Uncovering the Facts in Litigation And Investigations

Here's an overview of how best to design and conduct your reviews to get to the facts in a smarter and faster way.

Early Case Assessment: Understanding the Universe Of Data

Not all cases are created equally. It's easy to forget this when you consider the main similarity most major cases share, internal investigation and litigation matter alike: the millions, if not hundreds of millions of documents, that need to be reviewed.

When faced with such large quantities of data, many corporate counsel and outside counsel want to immediately dive into the review and push through as quickly as possible with as many people as possible. While this is understandable, it is not the best way to approach the process. The first step is to take an extra few days ' maybe even a week, if your timeline allows for it ' to conduct a thorough early case assessment (ECA) with outside parties and to carefully assess and understand the particulars of your case. But you can't fully understand the size and scope of the case, build a project map or manual, or factor in things like your budget, schedule, and legal objectives without first understanding your data. Before anything else, it's important to make sense of the universe of data within the case, at least at a high level ' which players are involved, their relationships to one another and to the case, and the type of data you're dealing with.

But where to begin? There's nothing more daunting than dealing with a seemingly unwieldy mass of data. You might even know what you're looking for, but how do you focus on sorting, filtering, and organizing your data in a way that makes sense and is logical and intuitive? This is where technology comes into play.

In an internal FCPA investigation we managed, for example, we were asked to assist a client in identifying whether sales people, distributors, and other employees in foreign countries were bribing government officials to increase sales. We had to identify key people, get a sense of their relationships with others inside and outside the organization and their potential level of involvement in the matter, and to get an overview of the surrounding facts before jumping into a full review.

Using technology on our review platform, we clustered documents by topics and grouped related documents across concepts that were potentially relevant to the matter. Another tool on the platform allowed us to immediately find high-volume communications between people we had already identified as key, and between previously unknown parties. Setting up this groundwork made running thorough, targeted searches a much easier proposition.

Most review platforms have similar tools. You should work with your outside counsel and other parties to ensure that you're using them to organize your data as efficiently as possible before review even begins.

Analyzing the Data: Tools And Approaches

The proper and exhaustive use of technology to cull and organize the data is a far more desirable and efficient option than simply directing an army of lawyers to sift through data. Technology is the ideal risk mitigation tool ' and it can be extraordinarily effective when you know when to use what tools and when you implement an approach that treats technology as the cornerstone of the discovery process.

Say you've conducted a thorough early case assessment, used technology to sort and filter your data, and have a good understanding of what you're looking for. The next step is then to cull down the amount of information that needs to be manually reviewed, while still following a process that is intuitive and defensible in court.

One traditional method involves building search strings that incorporate the information and knowledge that you have and running these searches across the potential data universe. A common mistake is stopping the process there and just reviewing everything that hits on these search terms. Instead, at a minimum, this should be an iterative process whereby you sample the results of the initial searches and refine the terms to hone in on the relevant documents and to narrow down the scope of review. In addition, the search term refinement process should be combined with advanced analytics tools to help further identify relevant terms and other ways to reduce the document population for review.

Another more advanced method of analyzing the data and approaching review is through the use of predictive coding. This can be used in many ways to analyze and cull data. Of course, in order to use predictive coding, you must have an understanding of the issues, or at least what is relevant in the matter, as it requires an initial review of statistically valid samples of documents. Once the sample documents have been coded, the system uses this information to translate this coding to other documents with similar characteristics.

In a recent litigation matter that we worked on, predictive coding allowed us to cull a significant number of documents. By coding different subsets of data within the initial set of documents ' around 1.5 million ' we were able to determine with a high degree of confidence that at least 20% of them were irrelevant to the matter. This process allowed us to get to the relevant documents much faster, increase the quality of review, and to get vital information to counsel more quickly than through linear review.

Letting Facts Dictate Review

Similar to developing search terms, the best way to tackle review is to build an iterative process that evolves as you find key, operative facts and information. It doesn't make sense to wait until the end of review to begin piecing together the facts and creating a narrative about what is going on. In litigation, for example, being able to gather and analyze facts at the outset and in the early stages of review can shape the entirety of your strategy. If, over the course of review, you find that the facts being uncovered overwhelmingly point to your culpability, you might decide it's a wiser move to just go ahead and settle, rather than letting the case drag on in court.

