Dispelling E-Discovery Myths in Internal and Government Investigations

To do that, we’ll take a look at some e-discovery “myths,” and then dispel those myths by examining reality. And we’ll use a hypothetical securities matter as the backdrop:

It’s 3:30 p.m. on the Friday before Labor Day, and your supervising partner appears in your office. You have been assigned to a new internal investigation for an important financial services client. The partner feels certain that the matter will lead to self-reporting, possibly filing a Suspicious Activity Report, and then a full-blown SEC investigation. Your role is to lead the e-discovery portion of the investigation. The partner explains that the client has already collected roughly 1.2 million documents from seven key custodians and expects the firm to complete its work on the investigation two weeks from today. Hard drives with the collected documents are on their way to your office via courier. After canceling your plans for the weekend (“after all, that’s why we call it Labor Day…”), you sit down to consider your options and map out a game plan.

Myth #1: As a junior level practitioner, you have the skill set and knowledge to manage this project— you’re an expert with your MacBook, your smartphone, the firm’s electronic timekeeping system, your Xbox. . .

Reality #1: Regardless of seniority or expertise with day-to-day technology, all lawyers need help understanding e-discovery best practices and can only gain e-discovery experience by handling matters over time.

So what does an inexperienced attorney do in this scenario? First, “know what you don’t know.” Ask for help. That help can come from more experienced lawyers in your firm, the litigation support group at the firm, or from a trusted service provider (more on finding one of those later). Second, draw on the good project management skills you developed working on other types of matters. Identify all the stakeholders in the project, and define everyone’s roles. Establish a protocol for communication. Document all decisions about the project. Finally, don’t be afraid to get your hands dirty. This project will likely require you to use new software and understand cryptic technical jargon. Get ready to dig in and work hard—you will not succeed if you over-delegate work to more junior associates, contract lawyers, paralegals, and/or litigation support staff.

Myth #2: For internal investigations, it’s OK to streamline and shortcut the process.

Reality #2: Internal investigations often involve faster timelines and larger quantities of unfiltered data than requests in litigation or from regulators. And you never know when an internal investigation will lead to something bigger. So it’s even more important to develop a defensible process and workflow, then document and follow it.

Why can internal investigations be so complex and difficult to manage? There are several reasons:

• Companies must sometimes self-report suspected issues within a specific timeframe, or often need to meet deadlines for filing current or amended financial statements—all of which puts time pressure on the investigation.

• When you’re looking for a “needle” in these cases—evidence of wrongdoing—the “haystack” of data can be difficult to reduce to a manageable size.

• Time frames often are undetermined. Unlike litigation, where a statute of limitations contains the temporal scope of discovery, in a securities investigation you may not know when key events happened.

• When investigating a potential “bad actor” within an organization, the heightened risk warrants a thorough, no-stone-unturned approach.  

Myth #3: All e-discovery vendors are the same.

Reality #3: The fact is, like any service company, e-discovery vendors have varying levels of expertise and experience. You must establish a relationship of trust and clear communication with your e-discovery vendor for your project to succeed.

Here are a few tips on vetting an e-discovery provider:

• What does a “typical” project look like? This is a bit of a trick question—there is no “typical” project. The most credible response from a provider will emphasize that every matter starts with a documented, defensible process. Even “typical” matters encounter problems, but a documented process can help determine what went wrong and why—which is critical if an opposing party or regulator questions your production.  
• How does the vendor staff their cases? You should come away with a full understanding of who will be staffed on your project. What is each person’s role, and what is their background and experience? Who will be your main point of contact, and how is that person the right fit for the matter?
• What is the provider’s review tool of choice? Be aware of vendors who have one proprietary tool that you must use. “One size fits all” is not necessarily a successful model for e-discovery. Look for a company that leverages best-of-breed technologies to achieve the results you want. Your vendor should listen to your needs and understand your comfort level and sophistication with different tools. They should be able to accommodate your request for a specific technology or, alternatively, recommend a new technology and offer training on the tool for you and your team.
• Ask for references. E-discovery vendors rely heavily on word of mouth referrals. Email your firm’s litigation support team and/or litigation team when vetting a new vendor. But keep in mind that people are more likely to give negative feedback than positive, so also request references from the vendor and give them an opportunity to introduce you to clients who have had good experiences. Recognize that e-discovery is an inherently imperfect process: Seek preparedness, reliability, and accountability—not perfection.

Keep in mind that you should follow the same rules when you vet internal litigation support that may be available to you. Not every member of that team has the skills to handle every matter—make sure your project is staffed appropriately.

Myth #4: When faced with large volumes of documents to review, “throwing bodies” at the project will help you meet your deadlines.

Reality #4: Good project management, including a thoughtful scoping of the project and the resources required to accomplish the work, will enable you to meet document review deadlines with better results and more consistency. A few tips:

• Start at the end. At the outset of your project, have a frank discussion with your vendor about the deadline for completion. Be honest about the nature of your deadline: Is it self-imposed? Is the deadline negotiable, so that you can add additional time if you encounter unexpected issues? Or is it a hard deadline imposed by a regulator? If “substantial completion” is required by the deadline, define what that means. Does a deadline for “production” mean that responsive documents must be provided by that date, and a privilege log can be turned over later? Or must the privilege log also be supplied by the deadline?

