Stionson Leonard Street

Big Data will significantly transform legal services and one of the key tools for that transformation will be the Document Management System (DMS). Like almost no other industry, legal service is shrouded in mystery. Years of training and apprenticeship are required for individuals to master practice skills. To protect hard-won intellectual assets, large law firms almost universally adopt the "black box" approach to their daily routines. Clients do not need to know the inner workings of a firm; what matters is the result. A laudable approach if it means protecting ethical obligations. Too often, however, it also hid a self-protective attitude toward technological change, which, as in so many other industries, threatens the revenue generated by traditional manual activity. The sacred cow of hourly billing is on the chopping block.

There is a better way to approach the challenge. If the goal is to serve clients' legal service needs, then a thorough ex­amination of inner routines is required. One of the great resources attorneys have at their fingertips is mounds of electronic files stored in document management systems. A typical large law firm may hold multiple terabytes of data representing tens if not hun­dreds of millions of documents, the residual artifacts of legal practice. The DMS is a powerful resource ripe for data analytics.

A DMS is filled with work product accrued over the course of years. "Work Product" is a term of art that means the legal opinions and supporting documents, contracts, agreements, corporate governance records, financial information, forms, procedural records, briefs, pleadings, filings, etc., that comprise the daily activity of lawyers. If this data is stored as structured information in a DMS, then it can be harvested and analyzed.

A signif­icant amount of legal activ­ity is routine—and as any technologist knows, anything that is routine can be turned into an algorithm. The challenge lies, really, in understanding what ques­tions to ask the data. It seems unlikely that trolling an entire collection of dis­parately acquired documents would be very productive, especially if you don’t have clear objectives in mind. The dirty little secret of the law firm DMS is that much of the content is junk, informa­tion gathered in pursuit of an approach that never materializes, chitchat among staff, or the meeting minutes of defunct corporations. The smoking guns are rare, while transactional documents are generally fairly well defined by market or regulatory requirements. In other words, smoking guns are outli­ers that can be treated as outliers, while routine document scan be isolated, and controlled statistical experiments run on them at relatively low cost.