The ECA Black Box

I read another great, and somewhat controversial, article in LTN last month – “Riding the Waves of Early Case Assessment” by George Rudoy.  Our old friend ECA returns! The last time I wrote on the topic of ECA appears to be December 2009 – “Early Case Assessment – What is it really? 

“Early Case Assessment (“ECA”) is an overused term that many in the eDiscovery business have adopted as a buzz word to define a process that is a combination of technology and professional expertise. It gets thrown about a lot, often by those that sometimes don’t know what it really means.”

Not much has changed since 2009 in terms of how ECA continues to be associated with technology features. Software solutions designed for early case assessment are widely available, but are vastly misunderstood and often underutilized, or worse, utilized with too much unverified reliance. The number of advanced technology choices out there is mindboggling, even for those of us who keep up with them. This has and will continue to cause considerable confusion in the market. ECA is the process of doing exactly what it says – determining what your case is about early. How early? As early as is humanly possible, that’s how early. Usually we need to know yesterday, right? Data reduction and “review cost avoidance” are pleasant bi-products of a well thought out ECA workflow. Technology is just a tool that helps us find and organize documents and information more efficiently. Speaking for myself, the more advanced the technology, the better. “Finding the truth and then putting your own spin on it” has never been faster – with the right workflow. T. Boone Pickens once said - “an idiot with a plan is better than a genius with none”. The same holds true with technology. The greatest technology on the planet is wasted at best with the wrong or no plan. The greatest problem we have today, however, is not a lack of technology availability. The problem is not just a lack of a deep understanding of that technology, although certainly part of the problem. The real problem is a lack of desire to understand. Lawyers want to see the documents that tell the story. They don’t want to stand in front of a judge and explain a “black box”. George Rudoy hit the proverbial nail on the head:

"Considering the risks at hand, this could be a significant step forward -- to be able to determine your risk exposure early on in a case, based on what human beings are seeing in the language rather than point to a "black box" purchased off the shelf. Clearly, ECA capabilities both pre- and post-collection continue to evolve.”

 Predictive coding, anagrams, Optimized Distributed Search ("ODS") technology or whatever new conceptual algorithm comes next are all very useful tools. I personally love black boxes! We should use them. We should not, however, expect our lawyers to understand and certainly not explain how that advanced technology works. Sound data reduction techniques based upon a simple human based methodologies that arrive at a set of validated terms, for example, is all that is needed at the end of the day. How information was organized and/or classified using technology to arrive at a validated set of terms is inconsequential with a properly documented methodology. A methodology that may well use conceptual clustering, searching, predictive coding and other advanced features. Again, just tools essential to efficiently organizing and analyzing information. Ultimately, however, a methodology that relies upon on a human call and tag rather than an algorithm is much more likely to be understood and followed. Advanced technology helps us get where we want to go faster and more efficiently. ECA - knowing what happened when, to whom and by whom - is a pleasant bi-product of a well designed and easy to understand and explain data reduction methodology.

Technology is not the silver bullet by itself and can even be dangerous in the wrong hands. Any ECA and data reduction workflow should not only confidently reduce the volume of information that ultimately needs review. It should also increase the relevant content that is ultimately in need of review. Be careful out there!