“Hit reports” have various names.[1] One issue that has often arisen is whether a producing party must produce every document on a hit report requested by an opponent, even if irrelevant or nonresponsive to discovery requests. Craig Ball has blogged on this issue since 2013. However, his recent blog explains that “the issue remains contentious….” [emphasis added]. Craig and others suggest that the issue be resolved by negotiation. Craig Ball, Federal Court Rules on Whether Documents Containing Agreed-Upon Keywords are Responsive Per Se | Ball in your Court (craigball.net) (Oct. 25, 2021).
At the end of this blog, I suggest some language to that effect. In my view, the most prudent course is a robust agreement concerning hit reports. Further, any effort to demand production of irrelevant or non-responsive ESI could lead to costly “document dumps” and other problems. It should be precluded as beyond the scope of permissible discovery.
A hit report shows the number of times a search term appears in a database and may show the number of unique documents containing that term. DISCO provides an example of a hit report generated by its litigation review software:
The term “hit report” covers a lot of ground.
This blog addresses only the first category.
Jill C. Griset and Anne B McCray point out that: “Opposing counsel may assume that because a document hits on a term it is relevant and then expect to receive at least the majority of documents that hit on the search terms.” E-Discovery Update: Producing Search-Term Hit Reports: Another Form of Discovery on Discovery | McGuireWoods LLP – JDSupra (Jul. 11, 2016).[2]
However, as pointed out by Josh Gilliland, that assumption may not be well founded. Responsive to Search Terms Does NOT Mean Relevant – Bow Tie Law (Aug. 16, 2016). Josh cited BancPass, Inc. v. Highway Toll Admin., LLC, 2016 WL 4031417, at *1 (W.D. Tex. July 26, 2016). In pertinent part, the BancPass court wrote that: “By its motion, BancPass wants to force HTA to produce all non-privileged documents responsive to certain previously agreed upon electronic search terms….” The court explained:
The essence of BancPass’s argument is that it and HTA’s email negotiations over these specific search terms—which are memorialized in a series of emails—constituted a binding and enforceable agreement between the parties. BancPass states that it believes the parties agreed that “all non-privileged documents responsive to the additional search terms will be produced,” which it understood to mean that every non-privileged document the search turned up would be produced, regardless of whether it had anything to do with the parties’ dispute…. BancPass argues that an agreement by the parties to run search terms and produce all non-privileged results prevents a producing party from later attempting to withhold documents based on relevance. [emphasis added].
However, HTA disagreed. It withheld 20,000 non-responsive documents. BancPass argued that this was a violation of the agreement. The court wrote:
HTA states that it believed that its email exchanges with BancPass did not amount to a binding agreement. Rather it understood the parties’ agreement to produce “all non-privileged documents responsive to the additional searches” to mean the parties would use the search terms “to capture a universe of potentially responsive documents, and then, subject to stated relevancy objection, the parties would produce non-privileged documents that were responsive to actual discovery requests.”
The court agreed with HTA: “If the Court were to construe the parties’ email correspondence on this topic to constitute a contract, it is likely that HTA’s actions would amount to a breach. The Court’s reading of the e-mail chain is that the parties agreed to produce all of the results of the searches save privileged documents. But the parties’ e-mail exchange is not a contract. Rather, it was a means to simplify and limit the scope of production responsive to BancPass’s requests for production….” It then permitted responsiveness review.[3]
More recently, it has been stated that:
There appears to be only a handful of courts that have addressed this issue. As discussed above, the court in FlowRider Surf held that an agreement to run search terms does not waive the producing party’s right to object based on relevance nor acquiesce to produce all resulting documents.
O’Donnell/Salvatori Inc. v. Microsoft Corp., 339 F.R.D. 275, 277 (W.D. Wash. 2021). The O’Donnell court concluded: “This Court similarly holds that a party’s agreement to run search terms does not waive its right to review the resulting documents for relevance so long as the review can be done in a reasonably timely manner.” [emphasis added].
