At firm lunches, I always prompt senior partners to tell some of their favorite old stories of their time as junior associates. Sometimes their narratives date back to the 1970s and 1980s, decades before current associates were born. The partners often detail the horrors of law practice without internet or cell phones (to which the associates in attendance always respond that, honestly, being unreachable on an airplane does not sound too terrible). They recall the prevalence of handwritten comments (truly, not much has changed after all) and consistently tell stories about their time serving as the discovery workhorses of the firm. They describe a world dominated by brand names that have since been genericized in the legal world: using hand-operated Bates stamps to label productions, carrying Redweld folders full of documents to and from the office, and measuring the size of each production by the number of Bankers boxes required to mail it.
These old stories often shed new light on the justifications for apparently idiosyncratic tendencies that linger on today. For example, one of the junior associates at a recent lunch had been chastised by a partner for using blue highlighting to denote privileged content in a document when she should have used pink. The junior associate has dutifully used pink highlighting since but never understood why. During the lunch, we learned that when that partner was an associate, he signified privileged hard-copy documents with pink slipsheets to exclude them from production. Now associates forever mark privilege with pink (and blue for non-responsiveness, of course).
Those stories of hard-copy productions are fascinating to young associates for the same reason rotary phones enthrall young children—they are foreign. Discovery looks nothing like it did in those days, and not only because the documents are electronic and depositions happen over videoconference. The way the world works has changed—and continues to change dramatically—and the corpus of discovery has changed with it. Our discovery strategies must adapt to this new landscape.
The basic goal of discovery bears repeating: Your discovery strategy must be driven by your dream exhibit and witness lists at trial, even if you never get there. If you begin with what you need to prove, you can work backward to fill in the gaps. This strategy holds true regardless of which side of the v. you happen to be litigating. Each side has its own story, and each side must find the evidence to support it.
But in an age where we all create enough written records every day to fill several Bankers boxes, reverse-engineering the funnel that leads to key evidence at trial is becoming more and more difficult, not to mention resource-intensive. The Federal Rules of Civil Procedure provide little concrete guidance. The relevant rule, Rule 26, which was revised almost a decade ago, is still often misquoted during discovery meet-and-confers, causing widespread confusion among junior associates who have never litigated using the prior version of the rule.
Indeed, the standard in Rule 26 is no longer that documents are discoverable so long as they are “reasonably calculated to lead to the discovery of admissible evidence.” Rather, since 2015, Rule 26(b)(1) has provided that discovery must be both “relevant to any party’s claim or defense” and “proportional to the needs of the case,” considering six factors: (i) the “importance of the issues at stake in the action,” (ii) the “amount in controversy,” (iii) the “parties’ relative access to relevant information,” (iv) the “parties’ resources,” (v) the “importance of the discovery in resolving the issues,” and (vi) “whether the burden or expense of the proposed discovery outweighs its likely benefit.” Despite the long list of considerations, most discovery fights focus on the last factor: weighing the burden of the requested discovery against the benefit to the party requesting it.
That burden has the potential to expand exponentially in light of technological advancements. In 2013, when the Senate Judiciary Committee considered the changes to the Federal Rules of Civil Procedure that eventually became the 2015 amendments—including the specific addition of the proportionality standard for the first time—the discussion opened with a statement highlighting that civil litigation in America “can be very expensive,” which is largely due to discovery costs. That statement was indisputable even then, and document discovery has only continued to balloon since.
Countless articles have been written, and continue to be written, exploring the ins and outs of collecting and preserving ever-changing data sources. Whether final contract terms were negotiated over the latest secure messaging app like Signal or whether important meeting notes were recorded in a company executive’s iPhone Notes app, discovering the correct digital records that will become the crucial exhibits at trial is a daunting task. Some estimates suggest that in one year, over six billion chats are sent in Microsoft Teams and over 17 billion in Slack. The potential sources seem never-ending and ever-expanding.
