When we lost access to a server with all of our blog posts, it gave us the opportunity to re-post archival material and look back on it from today’s perspective. This was one of our funnier marketing events as we actually held a protest at Legal Tech about the lack of acceptance of predictive coding. We were written up in LTN so to some extent it was a success. The bigger success came later in the year when the material we were compiling and workflows we were advocating for, were actually used in the Global Aerospace case.
Today we are less radical and spend some of our free time education lawyers on how to use and validate predictive coding results in the Predictive Coding Thought Leadership Series tour traveling around the country. Enjoy this humorous post from 2012….
Originally posted on January 27, 2012
As some of you know, I have been talking to over a dozen predictive coding vendors to get them to support an “occupy legal tech” movement. While the support was universal, getting competitors to band together in a common industry statement proved to be as politically challenging as organizing a TREC study or Electronic Discovery Institute Study. Though I must say that one vendor (OrcaTec) volunteered to put a sign supporting the protest at their booth.
So why am I launching this movement? To some extent I have been protesting since 2010 when I started educating judges on the value of technology assisted review. This was in response to the obstacles I encountered trying to sell similar technology to law firms. Everyone likes a new gadget, but lawyers don’t like the risk of wondering if a judge would approve of the use of these tools. This same tension was reported in the Electronic Discovery Institute’s Survey on Predictive Coding released on October 1, 2010 which cited 8 whitepapers from vendors that showed average savings of between 40 to 80 percent when compared to linear review. These vendors overwhelmingly reported that lawyers and their clients are most concerned about defensibility and ranked it as the number 1 obstacle to broader acceptance. While there are now more predictive coding options out there than are listed in the survey, this perceived defensibility obstacle still remains.
These defensibility concerns stem largely from a lack of education on the tools but there are also economic concerns that more technology means less lawyer time. As a result, if everyone agrees on both sides to use key words, then it is simpler to just continue doing what we have always done. That type of mindset is starting to crack but it is the majority view as despite substantial empirical evidence that ALL of these tools ALL work better than linear review, so why aren’t we using these tools on every large case? That is why a protest makes sense to me and why I have organized an ad hoc, non violent, no tents, business attired protest. Admittedly, this protest will probably be less fun than the range of Occupy Protests that have sprouted up over the country, but being in the 1%, I don’t deal very well with the cold now and the thought of being hauled off in handcuffs is not on my personal bucket list. However, I will be wearing some buttons describing the protest which are inserted below and will be handing them out in a limited supply if you promise to wear them during LegalTech.
I also want to clarify one other misperception I might create above that all the predictive coding tools work well. I stand by that statement that compared to linear review they are all better if appropriately used and can be validated statistically to test their results. But I don’t want to even suggest that all the tools work equally well when compared to each other. There are differences in approaches, workflows, algorithms, price points and capabilities. So I would suggest taking the time to become educated in these differences once you move over to the “protest” side and recognize that these tools do work better.
I have more data to support these positions available for review on my IPAD and am happy to convert visitors into protesters based on a few power point slides. Just email me at firstname.lastname@example.org. See you at LegalTech.