Article New York Law Journal

Building a Better ESI Agreement: Why Customization Matters

There are a wide range of potential challenges that arise when negotiating an ESI agreement, but here Kirkland partners Michelle Six and Vanessa Barsanti examine some of the more common issues encountered and provide practical tips to help address these considerations.

The preservation and collection of potentially relevant electronically stored information (ESI) is daunting, no matter the size or scope of your case. With ever-increasing sources of ESI to manage, including mobile device applications, structured data, ephemeral messages, multimedia, and messaging platforms, clients need a bespoke discovery plan that is thoughtful and targeted in order to keep costs manageable and discovery processes efficient.

Consider, therefore, the important strategic advantage that may be gained by implementing an ESI stipulation or protocol early on in the discovery process to help clients better manage their data and spend. Though not required in many jurisdictions, such an agreement can provide predictability and transparency in discovery efforts, and generally reduce the likelihood of costly, time-consuming discovery disputes. There is no one-size-fits-all model when crafting such an agreement, but counsel for the producing party is typically best situated to evaluate their client’s data map and determine what needs to be retained, collected, and reviewed, as well as how such data should be produced.
There are a wide range of potential challenges that arise when negotiating an ESI agreement (too many for a single article), but here we examine some of the more common issues encountered and provide practical tips to help address these considerations.

Understand the Client’s Data Sources
When included in an ESI agreement, invariably one of the most impactful provisions will address the various custodial (and non-custodial) data sources and file types the parties agree to review and produce. It is important to consider not only what data the client possesses, but also which sources reasonably could contain unique (non-duplicative), relevant information. While some data sources such as email and unstructured data are commonly collected and produced in major litigation, counsel should speak with their client—and in many cases their client’s Information Technology department—to wrap their arms around the various data systems the company has, and how burdensome it would be to collect and process this information. Agreeing blindly to include sources without considering the client’s IT landscape can prove an expensive and wasteful error.

Conversely, the parties may also consider agreeing about what sources may be excluded from discovery, potentially because they are unlikely to contain uniquely relevant data, are inaccessible, or are too burdensome to be included. Common sources excluded include voicemails, messaging platforms, backup tapes, and the like, but these exclusions are case-specific, and should be made with careful deliberation and clear information that supports the exclusion.

Even with careful investigation, it may be difficult to fully identify all pertinent data sources early in your case. One way to promote a fair and reasonable resolution in the event that new relevant data sources become known after the parties reach agreement on an ESI protocol is to include a schedule for both the disclosure of the newly discovered source and an agreement to meet and confer about the discoverability of data contained therein.

Production Format

One of the most often overlooked issues in negotiating an ESI agreement is the format for production. Without any agreement on the format, the Federal Rules provide that the requesting party “may specify the form or forms in which electronically stored information is to be produced.” Fed. R. Civ. P. 34(b)(1)(C). While there may be opportunity to object to any such specifications, the dispute can be avoided altogether by reaching an agreement prior to the start of productions.

Before reaching an agreement with opposing counsel regarding production format it is crucial to consult with your e-discovery vendor and/or firm technologists. Format requests are becoming less standard, trending towards ever-more-specific requests for additional metadata and even the production of native files (in addition to standard TIFFs). A vendor or technologist will be able to inform counsel as to whether such metadata even exists in the client’s data, whether it would be burdensome to collect and process such metadata (or even require re-collection if collection is already underway), and what added expense may be associated with the review and production of natives.

Other considerations for production format include how the parties treat privileged documents. This area tends to be rife with dispute when not resolved in advance. Consideration may be given as to whether all responsive fully privileged documents will be produced as a slipsheet or only some subset, such as those in families with at least one non-privileged responsive document. Counsel should consider not only their own client’s likelihood of responsive privileged documents, but also whether they expect the opposing party to have significant privileged content. As the receiving party, a potential advantage to consistently requiring slipsheets is the immediate knowledge of the withholding rather than waiting for the privilege log, making privilege challenge reviews more easily conducted within the review platform. As the producing party, consistently requiring slipsheets can simplify production logic requirements for the vendor, creating overall efficiency and less room for error. One possible downside in this scenario is the loss of opportunity to re-review and downgrade before claiming the privilege, as one would typically do through the logging process when fully privileged families are simply withheld without slipsheets. The interplay between the ESI agreement and any privilege logging agreement may be an important consideration in determining what suits the parties’ needs.

Another big-ticket format-related item worthy of early consideration is whether to allow threading; that is, where parties produce only the most inclusive email message when the less inclusive messages are fully contained therein. From both a producing and receiving party standpoint, the ability to utilize threading can mean significant cost-savings in terms of review volume. From a producing party standpoint, it also helps to avoid inconsistent calls across the thread and promotes further cost-savings in the privilege logging process as there are fewer documents to log. However, it is important to understand the email program(s) utilized before agreeing to threading as it may impact a reviewer’s ability to see all information. For example, some email programs do not keep all sender/recipient information intact on the face of the communication for each of the lesser emails as the chain continues. As the producing party, this may make it more difficult for reviewers to determine privilege if it is unclear what individuals were on the chain and at what time. As the receiving party, this could make accurate fact development more difficult.

If not otherwise contemplated by a protective order, an ESI agreement can be the right instrument to clarify permissible scope of redactions, as well as how such redactions will be labeled. Depending on your jurisdiction, if there are strong reasons a party desires the ability to make non-responsive redactions, such as non-relevant trade secret information, reaching an agreement may be the only avenue to permitting such redactions. Be cautioned, however, that as the producing party this is likely to add additional review time and costs. The receiving party may also request that field(s) be provided identifying when a document is redacted for non-responsiveness, and this frequently includes requests for categorizations of the non-responsive information.

Another type of redaction often taken for granted is personally identifiable information (PII). Though many courts’ local rules provide that specific categories of PII (e.g., personal contact information, social security numbers, medical information, etc.) must be redacted in public filings, they typically say nothing as to whether such information can or should be redacted within an initial document production. Counsel often unilaterally perform these redactions thinking that opposing counsel will be uninterested in the information, however relevancy is often in dispute. Performing such redactions without agreement can result in costly re-reviews and additional document production.

Search Methodology
Search parameters and methodology can be one of the most difficult and complex pieces to reach agreement on, especially early in the case, which is when the ESI agreement is typically negotiated. While the parties are unlikely to resolve every aspect of search parameters and methodology, early discussions about these factors can provide valuable insight, even revealing places where there is actually agreement. For example, it may reveal that both sides are interested in using Technology Assisted Review (TAR), or perhaps even agreement on the date parameter for discoverability.

Even if there are no pieces the parties can reach consensus on at the time of the ESI agreement, they can still consider including a future schedule laying out timelines for later meeting and conferring and, if necessary, briefing to resolve any remaining disputes. While not all cases benefit from having such a push, when opposing counsel is slow to respond or particularly difficult these types of roadmaps may prove useful in moving the discovery process along efficiently.


The most successful ESI agreements are those that are tailored both to the particular client and matter at hand. Using anything but a customized agreement presents risks to the client’s pocketbook, as well as the overall litigation strategy. Tailor-made ESI agreements can prove beneficial to all parties when entered into strategically, providing a clear roadmap, bringing a level of certainty to the process, and thereby maximizing efficiencies in the discovery process.

Michelle Six is a partner at Kirkland & Ellis in New York, where she advises global clients on strategic electronic discovery considerations and data protection and privacy issues. Vanessa Barsanti is a partner in the firm’s Chicago office, where she works with clients to develop practical, efficient, and defensible discovery strategies.