Litigation is not only costly but also extremely time consuming, especially when it comes to document review and evidence procurement. But in recent years, technological tools have enabled e-discovery to speed up the process and reduce expenses.
“When it comes to litigation, studies indicate that 60 to 90 percent of the cost is discovery,” says Lael Andara, a partner at Ropers Majeski Kohn and Bentley PC. “The person who understands the tools and adopts the technology can try the case for significantly less, and can get to the key information and be in a better position to resolve the dispute sooner.”
Smart Business spoke with Andara about using e-discovery tools and when they are an appropriate method to reduce the time and cost of a lawsuit.
What is e-discovery?
E-discovery is the same as discovery, with the recognition that business data is digital. Business data is 99.9 percent computer generated and therefore, so is all potential evidence.
In the past 10 years business data have evolved exponentially into the digital form as Electronically Stored Information (ESI), allowing for an exponential growth of data. A recent Wall Street Journal article explained that at this time in our history we are generating more data every 10 minutes than every written work and image created from the dawn of time until 2003.
To deal with this there are new technology tools in the form of predictive coding, or technology-assisted review (TAR), that ferret out the key information in pending disputes in less time and for less cost than the typical linear review previously used in the legal profession prior to ESI.
Predictive coding or TAR is already being embraced by different industries, especially those that are web based, because of how it maximizes sales. For instance, when you purchase a song from iTunes, the site shows other music that you might like based on your previous choices; it predicts your next purchase. When you buy groceries, coupons on the back of your receipt are based on the purchases that you made to increase the chance of generating business based on the preferences revealed in your current purchase.
The litigation profession is essentially doing the same thing to minimize costs. For example, if you are the plaintiff in a contract dispute, you probably only have a few key documents: invoices, the contract(s) and emails. You can take that information set, called a seed set, and plug it into the computer because you know those are the most relevant documents. The computer, based on an algorithm, will find all other documents that are similar in nature, which gets to the heart of the relevant evidence much quicker than a full review of all company records by an attorney. Predictive coding is an artificial learning that leverages computing capability to decrease the cost and time that often stands in the way of early case assessment or resolution.
Also, if you put in a seed set of five documents and the program brings back 100, you can tell it that 80 were not relevant and 20 were relevant, thereby allowing it to learn based on the expanded seed set. Predictive coding becomes more intelligent the more you teach it which data is on target and which is not, much like Netflix recommendations become more accurate the more content you watch.
How does this technology lower cost during litigation?
Cases today can require gigabytes or terabytes of information to be reviewed for responsive documents to be produced in litigation and ultimately narrowed further to several hundred or thousand pages that you will actually use as exhibits at trial. First and foremost, you have to narrow the universe of information that your attorney reviews because attorney rates of $200 to $600 per hour drive the cost of litigation. Predictive coding can narrow attorney review time from months to weeks or even days; time is money, and volume is time.
A business owner needs to understand the big-picture cost of TAR. For example, predictive coding may cost $750 per gigabyte, which initially may appear expensive, if you are dealing with a terabyte of data. However, instead of the attorney spending months reviewing documents, the information using TAR can be narrowed down to a manageable subset with the attorney spending only days or weeks reviewing the documents — a huge cost savings. Even greater than the cost savings, predictive coding gets you to the facts of your case more quickly so that you can make more informed decisions in terms of offering a reasonable settlement, if necessary.
When is predictive coding the appropriate method to use in document discovery?
Court rules require that the cost of discovery be proportionate to what’s at issue. For example, if you have a contract dispute with $1,000 at stake, predictive coding is probably not the best use of your time or money. But if you have a case such as the recent Oracle-Google case over the Android phone’s use of JAVA, in which there was a $1 billion demand and the volume of data was immense, then predictive coding is necessary. If you have a case that’s in the tens of thousands of dollars, you should at least consider predictive coding or some other version of technology-assisted review if the volume is significant.
Lael Andara is a partner in the Intellectual Property and Technology Group and chair of Electronics Services Protocol/Discovery ESP at Ropers Majeski Kohn and Bentley PC. Reach him at (650) 780-1714 or firstname.lastname@example.org.
Insights Legal Affairs is brought to you by Ropers Majeski Kohn & Bentley PC