Predictive Coding – Not for All Occasions

Several months ago, Vice Chancellor J. Travis Laster of the Delaware Court of Chancery, in EORHB, Inc. v. HOA Holdings, LLC, No. 7409-VCL, 2013 WL 1960621 (Del. Ch. May 6, 2013), issued a wide ranging order for the parties to cooperate. This included the selection of a single vendor to host both parties records, and instructing the parties to use predictive coding. On May 6, a new order was issued, allowing the defendants to retain a different vendor from the plaintiffs, and conceding that predictive coding was not necessary. As the number of relevant records was expected to be very small, the cost of using the technology outweighed any benefit.

The legal world tends to be slow in adopting technology. However, once a new technology is adopted, it seems to become a de facto standard, with advocates insisting that it be used in every case. It is important to remember that any technology is just one tool in the litigator’s toolbox, and just as a builder wouldn’t use a sledge hammer on a nail, some types of legal technology are just not suitable in certain situations.

The case above is a perfect example. The defendants rightly determined that, because the richness of the collection was very low, the effort to train a predictive coding system would be excessive. Other types of technology, such as conceptual clustering or even “archaic” keyword searching may return better, more cost effective results.

​Before jumping on the newest technological bandwagon, take a step back and determine if the selected legal tool is the best one for the job.