- The defendant had conducted custodian interviews that it used to produce 39,000 pages of highly relevant documents on the disputed issues;
- But the plaintiff demanded that the defendant also collect electronically stored information (ESI) from 69 custodians and their assistants, with the review limited only by 90 search strings, many of which contained “OR” Boolean connectors that took the effective number of search terms well past 100;
- When the court rejected the plaintiff’s initial ESI proposal and limited him to 10 custodians and search terms achieving an 85 percent responsiveness rate, the plaintiff mostly rejected the defendant’s suggested custodians and proposed 803 search terms (counting terms with the “OR” Boolean connector as multiples) that returned 468,595 documents (with families), while yielding an estimated responsiveness rate of about 7.8 percent;
- When the court instructed the plaintiff to craft new search terms, this time limiting him to 25 terms targeted to each custodian, the plaintiff proposed terms that dropped the responsiveness rate even further, to an estimated cross-custodian average of 5.1 percent;
- When the defendant proposed replacing search-based discovery with TAR, the plaintiff insisted on an 80 percent recall rate rather than the 65 percent recall rate proposed by the defendant;
- The defendant refused to pay the $40,000 estimated cost of reviewing the so-called “residual TAR documents”;
- The final responsiveness rate for the completed TAR review was only 3.3 percent; and
- The plaintiff could not show that the TAR process had yielded any documents of significance.
The court, of course, also considered the importance of the issues, the amount in controversy, and the parties’ resources. And its evaluation of those issues was no less interesting. The court found that the parties’ commercial dispute over a restrictive covenant was just not that significant. Though it recognized that $600,000 in estimated TAR expenses paled in comparison to the estimated $39–53 million in controversy, it also held that the monetary stakes were only one factor and could not, by themselves, justify speculative discovery. And it refused to give any weight to the corporate defendants’ significant resources, finding that the defendants’ operations had been affected by (among other things) the COVID-19 pandemic and, in any event, the plaintiff was himself well-resourced and had outside litigation funding.
What remains to be seen is if this is a one-off decision punishing a particularly unreasonable litigant or the start of a broader trend that reins in disproportionate discovery. The former would be just a footnote, whereas the latter would make a meaningful difference in the time and expense associated with civil litigation.