In a later decision, the O’Donnell court explained: “I am going to deny [ODS’s] motion to compel because I do find that Rule 26 does not require that Microsoft produce irrelevant information.” O’Donnell/Salvatori Inc. v. Microsoft Corp., 2021 WL 4948188, at *1 (W.D. Wash. Oct. 22, 2021)(emphasis added). The same argument was made in a Brief of Amicus Curiae Lawyers for Civil Justice:
While the Court has established through the Rules Enabling Act process that discovery is limited to relevant material, this Court’s guidance is needed because the lower courts are divided on enforcing the rules as written and intended.…
This conflict in authority is even more prevalent at the district court level. Many decisions have correctly denied compelled production of irrelevant material and upheld the producing party’s right to review the documents for relevance. See, e.g., … FlowRider Surf, Ltd. v. Pac. Surf Designs, Inc., 2016 WL 6522807, at *7-8 (S.D. Cal. Nov. 3, 2016) (rejecting effort to compel production of all search term hits without relevance review); Bancpass, Inc. v. Highway Toll Admin., LLC, 2016 WL 4031417, *2-3 (W.D. Tex. July 26, 2016) (rejecting attempt to compel production of all documents hitting on search terms); Gardner v. Cont’l Cas. Co., 2016 WL 155002, at *2-3 (D. Conn. Jan. 13, 2016) (rejecting as “untenable” plaintiffs’ request to compel defendant to turn over all documents hitting on search terms without relevance review, because “[a]s every law school student and law school graduate knows, when performing a computer search on WESTLAW and/ or LEXIS, not every case responsive to a search command will prove to be relevant to the legal issues for which the research was performed.”); Chen-Oster v. Goldman, Sachs & Co., 2014 WL 716521, at *1 (S.D.N.Y. Feb. 18, 2014) (rejecting plaintiffs’ contention that “defendants are obligated to produce all documents returned by the search without exercising further judgment with respect to responsiveness”); Wilson v. Rockline Indus., Inc., 2009 WL 10707835, at *1 (W.D. Ark. Oct. 22, 2009) (rejecting party’s request that the plaintiff be compelled to turn over all documents hitting on search terms, because “[i]n our system of law, we allow the party responding to discovery to filter his [or her] own documents and produce only those which are relevant to the litigation. In the absence of some showing that relevant information is being withheld-and here there is none-there is no basis to make the responding party produce all information. Indeed, to do so would make a mockery of F.R.C.P. 26(b)(1).”).
Other district courts, however, have erroneously ordered production of documents without review for relevance, resulting in compelled production of irrelevant material. See, e.g., Consumer Fin. Prot. Bureau v. Navient Corp., 2018 WL 6729794, at *2 (M.D. Pa. Dec. 21, 2018) (compelling plaintiff to produce all documents hitting search terms); Fairholme Funds, Inc. v. United States, 134 Fed. Cl. 680, 686-88 (2017) (compelling a “quick peek” procedure under Rule 502(d) over the defendant’s objection); UPMC v. Highmark Inc., 2013 WL 12141530, at *2 (W.D. Pa. Jan. 22, 2013) (compelling production of all documents relating to a particular individual); Carrillo v. Schneider Logistics, Inc., 2012 WL 4791614, at *10-11 (C.D. Cal. Oct. 5,2012) (ordering producing party to retain an outside vendor to collect documents from its servers and holding that “no documents identified by the vendor may be withheld on relevance grounds”); Williams v. Taser Int’l, Inc., 2007 WL 1630875, at *6 (N.D. Ga. June 4, 2007) (ordering party to produce all “presumptively responsive documents” that hit on particular search terms, subject only to privilege review).
Accordingly, this case provides an opportunity for the Court to re-establish the authority of the federal rules while resolving the conflict in the Circuits and district courts regarding whether courts may compel the production of irrelevant information without prior review for relevance and responsiveness.
Amicus Brief in Actavis Holdco US, Inc., et al., v. State of Connecticut, et al., 2020 WL 1479871 (U.S.), 11-14. However, the Supreme Court did not issue a writ of certiorari. See 141 S.C.t. 124 (2020).[4]
That logic is, in my view, compelling. Fed.R.Civ.P. 26(b) limits discovery to nonprivileged matters that are relevant to a claim or defense and proportional to the needs of the case.[5] Non-responsive ESI is not relevant or proportional.[6]
That said, the parties likely can alter the scope of discovery by agreement. That’s the rub with a loosely drafted hit report agreement.
Both Craig Ball and Douglas Austin have blogged on O’Donnell.
Craig points out that: “If you enter into an agreement with the other side to use keywords and queries for search, be clear about expectations with respect to the disposition of items hit by queries. Assuming the items aren’t privileged, are they deemed responsive because they met the criteria used for search or is the producing party permitted or obliged to further cull for responsiveness based on the operative Requests for Production? You may think this is clear to the other side; but, don’t count on it. Likewise, don’t assume the Court shares your interpretation of the protocol. Just settling upon an agreed-upon list of queries may not be sufficient to insure a meeting of the minds.” Federal Court Rules on Whether Documents Containing Agreed-Upon Keywords are Responsive Per Se | Ball in your Court (craigball.net) (Oct. 25, 2021).
Douglas Austin noted that he has been addressing the issue since 2016. Court Declines to Compel Microsoft to Produce All ESI Retrieved from Search Terms (ediscoverytoday.com).
Given the excellent available analysis of this issue by both writers and courts, any mistake in this regard is an unforced error. Good drafting and a Rule 26(f) discussion can prevent problems with hit reports and production requests. For roughly seven years, I have included language like the following:
In the spirit of cooperation and to further the goal of proportionality, we agree to run one or, within reason, more “hit reports” for you. The purpose is to permit you to focus your subsequent requests. We will use [name of litigation review software] and run the terms against electronically stored information in the [name database].
