Key considerations in using predictive analytics as a tool to help you reach your departmental and organizational goals.
In the September-October issue of CCBJ, my colleague Vince Venturella offered an introduction to predictive analytics and the benefits they can offer to law departments. Picking up on that topic, in this article, I offer further insight on what legal professionals should look for in this type of solution.
Although we sometimes discuss technology innovation as if the goal is to find a way to incorporate them into your processes, like any technology, predictive analytics solutions are not an end in themselves, but a tool to help you reach your departmental and organizational goals. I never expect a client to call me up and ask “Jeff, how can we add predictive analytics to our legal technology approach?”
However, it is very common for clients to say “we need a better way to select the right outside counsel for litigation,” or “we need to improve our budgeting accuracy.” In my experience, these kinds of operational challenges are best addressed by using Artificial Intelligence (AI) to fully, deeply mine the untapped value in the data that law departments collect every day on every matter and every invoice. Predictive analytics tools are a great way to do that.
How to judge the quality of a solution
When you are ready to look for a solution that can put your organization’s data to its best use, you are likely to find several offerings that brand themselves as predictive analytics. However, the term is not standardized and there is significant variation across the market. As a result, there are several traits you should look for to ensure you select a powerful solution that will deliver the value you need.
Predictive versus “predictive” – Looking at historical data can give you an idea of what future results may be, but a couple of key factors are missing. One is the intrusion of the outlier. Historical data in and of itself can’t spike out the red flags and other nuanced metrics that can make a matter fall outside the norm. Second, historical data doesn’t learn and adjust how it might forecast a future matter. Look for a predictive data model that can both understand how outliers affect predictive data and learn to get better at its job after the close of each matter.
A focus on counsel selection – The choice of which law firm to assign to a new matter is one of the most impactful decisions you can make in terms of value. So many factors are at play in determining which firm is best for which matter that attorneys need a huge amount of experience to even understand all of the considerations. And even with that experience, it can be difficult for in-house counsel to be sure that their impressions of past performance really represent the latest information. Solid, up-to-date data can ensure the counsel assignment that is most likely to deliver your desired outcome on budget.
Integration into your workflow – Dashboards are perfect for reporting performance to senior management; in-depth, drillable reports can help you deepen your understanding of the factors that affect that performance over time. However, when decision time comes and you are making an important choice about how to proceed with a matter, you need the right information at your fingertips. Choose a solution that displays the information you need on the screen where you’re working while making decisions.
KPIs that deliver relevant, high-level information – A solution that presents data in the right context on your screen should also deliver digestible, easy-to understand information that is clearly relevant to the choice at hand. Factors such as firm ratings, average cycle times, and performance to budget – for the relevant practice area and jurisdiction – provide insight that supports clear, straightforward comparisons at decision time.
Machine learning – There is no better technology than machine learning for processing huge amounts of data and helping you understand what it means. The wealth of data points on legal invoices can be overwhelming for humans, but it’s exactly what machine learning AI algorithms use to spot trends, compare performance, and constantly update the information they give you. An important point here is using a service with the highest volume, relevance, and quality of data (such as LegalVIEW®, the world’s largest source of legal performance data, including more than $128B worth of legal spend). Some solutions are billed as predictive analytics, but lack this fundamental component of effective prediction.
Long-term process improvement
A solution that meets the above criteria will deliver value on day one and will remain useful as your organization’s data and needs evolve. This is manifested in several types of improvements. To begin with, legal professionals will be more confident in their decisions because they will be armed with everything they need to make informed choices. A machine learning-assisted predictive model can catch details that might otherwise be overlooked. For example, if a party involved in a matter usually takes longer than average with e-discovery, the data presented will reflect that nuance. Even associates who have less experience can use this type of tool to draw on years of historic data. Meanwhile, more senior team members will be able to confirm that their decisions are objectively sound.
In addition, attorneys won’t need to choose between efficiency and sound decisions when they use a tool that puts the right information precisely where they need it. Because there is no need to break up their workflow for long research sessions, there is also no temptation to use guesswork or fall back on “what we have always done” to save time. Decisions are made quickly and are backed by data that reliably distills complex records into easily understood facts.
With a top-tier, AI-driven predictive analytics solution in place, law departments can expect to see a host of improvements as soon as associates have access to the information they provide. By turning facts and figures into useful information that your expert team members can act on immediately, predictive analytics offer a fast and reliable way to increase value and improve operations.