Moving at the Pace of TAR
Since first winning judicial approval in 2012, technology-assisted review (TAR) has continued to build momentum in the courts. However, despite case studies demonstrating the cost savings possible with TAR, that momentum has not been matched by its adoption among litigating organizations. A report by 451 Research from late 2014 indicates that TAR use rates hovered between 10 and 15 percent in all litigation matters; another survey indicated that TAR adoption for cases that reach the review stage is more likely between 5 and 7 percent. With organizations facing the dilemma of too much data and too little time to search and review it, how can that be the case?
It is clear that TAR can significantly reduce the volume of data that parties must review, whether from their own data stores or from large discovery productions by other parties. However, it has many other applications. For example, TAR can be a useful tool for quality assurance in discovery. TAR can improve information governance by minimizing the over preservation of documents, categorizing them into meaningful buckets for storage or deletion. In addition, TAR has applications in compliance and cybersecurity, assisting organizations to understand their data—including unstructured data—and identify data with a high-risk profile.
However, unlike other software, TAR is not an out-of-the-box solution that lawyers can simply add to their arsenal and immediately begin using. Given that lawyers are already cautious to adopt new technology, it is no surprise that a tool that moves the process of assessing information light-years ahead could be slow to penetrate the market. Although many lawyers now understand the basic concept behind how the software works from its use in a growing number of high-profile cases, its technical intricacies remain a mystery to some.
As a result of its complexity, implementing TAR requires expert guidance in many cases, including appropriate use cases for the specific task at hand. In addition, TAR use may require the expertise of TAR consultants, statisticians and linguists to work closely with subject matter experts to test, refine, and validate the efficacy of search terms, and generate statistically valid measurements of the precision and recall of the results on the seed set of documents used to train the TAR algorithm (which legal teams must be able to do). Depending on the flexibility the tool offers, experts can guide and customize precision and recall, margins of error and statistical variables for clients’ specific needs, yielding more accurate and cost-effective results. Then, once implemented, expert testimony may be required to defend the chosen TAR protocol. For instance, parties must be able to support their decision to exclude documents that fall below a certain threshold from review.
The key is for legal teams to realize they do not have to go it alone. Rather, they need to recognize that TAR is not merely a technology; it is a strategy whose results can be optimized when backed by the proven processes and expert guidance of a seasoned service provider.
Bill Schiefelbein is senior vice president, e-discovery consulting at Conduent. He can be reached at info@conduent.com.