E-Discovery in Intellectual Property Patent Litigation (Part 4 of 5)

E-Discovery in Intellectual Property Patent Litigation: Changing The Tune

In this fourth post of the five-part series on e-discovery in intellectual property patent litigation , Daniel Garrie analyzes the Federal Circuit’s model order regarding e-discovery in patent cases. [1]  In addition, the series briefly describes the purpose behind the model order, describes its key provisions, and analyzes the model order to identify some areas of continuing concern.


The Model Order Only Allows Email Production to Occur After the Parties Have Exchanged Initial Disclosures of Basic Documents and Information on the Critical Systems Storing The Email

The model order attempts to force the parties to hold off on email production until after initial disclosures regarding the patents, the prior art and relevant financial information.[16] However, to encourage focused and reasonable email production, it is respectfully suggested that the model order also should require the parties to define their respective technology systems involved with email. This information is critical to allowing the parties to draft email requests that are reasonable and narrowly tailored, as required by the model order.[17]

For example, a party might craft a request for email that is narrowly tailored and appears reasonable,[18] but that request still could be unreasonable if the party seeks email that is five years old and is only stored on disk backup in Germany.

In this example, the cost of production given the medium and location makes an apparently narrow and reasonable request unreasonable in practice, and may require an even more refined request. The parties should be required to identify and disclose their respective technology systems involved with email, so that such issues may be identified before email requests are issued. One possible solution is for the model order to be amended to require the parties to exchange information about their IT systems at the earliest stage of the litigation, enabling both sides to effectively organize their forthcoming search requests.

The Model Order Should Consider Requiring the Parties to Perform Email Sampling Before Limiting the Number of Search Terms and Custodians to Five People and Terms

The model order presumptively limits the number of custodians and search terms for all email production requests to five terms and custodians per producing party for all such requests.[19] The intent is to control the exorbitant costs of production by minimizing what parties can request.[20] While well intentioned, this presumptive limit presents a challenging paradigm, because it is impossible for parties to be 100-percent accurate on terms and custodians — especially when they do not control the data. Consequently, it is our belief that, prior to the court or parties selecting terms or custodians, they should apply common-sense:

1. Both parties should group search terms into high, medium, low value.
2. The parties should then take each group of search terms and identify applicable time frames and custodians. For example:

High Group

Dates: 02/2010 to 05/2011; 03/2005 to 04/2006
Custodians: D. Smith; M. Jane
Terms: Apple, Democrat, Republican, Libertarian
3. The opposing party should then sample each of the custodians using the search terms and dates for the group.
4. Re-order the terms and custodians.

Of course, the court should mandate the application of the model order’s strict number requirements, should the parties fail to mutually agree on a protocol, or if the terms the parties propose are inappropriate or indiscriminate in nature. In such circumstances, paragraph 11 of the model order provides for cost-shifting to the requesting party.

[14] Model Order at ¶ 5

[15] See D. Garrie & Y. Griver, “Dispute Resolution and e-Discovery,” Thomson Reuters (2012), at §2.3(B) and fn 15.

[16] Model Order at ¶ 8.

[17] See Model Order at ¶ 6 (“To obtain email parties must propound specific email production requests”) and ¶ 7 (“Email production requests shall only be propounded for specific issues, rather than general discovery of a product or business”).

[18] See, e.g., McGrath v. United States, No. 11–318C (Fed. Cl. March 6, 2012). In McGrath, the United States Court of Federal Claims considered a proposed discovery order that contained some of, but not all, the provisions from the Model Order. Among other things, the parties were eventually ordered to cooperate to identify the proper custodians, proper search terms, and proper timeframe before producing email, and “encouraged” to use narrowing search criteria (e.g., “and,” “but not,” “w/x”) to limit email production.

[19] Model Order at ¶¶ 10, 11.

[20] Introduction to Model Order at 2; Model Order at ¶¶ 6, 7.