Blog

Efforts to Control Discovery Costs Can Still “Implode” Even When Parties Adopt a Model Discovery Order

Why Do Bad Things Happen to Good Litigants? Efforts to Control Discovery Costs Can Still “Implode” Even When Parties Adopt a Model Discovery Order

In a recent patent infringement case in the Northern District of California, MediaTek v. Freescale Semiconductor, the parties attempted to rein in the costs and burden of discovery by doing exactly what they’ve been instructed. They entered into an agreed-upon discovery order, similar to the Federal Circuit Advisory Council’s Model E-Discovery Order, which contained limits on the scope and quantity of discovery. But as Magistrate Judge Corley observed when faced with more than a dozen letters and motions regarding discovery disputes, “the parties’ laudable efforts at controlling discovery costs have, unfortunately, imploded.”

So what happened? Well, as my colleague Matt Miller observed in several of his past blog posts, in theory the use of model discovery orders can reasonably limit discovery in patent litigation. In practice, however, it gets much more complicated than simply agreeing to a discovery order that places quantitative limits on discovery. As the old cliché goes, the devil is in the details. Here are a few lessons learned about those details from MediaTek:

  • Don’t agree to time limits on depositions without also reaching agreement on how the time is to be calculated. In MediaTek, the parties agreed to limit the number of hours of Rule 30(b)(6) deposition testimony, but they did not specify in their agreement how the time would be calculated. Not surprisingly, the litigants ended up fighting over whether the time limit had been reached, and needed the court to resolve the dispute.
  • Don’t agree to limit the number of written discovery requests without also clearly specifying how to count each separate request. The parties in MediaTek agreed to a numerical limit on interrogatories, “including all discrete subparts.” But what’s a “discrete subpart?” I’m not sure I know, and the litigants didn’t, either. So they ended up in motion practice, asking the court how to count interrogatories.
  • Don’t agree to a certain number of search terms without addressing how terms are to be counted, including how to count of variants of a term. If a party agrees to a limited number of search terms, and one of those terms is “patent,” are “patented” and “patentable” counted as additional terms? Or merely as variants of one term? What about the term “WLAN” — would searches for “Wi-Fi” and “wireless” be considered variants, or separate terms? This issue tripped up the parties in MediaTek, requiring court intervention.
  • Better yet, don’t agree to a numerical limit on search terms. Instead, reach agreement on a smart, statistically tested and validated set of terms. Frankly, I view the idea of a numerical limit on search terms as foolish, and very unlikely to result in reasonable, sensible constraints on discovery. What’s the point of agreeing to seven search terms per custodian, as the parties did in MediaTek, if running those terms returns a mess of non-responsive documents, and leaves behind even more responsive documents?  And what does a party gain by objecting to proposed terms as “overly broad,” if it “has not run a test search on a single identified custodian for any of the proposed searches,” as in MediaTek? If parties intend to use search terms, they should do so smartly. Engage in iterative discussions about what the right terms are based on the matters at issue in the case, regardless of how many terms it takes. Test along the way to be sure the terms are as precise as possible (not returning too much junk) but deliver reasonably high recall (not leaving too much behind). Whether the issues in a given case result in a search containing seven terms or 70 terms is not the point; the point is finding the relevant documents efficiently and cost-effectively. And as Magistrate Judge Corley observed, “the process is designed to be collaborative,” something the litigants before her had not done.

In the end, the parties in MediaTek wasted countless hours and dollars on unnecessary — nay, even silly — discovery disputes. Although they started down the right path by reaching an agreement on the scope of discovery, they strayed from that path by neglecting to consider some important details about their agreement. More importantly, they wandered completely off the path by failing to collaborate and cooperate around significant discovery issues.

Maureen O'Neill