In mid-2013, our small team of trial specialists had just begun defending the City of Fort Bragg, California in a complex environmental case. Our opponent, deep-pocketed Georgia Pacific and their 700-lawyer firm, was formidable. To build our case and prove the City was not responsible for contamination at a 120-year-old lumber mill, we needed to review more than 11 million documents in a wide variety of file formats spanning 12 decades. Millions of dollars in clean-up costs the City and its taxpayers couldn’t afford were at stake. Adding to the challenge, our trial deadline left us very little time to get the job done.
There was no way filling a room with a team of lawyers to do the old linear e-discovery process was going to work. It was clear that we needed to bravely go where few law firms had gone before.
Selling it Internally
My partners and I had been hearing a lot about predictive analytics from colleagues and the media. We began speaking with vendors about using the technology to tackle e-discovery in this case. As we dug deeper, we learned that outsourcing predictive analytics carried too high a price—both in terms of time and money. On paper, it was obvious we could reduce costs and gain control by bringing our own predictive analytics program in-house.
But first, the equity partners needed to agree.
We all know that law firms are risk averse and attorneys are prone to test every premise through strenuous argument and debate. So this topic temporarily converted our conference room into a courtroom. At first, many of the partners were extremely resistant to the idea of doing predictive coding in-house. Some thought of it as a “black box” technology due to the software’s reliance on proprietary algorithms. Many were concerned we would not get a return on our investment. Fortunately, cooler heads prevailed. The numbers didn’t lie.
Instead of paying a vendor $300,000 to get the job done less efficiently, we decided to purchase and implement our own system for about $22,000. When we considered software licensing fees; paralegal, attorney, litigation support training, and data migration, the total investment was around $45,000—a significant expenditure for a firm of our size. But we were able to defray $23,000 of the costs by partnering with our client, who was saving about $300,000 he would have paid to an outside vendor for processing, searching, and storing the data.
The vendor review and selection process had to happen very quickly. We considered three products:
- Concordance from LexisNexis.
- Relativity from kCura.
- Summation Pro from AccessData.
We selected Summation Pro after seeing technical demonstrations of all three products and speaking with references from AccessData who had gone through Summation Pro installations.
Summation Pro was attractive because it combined rich linear human coding tools and powerful search functions with a predictive analytics feature at a more reasonable cost than the other vendors. In addition, we already had an established relationship with AccessData through our existing investment in the company’s Summation iBlaze software. In the end, we determined the other software products were more complex than what we needed; the costs for implementation, maintenance, and licenses were too high.
After getting buy-in from the partners, teaching personnel to use the software was challenging and required several weeks of intensive training. Because the firm previously invested in a Private Cloud solution with virtualization and network storage technologies—and Summation Pro had the least demanding server requirements of the three vendors we considered—we were able to easily allocate the necessary server and data storage resources.
We were already deep into manual review on the Fort Bragg case, so one of the big hurdles in the beginning was migrating the data from iBlaze to Summation Pro. With the help of AccessData, we were able to accelerate that process by implementing version 5.0 of Summation Pro, which at the time was a recent release with significant changes from the previous version. After training and deployment we discovered some bugs that affected our ability to produce documents. We resolved these issues when we upgraded again, this time to version 5.1 of Summation Pro.
What We Learned
In retrospect, while we did move as quickly as possible, we wish we had implemented this technology sooner to reduce our reliance on outside vendors. By removing the “middle-man” vendor we are able save clients hundreds of thousands of dollars. But despite these jaw-dropping savings, many of our clients are still not aware of this new technology. For the clients who have embraced the idea of using predictive analytics for e-discovery, we could have better prepared them to think about how we have to shoulder hidden costs in the form of training and software fees. So we have had to expand our billing model to include set-up and maintenance of the case that goes beyond traditional litigation support.
In addition, we should have thought more about how moving from linear review to technology aided review causes a shift, but not entirely a reduction, in workload. Although the need for billing personnel is reduced, the need for technologists is increased. Moreover, the former typically demand lower compensation and are, by definition, billable. The latter demand relatively higher compensation and may not be billable absent a specific agreement with the client.
While predictive analytics is a powerful tool, the technology should not be considered a panacea. Some level of linear review by smart lawyers will be necessary into the foreseeable future. On the positive side, we now have the capability to narrow and train the software to search large document caches faster while simultaneously reducing costs. We are now better equipped to compete with AmLaw 500 firms to handle large pieces of litigation for Fortune 500 clients… which brings us back to the City of Fort Bragg.
By using predictive analytics and our deep environmental litigation experience, we obtained a dismissal with prejudice of Georgia-Pacific’s case against the City. As a result, the City will not be required to pay any portion of the clean-up costs or damages.
We brought predictive analytics in-house when very few law firms were adopting this technology and knew we did the right thing when predictive analytics helped us win the Georgia-Pacific case. As icing on the cake, we recently received the 2015 Legaltech Innovation Award for Most Innovative Use of Technology in a Law Firm for taking this step at a time when few law firms were adopting the technology. When achieved in the face of great risk and adversity, the taste of victory is all the more sweet.