Our discovery strategies must adapt, but the process is not hopeless. Consider your own day (or ask a junior associate about theirs). Professionals today transition seamlessly between a slate of electronic apps. The set for each of us may vary, but we are internally consistent and use our own electronic record keeping in specific ways. For instance, one team may use OneNote to take notes and Teams to instant message throughout the day, while another team may use Word and Slack for the same tasks. You simply do not know which you need, though, until you ask.
Efficient discovery, therefore, begins with an understanding of how relevant records were kept, before any collection takes place. The simplest way to find out is first to informally ask and then to confirm via formal discovery. Informal requests are often effective because counsel have an obligation to cooperate to manage discovery, and efficiencies can be achieved on all sides by limiting initial collections to locations or types of data that are likely to contain relevant information and that impose the most reasonable burden to collect.
The Sedona Principles, a series of publications that set out various principles to address issues related to electronically stored information and that are frequently cited as authoritative in the changing world of e-discovery, squarely place the burden on the responding party to assess the appropriate procedures, methods, and technologies for preserving and producing its own electronically stored information. This assignment arises from the idea that the responding party is “best situated” to do so.
Requesting counsel should trust opposing parties’ representations but should always verify them, whether through review, interrogatories, or confirmation during depositions. For example, if, during document review, messages appear to be missing or apps or platforms are referenced that were not previously identified during negotiations, the parties can negotiate supplemental productions. Relatedly, if a particular data source is proving especially difficult to collect—involving, for example, the need to manually export individual pages—counsel can negotiate more limited date ranges or custodians specific to the more burdensome data sources. They can reserve their rights to expand their requests, if warranted, after review of the initial production and as applied to other data sources.
The way we communicate today should also influence the structure of our requests. For example, during an important meeting, individual participants may be simultaneously communicating via text message or other messaging platforms without providing much context in the messages themselves because they see each other on screen. Thus, a request for all communication on any platform during a key meeting by meeting participants, without the use of search terms, may be more valuable than the more traditional approach of requesting all emails sent by the most senior individuals within a more extended date range and applying search terms. Downloads and debriefs often happen in real time across such platforms, and those real-time responses provide essential fodder for deposition questioning about the content of the meeting.
However, that sort of multi-platform communication requires careful review by individuals who deeply understand the facts of the case, because a short-form conversation over text message may be inscrutable to an uninitiated reviewer. Counterintuitively, deeply exploring communications and notes during narrower time periods and using broader search terms may be more fruitful, given the massive amount of data available and the lack of formality in modern daily written communication. This approach can be more effective than reviewing huge amounts of piecemeal data over several years, especially considering the speed at which areas of focus can change day to day.
Remember, the goal is to find the clips and snips to show a fact finder during a trial. Although the discovery process necessarily involves some amount of exploration, it is also ends-oriented. A more focused approach will make it possible to tell a more cohesive story of the critical moments, even if a stray helpful email between meetings is missed.
Once you have received the targeted discovery you collected or requested, the story building finally begins. This process of sorting through the equivalent of a building’s worth of Bankers boxes of documents is the area that appears most ripe for technological assistance. In 2012, Magistrate Judge Andrew J. Peck of the U.S. District Court for the Southern District of New York reached a groundbreaking decision in Da Silva Moore v. Publicis Groupe, 287 F.R.D. 182 (S.D.N.Y. 2012), approving the producing party’s use of technology-assisted review (TAR) to review and turn over responsive documents in discovery. Since that decision, the idea of automating some part of the review process—whether in relation to outgoing or incoming discovery—has gained increasing attention but has raised a dilemma for attorneys. On the one hand, many attorneys salivate over the opportunity to have a computer review a broad set of documents to efficiently identify the important ones and save time and money in the process. On the other hand, many attorneys express inherent distrust about relying on the judgment of technology to make critical determinations during the discovery process. This includes decisions regarding privilege or other protections, responsiveness, and which documents are interesting.
Incorporating emerging technologies into the discovery process only proves to become more complicated. TAR turned out to be merely a precursor to true artificial intelligence (AI). Ideally, these technologies can help our stodgy profession progress into the modern world. Someday, we will regularly use them to more efficiently digest data and thus quickly advise clients and avoid having document discovery drag on for years. AI has the potential not only to review large document sets and identify relevant documents but also to digest the documents and respond to questions about the data in the set it reviewed.
