDRM Project’s (www.edrm.net) reference model, e-discovery has a number of distinct phases, each of which describes the steps which must be taken with electronically stored information (or ESI) from the outset of litigation all the way through trial. These steps includeinformation management, identification, preservation, collection, processing, review, analysis, production and presentation. Perhaps not surprisingly, most of these stages—basically from information management through processing plus production and presentation as well—require little to no attorney time, as these functions can and often must be handled by more technologically savvy IT personnel with minimal legal oversight. While IT personnel are not inexpensive, they are far less expensive than their attorney counterparts and are also far more comfortable automating as many processes as possible, which helps keep the costs from these stages relatively low. The net result of these circumstances is that the majority of e-discovery stages (essentially all stages except review) represent just a fraction—roughly 30%—of the cost of the entire e-discovery spectrum. (For clarity, review is the step at which attorneys literally read each document to discern the document’s meaning and importance to the case, which allows them to appropriately "tag" or code each document along several parameters like privilege, responsiveness, relationship to issues, etc.). Simply put, attorneys are expensive, so it stands to reason that a process requiring a significant amount of their time would similarly be quite costly, representing fully 70% of the cost of the entire e-discovery spectrum.

Review 1.0

Exacerbating the situation is a phenomenon we’ll call "review 1.0," which is simple to explain but lethal to clients and outside counsel trying to keep e-discovery costs under control: today’s review attorneys use processes and tools that were created in yesteryear’s paper-driven environment to review today’s electronic documents in a digital world. The result is predictable: not only are e-discovery bills skyrocketing, but the percentage of these bills being driven by the review stage is actually going up, not down. The core challenge represented by the review stage is that we are all creating more and more data, much of which must be eyeballed by a competent attorney. No amount of deduplication or email threading will change this fact.

The ramifications of review 1.0’s overly manual, time-consuming and inefficient nature are numerous and significant. Using review 1.0 tools, a typical review process requires several "passes" (i.e. instances of looking at each document for different reasons) through a single document collection—which is highly inefficient and extremely expensive, especially with large collections of millions of documents. To make matters worse, with each pass at each document attorneys must "recreate the wheel" by re-reading significant portions of the document to (again) discern what each document is about in order to conclude its meaning and value for whatever the typical pass involves. In other words, reviewers must recreate context for each document every time the document is reviewed. Additionally, review 1.0 tools require attorneys to individually code each document on each and every relevant parameter time and again.