Fact development can also play a major role in controlling the amount of data you need to review. This was especially true in the FCPA case I mentioned previously. After we reviewed an initial set of 50,000 documents (out of over two million), we confirmed that the only relevant and key documents were email conversations. We therefore removed all documents unrelated to e-mail, and refined our search terms to better reflect the facts we were already gathering. This had an enormous effect on the review: we were able to bring the total volume of data down to 400,000 documents and minimize the amount of unnecessary review.

Work closely with your outside counsel and other legal partners during reviews to ensure that you're building an iterative search process and using tools like predictive coding to narrow down the amount of documents that need to be reviewed. Getting to the important facts as quickly and efficiently as possible is paramount in developing a smart strategy.

Shaping Future Reviews and Real-Time Compliance

Large-scale e-Discovery and document reviews yield tremendous amounts of data and insight. It would be wrong to think that the data you analyze and become familiar with in a given case can only be understood and used in the context of that same case. Rather, data gleaned from past reviews can be enormously helping in informing and shaping future reviews ' and in building compliance programs that utilize fact development as the first step in identifying and solving issues as they arise.

The most obvious application of data from past reviews is to keep it readily accessible for future use. For example, a government investigation into off-label promotion of a product could lead to a shareholder class action, as well as product liability matters all related to the same product and involving similar issues. If you engaged in a full-scale review during discovery for the initial investigation matter, you could re-use that same data and all of the information gleaned from the initial review regarding manufacturing, marketing and sales practices.

In the situation described above, one way that you could use data previously collected is to look through the initial data set to see if there are any mentions of the plaintiff in a product liability matter or any instances of off-label promotion that could be possibly related. The company would obviously know the patient's location and would have sales records and marketing reports from that region. By running searches across the existing data for the five to 10 custodians who may have been involved in promoting the product in that region, you could try to find what doctors, hospitals, and patients may have been involved and impacted.

Even if this process does not evolve exactly in the manner described above, this is still how companies should be treating legacy data. Documents collected for each matter should be adequately stored for future matters in the event that these documents or subsets of these documents become relevant again. Re-using and re-purposing data can only help you become more familiar with your own company, your organization's lexicon, the way you do business, and the issues that affect you across business functions.

Moving Beyond Review

Fact development is not just about looking for and interpreting facts on a one-off basis as an individual case arises. Instead, companies should treat their databases like a work in progress ' something to be turned to in times of need but also something that should grow and be kept up-to-date.

Instead of waiting until an issue builds to the point of mandated or court-ordered discovery (when data sizes will inevitability be much higher) companies should engage in a regular analysis of their data and tackle points of suspicion immediately. Discovery should become a continuous, institutional process; literally “discovering” issues as they emerge and analyzing them. The end goal should be to build an internal compliance system that allows companies to track their electronic communications in real time and flag-up suspicious behavior for immediate review.

And how about using data for non-compliance reasons or turning the legal division into a profit center instead of just a cost center? Imagine being able to take all the knowledge you have gleaned during reviews ' for example, which sales and marketing practices and individuals are the most likely to cause issues downstream ' create reports, and sell it back to different internal business units. Or say you have an enormous amount of data surrounding your sales initiatives. Why not examine your strategies, see what's working and what's not, and share your learnings with your marketing department?

Facts are stubborn things, John Adams once said. They're also incredibly useful.

Laura Jungels is an assistant vice president at Clutch Group, where she manages key client relationships and litigation and investigation matters. Prior to Clutch, Ms. Jungels was with Abbott Laboratories, where she worked closely with eDiscovery vendors and outside counsel to implement new strategies and processes for discovery.

Over the past decade, the volume of data in litigation and investigations has exponentially increased. As data has become more and more vulnerable to subpoenas and regulators, people have become singularly concerned with making relevance decisions for production. In the process, we have forgotten about focusing on the facts themselves and uncovering the stories within data. The art of fact development ' the uncovering of essential data to develop case narratives ' has been lost, and it needs to be brought back again.