Navigating the usefulness of AI in the world of discovery will require a new literacy in technology that many of us lack, however. In our profession, the nuts and bolts of document discovery have long been left to the younger attorneys who commiserate with their counterparts at opposing law firms about the miseries of document production details. Expecting senior attorneys who have never logged into a Relativity database or designed a coding panel, and who certainly have no idea what the difference is between an MD5-hash duplicate and a text-based duplicate, to jump straight into effectively using generative AI in litigation is optimistic at best. Unsurprisingly, recent surveys suggest that a majority of legal professionals do not feel the industry is ready to handle the implications of generative AI.
The impact of TAR and AI in the proportionality analysis of Rule 26 is also ripe for debate as the technologies become more prevalent. If using technology makes a review more efficient and less burdensome on individual reviewers, a “proportional” approach may require the party using the technology to collect and review a significantly higher volume of data than it otherwise would. However, such a result would misplace the relevant incentives and risk fishing expeditions based on the low burden on the responding party, rather than produce an overall increase in efficiency. But it is too early—particularly in the AI space—to know exactly how courts will address the issue going forward.
Oddly enough, despite the drastic changes to document discovery in recent years due to the over-proliferation of written records and varying data sources, depositions have changed very little. They now occur more frequently over video, although many attorneys prefer to do them in person when possible. But the basic strategies that everyone has used for years appear exactly the same. And the war stories the senior partners tell at lunches about deposition theatrics, unlike their document discovery stories, sound very similar to the depositions we still experience today.
Attorneys generally begin depositions by laying out the rules for the deposition, asking background questions, presenting the witness with and authenticating documents, and then transitioning to more challenging or confrontational questions after the lunch break. For every block of questions, counsel tend to begin with general open-ended questions. Counsel then pivot to more direct questions to elicit shorter answers toward the end of each block to funnel information into the snippets that will ultimately be played at trial.
There have been some changes to depositions, although subtle. Many attorneys now spend time asking deponents about where and how they communicated with their colleagues and where they kept their documents. Presumably, they ask these questions because of the variety of possible answers and opportunities for further collections; many depositions I have attended have been followed by supplemental document requests. Relatedly, attorneys spend significant time merely authenticating the many potentially relevant documents custodial to each deponent. This exercise tends to bore everyone in the room, especially the deponent. And it is likely unnecessary in light of the common practice of stipulating to the authenticity of electronically stored information produced in the ordinary course of discovery.
Deposition strategy is ripe for improvement and growth along with the rest of the discovery process. Because of the large volume of record evidence, counsel may be lulled into assuming that exploratory questions at depositions are unnecessary, given that they have already pieced together the information they need from the written record. But that is a trap. Although we write down more information than we used to, that writing is often incomplete and informal. Making assumptions about the facts based on an amalgamation of those partial writings is risky, and depositions are the perfect time to challenge assumptions about the facts and to test theories. Those exploratory questions, like asking why a decision was reached or why someone important was excluded from an email, can feel riskier but will ultimately be more rewarding.
Going forward, I am eager to see how depositions change to adapt to the new world of technology. Using a video or spreadsheet as an exhibit at a deposition is already much easier than it used to be. Future legal and strategic questions will almost certainly involve wrestling with the use of AI in writing deposition questions. This will require addressing ethical dilemmas and acquiring the technical expertise to use AI to effectively request information germane to the case and to form effective questions from the large volume of produced data.
Discovery is the bread and butter of civil litigation practice, and some things will never change. Discovery battles will continue to be the place where many young litigators cut their teeth. There will always be lingering fears that our extensive efforts did not find that one important document that would make or break the case (and we will always cross our fingers that it does not show up for the first time at a deposition or, even worse, at trial). But there are many lessons to be learned from the war stories of years past and the trends that are developing daily to develop a focused, efficient, and strategic approach to the process.
I look forward to learning which parts of our current discovery practice will sound delightfully arcane over lunches in years to come.
