When I took on an e-discovery role that required approaching document review without the help of a hundred contract attorneys, I found myself looking for ways to innovate. This led me to text analytics, which has sped up review for me in so many ways. Although analytics is becoming more common, many attorneys and case teams are still hesitant to use it. Why?
One major reason is that they’re not used to the technology. They don’t understand how it can work for them. With high stakes, they’re worried that if human eyes aren’t on every document, they won’t be coded correctly. People also steer away from analytics because they assume that if they don’t have five million records, they don’t need it.
How can you get your case team on board with analytics? First, explain and demonstrate the time and cost savings possible for cases of all sizes, then make it as easy as possible for them to get up and running.
Prove a Little Time Up Front Means a Lot Less Time Later
To demonstrate the benefits of analytics, I ask attorneys to let me spend one hour in their data. In a recent case, the team insisted they needed to start reviewing right away to make a very close deadline. I suggested we put some front-end time in before starting the review to index and run a few analytics features on the data. Because we didn’t have a full day to sit down and get everything ready, I planned for five or six hours over three days to get them up and running. The attorney was frustrated with this approach because she thought we were going to waste three days of review. What was critical to explain to her and the case team was that all of the time we were spending organizing and indexing the data, with tools such as clustering and email threading, was going to save time in the end. Although they were resistant, they agreed with our analytics plan. In the end, they saw the proof: we knocked out the review in 4 weeks when they had estimated it would take 16.
The attorney was skeptical of the results because we got through the review so fast. I took this as an opportunity to jump in and show her how we can use analytics to QC the work, as well. We created clusters from the different sets of coding—responsive, non-responsive, and privileged documents—so she could do spot checks, while also allowing her to find what was conceptually similar. The review was efficient and accurate, and now she’s a believer.
Emphasize That Case Size Doesn’t Matter
You can also find opportunities to demonstrate that case size doesn’t matter when it comes to analytics. Even if you’re going to manually review a thousand records and look at every document linearly, from one to one thousand, you can still use analytics to do it faster, more efficiently, and more consistently by organizing and prioritizing your data—based on email threads and concepts, for example—from the start.
In one case, the team had about 1,000 documents to review, but didn’t think they had enough time with the deadline that was provided. I asked for one hour to show how tools in analytics can help. I used everything that I could—from standard review features like search term reports and persistent highlighting, to analytics like clustering and email threading. Additionally, since they were reviewing for FOIA and litigation, I created one review layout covering all coding issues and categories. We organized the documents in batches that grouped conceptually similar documents, and pulled in families and email threads. Combining these analytics tools with customizable review options was crucial to saving time.
It goes back to the cost and time benefit. You have to show case teams that you’re going to cost a few hours in the beginning, but you’ll save them more time in the end.
Use Familiar Concepts
The other large element is getting people familiar with the technology and comfortable with the control it provides. I love having control over the software, but some find it overwhelming. Part of our job is providing hands-on customer service so that it doesn’t feel overwhelming. If I’m doing my job properly, users should be able to get to whatever data they need in about three clicks.
The goal is to make analytics familiar to users. I use my folders and subfolders in Outlook to explain the field tree and clusters, and email conversation groups to explain email threading. I explain that if I use email threading and clustering holistically like I use folders and conversations in Outlook, I can get through the data set a lot faster than if I’m going through every single document incrementally.
Almost everybody uses Outlook, making it an easy way to show how familiar analytics can look and feel. It’s important to consider what case teams are familiar with, and use that concept to train them in analytics.
If we continue to educate case teams, even if it’s in bits and pieces, we’ll get buy-in. e-Discovery is a customer service industry, and at the end of the day, our goal is to help people. The more we can teach people by demonstrating how we can use technology to help them, the more success we’ll have.