A common sentiment seems to be that focusing on individual facts just isn't possible due to high volumes or tight deadlines. But not only is it possible ' it's essential for building the narrative of your case. There are many benefits to incorporating fact development early on in the review process. It can lead to a more streamlined, efficient review. It can substantially reduce the volume of data you need to analyze. Furthermore, organizing and harnessing the data you uncover can have reverberating effects across current and future reviews.

Rather than waiting until the end of review to consider all the facts, corporate counsel should place a stronger emphasis on conducting fact development in a more proactive manner and on building an early case assessment process that is designed to prioritize data analysis as review unfolds. Instead of waiting for outside counsel to conduct fact development and treating it as an end product, corporate counsel should build fact development into the actual review process itself ' and before review even begins.

Here's an overview of how best to design and conduct your reviews to get to the facts in a smarter and faster way.

Early Case Assessment: Understanding the Universe Of Data

Not all cases are created equally. It's easy to forget this when you consider the main similarity most major cases share, internal investigation and litigation matter alike: the millions, if not hundreds of millions of documents, that need to be reviewed.

When faced with such large quantities of data, many corporate counsel and outside counsel want to immediately dive into the review and push through as quickly as possible with as many people as possible. While this is understandable, it is not the best way to approach the process. The first step is to take an extra few days ' maybe even a week, if your timeline allows for it ' to conduct a thorough early case assessment (ECA) with outside parties and to carefully assess and understand the particulars of your case. But you can't fully understand the size and scope of the case, build a project map or manual, or factor in things like your budget, schedule, and legal objectives without first understanding your data. Before anything else, it's important to make sense of the universe of data within the case, at least at a high level ' which players are involved, their relationships to one another and to the case, and the type of data you're dealing with.

But where to begin? There's nothing more daunting than dealing with a seemingly unwieldy mass of data. You might even know what you're looking for, but how do you focus on sorting, filtering, and organizing your data in a way that makes sense and is logical and intuitive? This is where technology comes into play.

In an internal FCPA investigation we managed, for example, we were asked to assist a client in identifying whether sales people, distributors, and other employees in foreign countries were bribing government officials to increase sales. We had to identify key people, get a sense of their relationships with others inside and outside the organization and their potential level of involvement in the matter, and to get an overview of the surrounding facts before jumping into a full review.

Using technology on our review platform, we clustered documents by topics and grouped related documents across concepts that were potentially relevant to the matter. Another tool on the platform allowed us to immediately find high-volume communications between people we had already identified as key, and between previously unknown parties. Setting up this groundwork made running thorough, targeted searches a much easier proposition.

Most review platforms have similar tools. You should work with your outside counsel and other parties to ensure that you're using them to organize your data as efficiently as possible before review even begins.

Analyzing the Data: Tools And Approaches

The proper and exhaustive use of technology to cull and organize the data is a far more desirable and efficient option than simply directing an army of lawyers to sift through data. Technology is the ideal risk mitigation tool ' and it can be extraordinarily effective when you know when to use what tools and when you implement an approach that treats technology as the cornerstone of the discovery process.

Say you've conducted a thorough early case assessment, used technology to sort and filter your data, and have a good understanding of what you're looking for. The next step is then to cull down the amount of information that needs to be manually reviewed, while still following a process that is intuitive and defensible in court.

One traditional method involves building search strings that incorporate the information and knowledge that you have and running these searches across the potential data universe. A common mistake is stopping the process there and just reviewing everything that hits on these search terms. Instead, at a minimum, this should be an iterative process whereby you sample the results of the initial searches and refine the terms to hone in on the relevant documents and to narrow down the scope of review. In addition, the search term refinement process should be combined with advanced analytics tools to help further identify relevant terms and other ways to reduce the document population for review.

Another more advanced method of analyzing the data and approaching review is through the use of predictive coding. This can be used in many ways to analyze and cull data. Of course, in order to use predictive coding, you must have an understanding of the issues, or at least what is relevant in the matter, as it requires an initial review of statistically valid samples of documents. Once the sample documents have been coded, the system uses this information to translate this coding to other documents with similar characteristics.

In a recent litigation matter that we worked on, predictive coding allowed us to cull a significant number of documents. By coding different subsets of data within the initial set of documents ' around 1.5 million ' we were able to determine with a high degree of confidence that at least 20% of them were irrelevant to the matter. This process allowed us to get to the relevant documents much faster, increase the quality of review, and to get vital information to counsel more quickly than through linear review.

