Beginning Next Week: InsideCounsel will become part of Corporate Counsel. Bringing these two industry-leading websites together will now give you comprehensive coverage of the full spectrum of issues affecting today's General Counsel at companies of all sizes. You will continue to receive expert analysis on key issues including corporate litigation, labor developments, tech initiatives and intellectual property, as well as Women, Influence & Power in Law (WIPL) professional development content. Plus we'll be serving all ALM legal publications from one interconnected platform, powered by, giving you easy access to additional relevant content from other InsideCounsel sister publications.

To prevent a disruption in service, you will be automatically redirected to the new site next week. Thank you for being a valued InsideCounsel reader!


3 steps to becoming a "predictive coding guru"

Does predictive coding usher in the man vs. machine debate?

National Public Radio recently said that predictive coding could let "one attorney do the work of 500." Despite the apparent hyperbole, it’s still easy to get excited about new technological advances. Perhaps more important, though, is to divine what this tectonic shift might mean for practicing attorneys in reality. Several questions naturally arise, including what will the face of document review look like over the next decade, and how will this new landscape affect the practice of law? 

More than a decade ago, even before electronic discovery became mainstream, many in the legal field saw another development underway: a gradual shift away from using hard copy documents for managing discovery toward converting those same documents into electronic images. The trend developed because converting paper documents to searchable image files with scanning and optical character recognition software was simply far more cost-efficient and manageable than dealing with paper. While this paradigm shift dramatically impacted the litigation playing field, it didn’t threaten to steal billable hours like the type of automation seen on the horizon with new technology like predictive coding.

Predictive coding is a type of machine learning technology that enables a computer to automatically predict how documents should be classified based on a limited, but significant, level of human input. The technology is exciting for legal departments attempting to manage skyrocketing litigation budgets because the ability to automatically rank and then “code” or “tag” electronic documents based on criteria such as relevance and privilege has the potential to save companies millions in e-discovery costs.

The potential cost savings of predictive coding technology are so great that corporate legal departments inevitably will expect the law firms that represent them to learn how to use the technology. Much like the shift from paper to electronic file conversion was driven by the potential for cost savings, so too is the buzz surrounding predictive coding technology. 

While the days of “brute force,” linear document review are rapidly disappearing, the role of savvy e-discovery practitioners is still vital, and predictive coding provides a land of opportunity for those eager to capitalize on the trend. The emerging body of case law (Da Silva Moore, Kleen Products and Global Aerospace) shows that there is no shortage of work required to negotiate and craft protocols for the use of this new technology, even before the review process begins.

After protocols for sampling and training are sorted out, there’s still a need for skilled attorney reviewers (likely billing at higher hourly rates) to tag and code documents so the system can be accurately trained. Similarly, the role of attorneys in evaluating the quality of the computer’s decisions is increasingly important since most believe that the documents designated as highly relevant by the computer should undergo a final human review prior to production.

In order to capitalize on new technology trends like predictive coding, though, it’s important to embrace the new changes instead of taking an ostrich approach. While the task might seem daunting, the complexity surrounding early predictive coding tools has created an opportunity to learn a new method that many less-enterprising practitioners may foolishly pass up. That means the area is still ripe with opportunity—and the good news is that becoming an organization’s go-to person for predictive coding does not necessarily require obtaining an advanced degree in statistics or computer science.

For a simpler approach to reaching “predictive coding guru” status start with these three basic steps:

1. Read up on the topic

There is a tremendous amount of information available on the Internet that covers everything from predictive coding 101 to recent case law. Reading articles from trusted sources will help you understand predictive coding at a high level, particularly surrounding the role of inside/outside counsel.

These resources provide a good primer:

2. Understand the market dynamics and tools

The predictive coding software market is flooded with numerous companies claiming to provide “leading” predictive coding solutions. The challenge with any burgeoning market is to separate the players from the poseurs.

The good news is that independent industry analysts like Gartner prepare several reports annually that evaluate e-discovery market trends, so there’s no need to start from scratch. Gartner’s 2012 “Magic Quadrant for E-Discovery Software” report is a particularly good starting point for understanding current e-discovery trends like predictive coding, as well as the major players in the broader e-discovery and information governance space. 

3. Test drive the software

The next step for gathering information about predictive coding technology should include observing live product demonstrations. Specific things to look for in any predictive coding product include transparency and ease of use for end users, recommended workflow, chain of custody reporting and broader integration with other important e-discovery modules. Finally, be wary of e-discovery providers suggesting that predictive coding technology tools will replace all other e-discovery technologies. Providers belittling the broader e-discovery and information governance picture probably don’t offer much beyond predictive coding technology, so they may try to focus the conversation exclusively on document review and analysis.

Other valuable steps in the e-discovery process that must be accounted for (in addition to document review) include proactively establishing an internal records retention schedule, legal hold notification, data identification and collection, early case assessment and data filtering. 


For some outside law firms and contract attorney service bureaus, it’s been all too easy to become addicted to the high utilization associated with e-discovery document review projects. But, for inside counsel, those massive numbers of e-discovery hours have become an easy target, at which the latest predictive coding technology is squarely aimed.

And yet, the demise of attorneys in the review process has been vastly overstated. Savvy legal professionals interested in advancing their careers and being on the cutting edge of something new will invest time in understanding the future of predictive coding or risk becoming less relevant.

author image

Dean Gonsowski

A former litigator, general counsel and associate general counsel, Dean Gonsowski is the global head of information governance at Recommind and has more than 15...

Bio and more articles

Contributing Author

author image

Matt Nelson

Matthew Nelson is an attorney and legal technology expert with Symantec who has spent the past 15 years helping organizations address a wide array of...

Bio and more articles

Join the Conversation

Advertisement. Closing in 15 seconds.