Inter Partes Review and Limited Discovery: Applying the Garmin Factors in Practice

Parties coming to an inter partes review from district court often expect broad discovery. They will not get it. The statute allows depositions of declarants and otherwise only discovery “necessary in the interest of justice” (35 U.S.C. § 316(a)(5)), and the Board measures that standard with five factors first set out in Garmin International, Inc. v. Cuozzo Speed Technologies LLC, IPR2012-00001, Paper 26 (March 5, 2013).

The case note on Paper 26 explains what the Board decided there. This page takes a different angle: how the factors work as a practical test, and what a motion for additional discovery generally needs to show to have a chance.

Where additional discovery fits

IPR discovery comes in three layers under 37 C.F.R. § 42.51:

• Mandatory initial disclosures, by agreement or on motion (§ 42.51(a)).
• Routine discovery (§ 42.51(b)(1)): exhibits cited in a paper, cross-examination of declarants, and relevant information inconsistent with a position the producing party has taken. No authorization is needed.
• Additional discovery (§ 42.51(b)(2)): anything else. The parties may agree to it. If they cannot agree, the requester must move and show the discovery is in the interest of justice.

A motion needs prior Board authorization (37 C.F.R. § 42.20(b)), usually obtained on a conference call. In Garmin, the Board described the factors on that call, before the motion was filed. Parties should expect the panel to raise them early.

Post-grant review is different: § 42.224 uses a good-cause standard limited to evidence directly related to factual assertions advanced by either party. See discovery in post-grant review versus inter partes review.

The factors as a checklist

Factor	What the Board looks for	How it played out in Garmin
1. More than a possibility	Evidence already in hand tending to show, beyond speculation, that something favorable to the requester will be found	No threshold evidence of copying, nexus, failure of others or long-felt need
2. Not litigation positions	Requests aimed at facts, not at the opponent’s arguments or trial evidence	A request for documents the petitioner meant to rely on was refused
3. Not obtainable otherwise	An explanation of why the requester cannot generate the information itself	The patent owner could run its own market survey and obtain file histories
4. Understandable instructions	Short, clear instructions	Two-page instructions passed
5. Not overly burdensome	Narrow requests tied to a genuine need, with cost and schedule in mind	Cost estimates of $52,500–$82,500 and an all-documents-considered request weighed against the motion

Factor 1 does most of the work

Most motions that fail, fail here. Two points from Garmin are often overlooked:

• “Useful” means favorable, not merely relevant or admissible. Restating the factor as “likely to yield relevant, admissible evidence,” as the movant did in Garmin, applies a district court test the Board rejected.
• The showing comes first. The requester should already have some evidence or reasoning that points toward the result it hopes to prove. A belief that the documents probably exist is not enough.

For secondary considerations this usually means a threshold showing of nexus. In Microsoft Corp. v. Proxyconn, Inc., IPR2012-00026, Paper 32, the Board relied on Garmin to deny discovery of sales data for a complex product with thousands of features. The patent owner had shown no connection between the accused feature and the product’s sales. See the Proxyconn discovery note.

What a well-supported motion tends to include

Drawing the factors together, a motion the Board takes seriously usually:

• identifies each request precisely, with a defined subject and time period;
• attaches the threshold evidence for each request and explains what favorable fact it points to;
• ties each request to an issue in the trial, such as an instituted ground or an asserted secondary consideration;
• explains why public sources, the requester’s own records or expert analysis cannot supply the information;
• keeps instructions short and avoids catch-all requests;
• addresses burden and schedule, often with a proposal that narrows the request; and
• is raised promptly, given the one-year deadline in § 316(a)(11). See discovery and the one-year time period.

A request for a corporate deposition on the same topics as written requests will usually share their fate, as it did in Garmin.

Common pitfalls

• Labeling broad requests as routine discovery. “Inconsistent information” under § 42.51(b)(1)(iii) means specific information the producing party knows contradicts its position, not a category to search.
• Seeking the opponent’s evidence before it is due. The Board’s schedule determines when each side presents its case, and cross-examination of the other side’s declarants is the usual check.
• Asking for information about real parties in interest or privies without explaining how it bears on a live issue.
• Treating the factors as boxes to tick. The Board weighs them together and with the case schedule in mind.

The factors have been applied in many later decisions, and the Trial Practice Guide discusses them. Practice can shift over time, so readers should check current USPTO guidance.

Key takeaways

• Additional discovery in IPR requires a showing that it is necessary in the interest of justice, assessed under the five Garmin factors.
• The first factor is decisive in most cases: the requester needs evidence, beyond speculation, that the discovery will turn up something favorable.
• Requests should be narrow, clearly worded, tied to trial issues and limited to information the requester cannot get itself.
• PGR uses a different good-cause standard under § 42.224.