Letting Facts Dictate Review

Similar to developing search terms, the best way to tackle review is to build an iterative process that evolves as you find key, operative facts and information. It doesn't make sense to wait until the end of review to begin piecing together the facts and creating a narrative about what is going on. In litigation, for example, being able to gather and analyze facts at the outset and in the early stages of review can shape the entirety of your strategy. If, over the course of review, you find that the facts being uncovered overwhelmingly point to your culpability, you might decide it's a wiser move to just go ahead and settle, rather than letting the case drag on in court.

Fact development can also play a major role in controlling the amount of data you need to review. This was especially true in the FCPA case I mentioned previously. After we reviewed an initial set of 50,000 documents (out of over two million), we confirmed that the only relevant and key documents were email conversations. We therefore removed all documents unrelated to e-mail, and refined our search terms to better reflect the facts we were already gathering. This had an enormous effect on the review: we were able to bring the total volume of data down to 400,000 documents and minimize the amount of unnecessary review.

Work closely with your outside counsel and other legal partners during reviews to ensure that you're building an iterative search process and using tools like predictive coding to narrow down the amount of documents that need to be reviewed. Getting to the important facts as quickly and efficiently as possible is paramount in developing a smart strategy.

Shaping Future Reviews and Real-Time Compliance

Large-scale e-Discovery and document reviews yield tremendous amounts of data and insight. It would be wrong to think that the data you analyze and become familiar with in a given case can only be understood and used in the context of that same case. Rather, data gleaned from past reviews can be enormously helping in informing and shaping future reviews ' and in building compliance programs that utilize fact development as the first step in identifying and solving issues as they arise.

The most obvious application of data from past reviews is to keep it readily accessible for future use. For example, a government investigation into off-label promotion of a product could lead to a shareholder class action, as well as product liability matters all related to the same product and involving similar issues. If you engaged in a full-scale review during discovery for the initial investigation matter, you could re-use that same data and all of the information gleaned from the initial review regarding manufacturing, marketing and sales practices.

In the situation described above, one way that you could use data previously collected is to look through the initial data set to see if there are any mentions of the plaintiff in a product liability matter or any instances of off-label promotion that could be possibly related. The company would obviously know the patient's location and would have sales records and marketing reports from that region. By running searches across the existing data for the five to 10 custodians who may have been involved in promoting the product in that region, you could try to find what doctors, hospitals, and patients may have been involved and impacted.

Even if this process does not evolve exactly in the manner described above, this is still how companies should be treating legacy data. Documents collected for each matter should be adequately stored for future matters in the event that these documents or subsets of these documents become relevant again. Re-using and re-purposing data can only help you become more familiar with your own company, your organization's lexicon, the way you do business, and the issues that affect you across business functions.

Moving Beyond Review

Fact development is not just about looking for and interpreting facts on a one-off basis as an individual case arises. Instead, companies should treat their databases like a work in progress ' something to be turned to in times of need but also something that should grow and be kept up-to-date.

Instead of waiting until an issue builds to the point of mandated or court-ordered discovery (when data sizes will inevitability be much higher) companies should engage in a regular analysis of their data and tackle points of suspicion immediately. Discovery should become a continuous, institutional process; literally “discovering” issues as they emerge and analyzing them. The end goal should be to build an internal compliance system that allows companies to track their electronic communications in real time and flag-up suspicious behavior for immediate review.

And how about using data for non-compliance reasons or turning the legal division into a profit center instead of just a cost center? Imagine being able to take all the knowledge you have gleaned during reviews ' for example, which sales and marketing practices and individuals are the most likely to cause issues downstream ' create reports, and sell it back to different internal business units. Or say you have an enormous amount of data surrounding your sales initiatives. Why not examine your strategies, see what's working and what's not, and share your learnings with your marketing department?

Facts are stubborn things, John Adams once said. They're also incredibly useful.

Laura Jungels is an assistant vice president at Clutch Group, where she manages key client relationships and litigation and investigation matters. Prior to Clutch, Ms. Jungels was with Abbott Laboratories, where she worked closely with eDiscovery vendors and outside counsel to implement new strategies and processes for discovery.