Please note that a “hit” may be beyond the scope of discovery, irrelevant, privileged, work product, non-responsive, a trade secret, confidential, proprietary, etc., and all rights are reserved to object to the production of any or all “hits.” This includes without limitation the right to assert that any or all production is not proportional to the needs of this case. Further, all rights are reserved as to any and all other objections under the Maryland Rules [or Federal Rules of Civil Procedure], including the right to move for a protective order for any reason permitted under the Rules. All objections to any discovery request are incorporated herein, and none are waived by the hit report. No privilege is waived by production of a hit report because the hit report is not an offer to produce any document. Additionally, the right to mark a “hit” as confidential or highly confidential under the agreed protocol or protective order is fully reserved. Further, we reserve the right to add additional data to, or remove data from, the database and that may vary the number of hits. All rights are reserved, including without limitation, rights under Fed.R.Evid. 502.
At least since the date of Craig’s 2013 post, the road map has been clear.[7]
_____
[1] For example, Relativity uses the phrase “search terms reports.” Search terms reports (relativity.com).
[2] The J.D. Supra article lists decisions addressing discoverability of a party’s internal hit reports. They argue that “the courts have given approval to keep the hit report private.” That issue is not addressed in this blog.
[3] In his blog, Josh agreed: “Plans for producing ESI solely based on broad search terms can end with lawyers on a Highway to Hell with excessive irrelevant data to review.” See also Josh Gilliland, Think in Terms of What is Relevant, Not Producing Keyword Hits – Bow Tie Law (“Just because something is a hit to a search term does not mean it is relevant.… When requesting parties want all search hits in an email mailbox containing a common search term, they are asking for trouble.”)(Dec. 10, 2018), and The Codename is “Motion to Compel Review of the Search Hits” – Bow Tie Law (“Document review of search term hits can be described as ‘trust but verify.’ No matter how confident a party is that search terms will identify relevant hits, it is radically dangerous to trust the results without verifying the accuracy of the hits.”)(Nov. 20, 2018).
[4] The District Court order stated: “Defendants shall apply the agreed search terms to the agreed custodial files and may review the identified documents for privilege, but may not withhold prior to production any documents based on relevance or responsiveness…. Any dispute arising out of the above provisions shall be brought to Special Master Marion and ESI Master Regard via simultaneous letter briefs….” In re Generic Pharms. Pricing Antitrust Litig., 2019 WL 8106511, at *1 (E.D. Pa. Oct. 24, 2019).
[5] Maryland Rule 2-402 permits discovery of any matter that is not privileged if the matter sought is relevant to the subject matter of the action. Although that is broader, in my view, the same result would be reached under that rule.
[6] To the same effect, Josh Gilliland wrote: “Litigants frequently battle over the production of search term hits, regardless of whether the ‘hits’ are even relevant. This is absolutely the wrong beach to die on in a lawsuit. There is the brutal reality of the costs of motion practice for the client and judicial resources spent refereeing that melee. Moreover, the ugly truth is no one should want gigabytes of irrelevant information ‘just cause’ they want irrelevant information.” J. Gilliland, “Reviewing Search Hits for Responsiveness is a Good Thing,”(Oct. 25, 2018), citing Umg Recordings, 2018 U.S. Dist. LEXIS 164761, at *17 (W.D. Tex. Sep. 26, 2018). “Reviewing search hits for responsiveness is normal to do a good faith production. Courts loathe ‘document dumps’ where a requesting party is buried in non-responsive information. Moreover, reviewing search hits to identify potentially privileged should be standard operating procedure to identify potentially privileged information and that the ‘hits’ are actually relevant.” Id.
[7] The flip side of this dispute is shown by a “belts and suspenders” agreement that: ““ESI that is known to a Party to be responsive to a discovery request or relevant to the subject matter of this action may not be withheld on the grounds that it was not identified as responsive by the protocol described in, or developed in accordance with, this Order.” Josh Gilliland, “A Subscription for Search Terms” (Oct. 11, 2018), citing Price v. Synapse Grp., Inc., 2018 U.S. Dist. LEXIS 155637, at *2-3, *17-18 (S.D. Cal. Sep. 12, 2018). That issue is not addressed here. Parties should, however, be alert to the concern that an opponent may have actual knowledge of a responsive, non-privileged document that does not “hit” on a search term. See, e.g., “9th Annual Georgia Symposium on Ethics and Professionalism” 60 Mercer L.Rev. 845, 876 (2008)(quoting Jason Baron re mistakes in keywords).
Another issue not addressed in this blog is the difference between “hits” and “unique hits.” Relativity explains that: “Unique hits is the count of documents in the searchable set returned by only that particular term. If more than one term returns a particular document, that document is not counted as a unique hit. Unique hits reflect the total number of documents returned by a particular term and only that particular term…. Unique hits can help you identify terms in your search terms report that may be overly inclusive.” Search terms reports (relativity.com)