When does the cost of using predictive coding assistance outweigh any practical benefit of its use?

By Ian Bagger, Esq.

There is a lot of talk about predictive coding and other technology-assisted  (“TAR”)  review tools.  Oracles on many prominent e-discovery blogs have generated countless articles singing the praises or questioning the validity of such technologies, and there is a small industry devoted to reading the judicial tea-leaves regarding TAR.    Overall, the trend has been toward heightened judicial awareness and acceptance – even in some cases endorsement – of such tools.  But despite the optimistic pronouncements of e-discovery seers, nobody expected the Delaware Court of Chancery’s order on October 15, 2012.  Legal Expenditures

In EORHB, Inc. v. HOA Holdings, LLC, a contract dispute arising out of the purchase and sale of the Hooters restaurant chain, the parties appeared before Vice Chancellor J. Travis Laster of the Delaware Court of Chancery to argue a motion for summary judgment and dismissal of counterclaims.  The Court disposed of the motion by granting partial summary judgment. Having done so, the Court then – without prompting from either party, and with no record of dispute as to discovery – ordered the parties to show cause why they should not jointly select and share a predictive coding vendor.   

Said the Court:  

This seems to me to be an ideal non-expedited case in which the parties would benefit from using predictive coding. I would like you all, if you do not want to use predictive coding, to show cause why this is not a case where predictive coding is the way to go.

Later, on May 6 of this year, Vice Chancellor Laster entered a terse order indicating good cause had been shown, and the parties need not jointly select and share a predictive coding vendor after all.  EORHB, Inc. v. HOA Holdings, LLC, No. 7409-VCL, 2013 WL 1960621 (Del. Ch. May 6, 2013).  

What happened in the intervening five-and-a-half months?  The record is slim, but there are several reasons the Court could have overruled itself.

The order does not linger over its analysis of what “good cause” constituted in this case, but the meager facts recited can give litigants some guidance.  The Court’s order does acknowledge that the plaintiffs faced costs if they adopted predictive coding technology, and expected very little value from its application.  Defendants were more willing to adopt such technology. Specifically, the plaintiffs avoided the expense by doing a cost-benefit analysis, described here by the Court:

WHEREAS, the parties have agreed that, based on the low volume of relevant documents expected to be produced in discovery by [the plaintiffs] the cost of using predictive coding assistance would likely be outweighed by any practical benefit of its use[.] (Emphasis added.)

The court may have been demonstrating an understanding of one of the primary rationales for TAR.  One of the main justifications – indeed the impetus for the development of such technologies – is the fact that data sets are getting bigger every day.  The Rand Institute for Civil Justice report on e-discovery expenditure teaches that review is the single largest cost for litigants.  See Where The Money Goes: Understanding Litigant Expenditures for Producing Electronic Discovery, Pace, Nicolas M. and Zakaras, Laura; Rand Corporation, 2012.  Much of the Rand  report addresses how technologies can and should be used to help characterize very large document sets and reduce review cost.  At the same time, however, vendors of such technologies – such as kCura’s Relativity Assisted Review – will tell you that the technology is not intended for use on small bodies of data.  

In this case, the plaintiffs contended that the volume of data they expected to produce would likely be quite small.    Very small data sets are inappropriate for use of these technologies. Thus, the court could have adopted the plaintiff’s alleged  cause for not using TAR tools.  If the plaintiff’s cost-benefit analysis led to the conclusion that TAR would be more costly than other defensible techniques for arriving at the same result, then TAR should not be employed.

Alternatively, the Court may have been unusually open to any argument that would allow it to save face, having made an order that would have substantially altered the time-honored tradition (and case law) governing how discovery takes place in American courts.  Traditionally, the litigants manage discovery themselves, and the power of the court to manage discovery is only invoked when the parties cannot resolve disputes themselves, or when a party is shirking its discovery obligations.   

Vice Chancellor Laster’s oral order at the tail end of the hearing in October crossed an unusual line in addressing a matter not properly before the Court – indeed, one that was not in dispute at the time.  Moreover, that order would have imposed a discovery regime very much at odds with common practice.   By entering its May 6 order, the court restored the traditional balance of responsibilities in management of discovery, placing it back in the hands of the parties.







Leave a Comment

You must be logged in to post a comment.