Buddy, Can You Spare a Dime?
by Dennis Kiker, Esq., Director, Professional Services, Fios, Inc.
Cost containment continued to be the most recurrent theme on day two of the Georgetown Advanced E-Discovery Institute. And by a pretty wide margin (with no disrespect to any of the other excellent speakers), Ralph Losey and Jason Baron’s presentation on advanced search had to be the most entertaining of the day. The message was simple: the volume of data created and stored on computers continues to increase exponentially, and traditional methods of identifying, collecting and reviewing that data are becoming impractical. Indeed, they are impractical already in many cases.
As a result of the work being done by TREC Legal Track, the eDiscovery Institute and others, a number of truths that many of us have known intuitively are becoming manifest in data. Traditional, linear review, the tried and trusted “attorneys-eyes-on-the-page” method, just doesn’t work. Statistical analysis reveals that our legions of highly-educated and highly-compensated attorneys can find about one in five relevant documents in a large collection. Of course, all of us who’ve been involved with any major document production knew this already. Who, two weeks into a several month paper review project, has not sighed to him or herself, and wished we could just start over knowing what we’d learned in those initial weeks? Or wondered, as we tagged a document as potentially privileged, whether that wasn’t very similar to a document we saw last week, and whether we tagged it then? In the context of the large production, the effectiveness of traditional review has not changed, it has only been exposed. But we’ve always known what Ralph Losey said today, “We can’t afford the whole truth.” Beyond some indefinite threshold, less than perfect is the very best we can hope for.
On the other hand, the degree of perfection achievable is beginning to change. The TREC Legal Track study suggests that by combining available methods such as Boolean search, concept clustering, etc., we can improve our 20% hit rate to as high as 78%. That is remarkable. Combine that fact with another that becomes apparent from presentations like this one, and it is obvious that our traditional document review practices are untenable. Consider what would not be that unusual a case today, and certainly not in the near future: 1 million documents. Assume further that you have 20 attorneys reviewing those documents at a rate of 50 documents per hour – not atypical numbers for most review shops to quote. It will take that crew about 125 days to complete the review. Now let’s scale that scenario just a bit. Assume instead 10 million documents. The math is pretty simple, but the results are disturbing. The same 20 lawyers now need nearly 3.5 years to finish the job. Let’s not even discuss the cost. And, when they pack up and go home, they will have left 80% of the relevant information behind.
When the facts are so clear and the evidence undeniable, why then have most of us not rushed to embrace new search and retrieval technologies and methodologies? It can’t just be risk aversion. The statistics tell us that the risk of missing a relevant document or inadvertently disclosing a privileged one would be much lower using currently available technologies. So what is it? I think it is, in part, a matter of choice – meaning, it is too difficult to choose an approach and run with it. Which of the different technologies and methodologies currently available is the best one on which to invest your money and reputation? I certainly do not criticize the majority that is sticking with keyword searches and linear review for the time being. It is a tough thing to walk up to the ledge and jump, even though the studies show that it is not merely a leap of faith. A great deal is at stake, even in the smallest cases. Our reputations, our licenses, our livelihoods. What we really need are more evangelists amongst those that have already taken that leap, to give credibility and build confidence in the new technologies. We need some brave souls to sit down at a Rule 26(f) conference and agree upon the fact that the numbers don’t lie. That there is a better way to get to the truth without wasting resources turning virtual pages that could be better spent on resolution. Until, then, companies will continue to be frustrated with escalating discovery costs and reduced resources with which to cover them. Buddy, can you spare a dime?
Filed under Discerning e-Discovery, Uncategorized.