By Richard Lau, Manager, Production

As the pandemic progresses, commentators are still trying to figure out what the new normal will look like for the global economy. However, the uncertainty created by the global COVID-19 pandemic has only intensified a decade long trend in the legal industry of more clients becoming more critical of conventional wisdom and more skeptical of the value being provided by legal services.  We have already seen this dynamic appear in staffing: clients are no longer willing to pay the same rate per billable hour for work performed by associates as they do with work performed by partners. But even among attorneys of similar seniority and experience, each hour worked is not necessarily as productive as any other. How can we valuate the time attorneys spend on given tasks? At the same time, law firms must select which clients to represent and which cases to take. The intensifying pressure to generate value means that firms must also become more discerning about which cases they take. Taking a financial loser of a case can be  devastating.

The emerging  answer appears to be legal analytics. For the past 10 years, the legal industry has been moving to a more data-driven approach to decision making to reduce waste and provide greater value, and now in the midst of the COVID pandemic, this increased efficiency will not be just another selling point for a given law firm, but a matter of that firm’s survival.

The Old Paradigm

Discussion of big concepts like analytics in the abstract can be disorienting, and lead to those important ideas being perceived as buzzwords. The easiest way to see the value of the legal analytics is to examine the systems and procedures they are replacing and their weaknesses.

In the past, staffing and trial strategy decisions were made primarily based on an experience. The supervising attorney in charge of a case would rely on their informed intuition to decide which members of his team to assign to a given task and how much overall time to spend on a given task.  In this way, a competent supervising attorney will usually create a strong and effective overall strategy but is not very useful when trying to justify your prices on an itemized bill. Moreover, this approach will result in unnecessary and potentially wasteful tasks being codified along with the valuable tasks. This is simply human nature: when people experience success they want to do things the exact same way in the future, and without a clear picture of how each individual step contributes to the success of the overall endeavor, why would they compromise a proven strategy by changing it? Personnel decisions were made in a similar way, people being given tasks based on their perceived or general expertise. Over the past few decades, the industry has been moving to a more objective, quantitative means of measuring value, through standardizing task and pricing data in (UTMS and LEDES respectively) so that we can more effectively make comparative evaluations. But these categories must be manually assigned, often by attorneys unfamiliar with the system or uninvested in tasks that do not directly pertain to winning the current case.  

On the case valuation front, firms typically task an attorney to read the complaint and looking for three to five previous cases with similar facts, or at least the same general topic. The attorney will then look at the billing data and make an estimate regarding the revenue and cost of handling the current case based on those previous cases. The process resembles searching for precedential decisions in litigation: the attorney reads the complaint and makes a judgment call about whether its “on-point” enough. With three to five complaints, there is simply not enough data to perform a more granular comparison of specific facts (e.g. evaluating the impact of the plaintiff’s age in an age discrimination case, in isolation of the other case facts). However, reading more than three to five similar complaints per complaint is not cost effective: the information gained is not worth the time spent. This is especially true when using these comparisons to determine which cases to take, since these assessments would need to be performed on cases the firm turns down as well.

The New LegalMation Paradigm

The biggest obstacle to any analytics solution is often obtaining the data. Law firms and inside counsel often have a resistance to analytics because they believe that any solution would necessarily involve the tedious and time-consuming task of labeling all their unstructured data. However, large amounts of unstructured data are no longer the obstacle it once was.  Advances in natural language processing allow AI platforms like ours at LegalMation to analyze work product to automatically determine the type of work performed and which litigation phase in which it belongs. Firms no longer must rely on the subjective and sometimes incomplete manual identification of these phases, consistency when comparing the work of multiple attorneys. Instead of having an attorney read complaints and look for similar complaints, we have an AI platform that analyzes the complaint and identifies the actual facts of the case (such as the specific medical condition in a discrimination case). Such an approach has the advantage of being both more granular and being capable of considering far more than just three to five complaints at once.

From there it becomes a relatively simply matter of incorporating the billing information from the law firm’s existing enterprise legal management software (which already stores this billing data in a semi-structured format).

Here are a few examples of the actionable insights our platform can derive:

Conventional Wisdom	What the Data Actually Shows:
Discovery does not directly require an attorney’s expertise (applying law to fact) and thus should be assigned to associates and paralegals.	Senior attorney involvement in the discovery phase tended to have better outcomes, suggesting that their involvement and legal expertise make an impact here.
In an age discrimination case, the older the plaintiff, the better the cases odds of success.	Age discrimination cases where the plaintiff is between 50-55 actually have the greatest odds of success.
Opposing counsel and jurisdiction make a difference	This is true, but now we know specifically how and how much (e.g. being able to differentiate between a jurisdiction that is slightly more favorable and a jurisdiction that is much more favorable).
An employment attorney with the best success rate is the best person to handle any employment case.	An employment attorney may excel at specific subjects (e.g. age discrimination vs racial discrimination) or specific tasks (e.g. talented at discovery vs talented at motion practice) and is thus more valuable on some employment cases than others.

Our platform derives these insights using a specific firm’s data, so our client knows that these insights pertain specifically to them.

                Moreover, our system can look at all your data at once and generate customizable visualizations. In the time-intensive old paradigm, analysis typically required a specific hypothesis to test. However, this new unified approach allows clients to conduct open-ended exploratory analysis looking at multiple variables at once. This could potentially reveal relationships that it would not even have occurred to us otherwise.

Conclusion

LegalMation’s data analytics solution replaces the slow, discrete, intuition-driven tasks described in the previous section with a single unified platform that is fast, holistic, and data-driven. Where other legal analytics solutions focus on external, public data to provide mono-variate insights regarding case facts and law, the Legalmation Data Analytics leverages a firm’s internal data as well, to create a more complete picture of how internal decisions about procedures and staffing impact outcomes. This platform reveals how both datasets interact: How the value added by legal spend in various litigation phases would vary according the facts of the case.