• When you’re done, you’re not done. Remember that even when document review is finished, you are not done—you must build in time for preparing the production, which often can require several days of work. At the beginning of the project, discuss with your vendor the required production format and other necessary deliverables.

• Before you begin review, develop the substantive strategy. Rushing into document review without a strategy in place will result in inefficiency and poor quality down the line. Before you start, consider carefully what you are trying to accomplish with the review. What documents are you looking for? Establish the “story” of the case, including relevant dates, parties, and underlying facts. Memorialize your strategy in a concise memorandum. Provide other study guides to the review team if needed—for instance, if a complex financial instrument is at issue, ask the business unit for a summary of how the instrument is structured. What are likely to be the “close calls” separating responsive from non-responsive documents? Are there examples you can provide the review team?

• Once you understand the strategy, define the workflow. After the strategy is developed, then you can structure a process to implement it. What decisions will the reviewers make about each document? Will you have confidentiality and privilege coding? Will you require a second-level review for redactions? How will you provide substantive feedback to the review team? What is the quality assurance protocol? Rely heavily on your vendor for help with the workflow creation—that’s why you hired them.

• Bigger is (usually) not better. Consistent with your deadline requirements, keep the review team as small as possible. You can always adjust the number of reviewers after you assess the team’s throughput. But smaller teams typically produce more consistent and accurate decisions. The larger the team, the harder it is to effectively supervise the reviewers, provide them with feedback, and keep tight control on the quality and consistency.

In e-discovery, throwing bodies at the project without proper planning will waste time and money. You might as well set your client’s money—and your client relationship—on fire.

Myth #5: Using predictive coding or other forms of technology-assisted review (TAR) will significantly reduce your timeline.

Reality #5: As powerful as these TAR tools are, they will not necessarily save time (or money) in every case.

A few things to keep in mind about using TAR for a document review project:

• TAR software does not operate with the quick press of a button. These tools require significant substantive input from attorneys knowledgeable about the case to “train” the system on what is relevant. Just as with manual human review, the process and workflow must be developed and defined. The results of these tools are only as good as the planning and input that goes into them—the classic example of “garbage in, garbage out.”

• TAR isn’t effective for every document collection. These software tools rely on an algorithmic analysis of document text. If your document collection is heavy with non-text documents—numeric spreadsheets, for example—the technology will not provide good results. Also, the tools need a minimum quantity of relevant content to “learn” from—if you’re looking for one needle in a haystack, TAR is not a good option to find the needle.

• Use technology to improve human review rather than replace it. Even if you decide that you will stick with old-fashioned human review for your matter, consider using TAR in other ways to improve your project. For instance, the technology can be used to prioritize review, so that the documents most likely to be relevant are reviewed first. This can be a huge help with internal investigations, especially when you cannot interview the custodian (because she is no longer with the company or may be the subject of a suspicious activity report and cannot know that she is under investigation).

Technology should be implemented strategically when it can help achieve a well-defined and specific goal in an easier, quicker, and more accurate manner.

Myth #6: Traditional search methods such as keywords strings can effectively find documents containing evidence of wrongdoing. And predictive coding tools work even better.

Reality #6: In many investigations involving potential violations of securities laws, the evidence you’re looking for simply can’t be found using keyword searches or even the sophisticated predictive coding tools available today.

In internal investigations, often you are trying to assess whether an employee was a bad actor, for how long the bad acts took place, whether there was a systemic negligence in locating problems, and whether other employees can be linked to these actions. It’s almost impossible to train a computer to recognize documentary proof of a bad act because humans generally try to mask their bad actions. Likewise, using specific search terms to find this evidence often is futile, because you don’t know what words and phrases were being used in connection with the wrongdoing. (You certainly wouldn’t expect an employee engaged in insider trading to use the phrase “insider trading” in her emails!) Typically, keyword searches and TAR tools will only shrink the size of the haystack you are searching—the documentary proof you’re looking for is still just a needle in that haystack.

Nevertheless, there are some technologies that can help. For example, software that clusters documents by themes and concepts can be helpful to locate patterns of a custodian speaking in code to hide bad behavior. Social networking analysis can also be helpful; using the example of insider trading again, there are tools that can analyze the frequency and timeline of your suspect communicating with others, which you can then compare to trade dates.

Myth #7: All government regulators will resist your use of predictive coding to find and produce documents.

Reality #7: If you can demonstrate the effectiveness and defensibility of predictive coding in your matter, many government lawyers—especially those in the SEC and Department of Justice—will at least consider your request to use the technology.

Even within a government regulatory agency, some attorneys will be open to your use of TAR to satisfy a production request. Be patient—if a regulator seems interested in the idea, take the time to train them about the technology and your selected tool. Show them your proposed workflow, and help them understand why the process works and how you can validate it. Don’t make the mistake of being condescending: Many of the staff attorneys at regulatory agencies simply have no access to training on these tools, and they are eager to learn. You can make significant inroads for your client (and your own personal reputation within the agency) by providing that knowledge.

I hope this examination of some of the “myths” of e-discovery will help you better plan and execute your next significant document review and production for an internal investigation or government request. As unglamorous and tedious as this work can be, it is critical to get it done right, on time, and within budget (another critical consideration that deserves its own separate article). Plus what a great thrill it is to be the lawyer who, against great odds, finds the smoking gun your client needs to identify an instance of securities fraud!  (And there will always be another Labor Day to enjoy).