Avoiding the “Shiny Object” Syndrome in E-Discovery Technology Purchases

06
Jan
2014

Peter Mercer

Here's a dirty little secret in the e-discovery industry: Many practitioners don't really understand the right questions to ask when evaluating and purchasing e-discovery technology and services. Although most people enter the purchase process with specific objectives in mind, such as reducing costs, eliminating mistakes and improving staff efficiency, they are easily swayed by vendors touting the latest “advanced” features and functionality, and by whatever new “shiny object” is currently generating buzz in the marketplace. 

Many practitioners begin the search for better technology with a set of questions focused on a group of “must have” features, such as technology-assisted review (TAR) or predictive coding (PC), that deliver some form of advanced analytics, artificial intelligence or machine learning with the promise of lowering e-discovery costs. Just about anyone involved with e-discovery has heard about TAR and PC, whether from a vendor's sales team, on a blog or in an industry publication. What most people fail to realize is that these analytics-based technologies alone may not fully address the e-discovery challenges faced by organizations on a daily basis. While it's true that some practitioners can derive meaningful cost and efficiency benefits from analytics-based technologies, it's typically only realized on the largest, most complex cases. It's the organizations with the highest litigation volumes that are likely to realize those benefits with any regularity.

Most practitioners ultimately want solutions that will help them lower the total cost of ownership (TCO) of their e-discovery investments (e.g. time, cost & risk). In order to successfully address TCO and avoid the “shiny object” syndrome, legal teams need to ask the right questions when evaluating e-discovery technology and services. Instead of beginning inquiries by approaching vendors with a set of preconceptions about the technology needed, it is much more useful to ask about the kinds of legal matters the organization faces most frequently. These questions include:

• What kinds of data and file types are typically involved in the average cases? 
• How much data is processed in an average case? 
• How are data collections typically performed? 
• Do the people doing the work have broad experience collecting, searching and culling the data typically encountered? 
• Who in the organization is in charge of cleaning up and de-duplicating the data?  
• What types of production sets are required at the end of typical projects? 
• Is the legal team experienced and able to move beyond the basic e-discovery tasks?

Remember, TCO is not determined just by the costs of acquiring technology but also by the operating costs associated with that the use of that technology over time and across matters. The more data pushed through the various phases of the e-discovery lifecycle, the more expensive and time-consuming it will be. While most e-discovery technologies can reduce data volumes to some extent, it's usually done late in the game-after collection at the review stage-which rarely saves enough money to justify the purchase of an expensive technology to perform the task. In fact, simply culling down the data early in the e-discovery process before collection can have a much greater impact on costs than new technologies, such as CAR and PC. 

Instead of relying on analytics-based technologies relatively late in the e-discovery process, less expensive forensic search tools, which are easier to operate, may hold the answer for many organizations. Forensics search tools allow organizations to narrow the data set early in the game without the need to hire expensive specialists. Newer forensic search tools enable legal teams to quickly analyze email and cell phone data, instantly visualizing relationships and timelines, as well as attachments, embedded images, archives, headers and metadata. They can also preview, cull and de-duplicate data, as well as export it in appropriate production formats that are compatible with other e-discovery platforms. This capability is critical for the portfolio of smaller matters that legal teams manage on a daily basis.

For organizations that typically have only 100 GB of data per matter subject for e-discovery, it is prudent to ask, “What are the hidden costs of advanced analytics technologies?” Before acquiring that shiny object, legal teams need to carefully consider the following to determine true TCO: 

• The cost of data collection. Competent, targeted collections can save a huge chunk of the cost of e-discovery. Get this step wrong and the costs and related results can be dramatically impacted throughout the entire the matter, regardless of the software benefit claims. It's important to make sure that the person overseeing the collection can provide a complete a breakdown of what has been collected, why it was collected it and what data can be sensibly and defensibly removed to save time and money.   
• The cost of hiring the technology-savvy specialists. There's no getting around this. With technology acquisitions comes the requirement to ensure that the people using it know what they're doing to optimize results (e.g., creating a seed set that the system will use to learn and find similar documents or concepts). 
• The cost of establishing, training and maintaining litigation support specialists. Expensive, high-risk e-discovery projects with tight deadlines are no time for on-the-job training. 
• The cost of getting it right from a data perspective. If the tools being used are difficult to use and unable to easily provide insights into the data, it becomes that much easier to get it wrong, which can be very risky both from a cost perspective and a legal perspective.
• The cost of being able to competently and defensibly explain the predictive coding processes. The legal team must be able to clearly explain how the advanced technologies used were applied and validated. Without this knowledge, it will be harder to defend the process in court and convince opposing counsel or a judge as to the accuracy and efficacy of those methods.

Reducing data volumes early in the e-discovery process holds the key to lowering the TCO. Advanced, analytics-based technologies that turn more of the decision-making over to computers may sound appealing, but unless the organization is ready to incur the hidden costs of operating that technology over the long run, the “shiny object” may not be the answer. Deploying an intuitive forensic search tool that allows legal teams to analyze, sort and reduce data volumes before collection may be the better option.