Migrated from eDJGroupInc.com. Author: Barry Murphy. Published: 2010-11-15 10:47:36Format, images and links may no longer function correctly. Our earlier post on predictive coding potentially being the future of document review included a poll on whether or not predictive coding is defensible. The results so far are somewhat surprising in that the majority of folks believe predictive coding is not defensible.
I suppose I shouldn’t be surprised; the legal industry is slow to change (and for good reason – it makes sense to proceed with caution when going into unchartered territory). When I first came into the eDiscovery space wondering why more organizations weren’t adopting technology, I was constantly reminded that most lawyers hadn’t yet migrated off WordPerfect. In fact, that argument was used so many times that I came to despise it.
Yes, the surprise quickly wore off and I can understand why most believe that predictive coding is not defensible. It hasn’t seen the battle test of case law and there’s nothing like precedence to make the legal community feel more comfortable about certain practices. One could argue that the lack of case law around predictive coding is actually a tacit acceptance of it because it is not being challenged. However, the poll results seem to challenge that line of thinking.
Is this bad news for predictive coding? I don’t think it is. That predictive coding will become commonplace is inevitable. This poll simply suggests that adoption will occur slowly. And history supports the move to reliance on analytics. Several years ago, when I first began as an enterprise content management (ECM) analyst at Forrester Research, I covered document imaging. One of the core components of document imaging is optical/intelligent character recognition (OCR/ICR). When OCR first hit the market, the accuracy rates were low – less than 50% at times. Most in the industry scoffed at anyone that believed the claims that OCR could reach 99% accuracy. But, with the advent of multiple OCR voting engines and algorithms, OCR rates quickly reached the 99.9% accuracy levels necessary for broad adoption. Clearly, it’s more expensive to get that accuracy level, but the technology can get there.
The same pattern should emerge with predictive coding. The technology will get better and better. There will be challenges and precedents set. The cost savings and review efficiency will become impossible to ignore. It might take 5-10 years for broad adoption, but we will get there. In the meantime, predictive coding will still have a home today as a core activity used in ECA initiatives aimed at quickly getting to the heart of a given matter.