Ask Dr. Paola: Choosing the Best Path for Big Data — Using Descriptive and Diagnostic Analytics

Topics: Data Analytics, Harvard Law School, Law Firm Profitability, Law Firms, Legal Innovation, Midsize Law Firms Blog Posts


The Legal Executive Institute blog is honored to be working with Dr. Paola Cecchi-Dimeglio, a behavioral economist and senior research fellow for Harvard Law School’s Center on the Legal Profession and the Harvard Kennedy School, on a monthly column.

Each month, Dr. Cecchi-Dimeglio will be answering questions about how law firms and legal service firms can navigate a dramatically changing legal environment by using data analytics and behavioral science to create incentives for law firms and their lawyers to change behavior. (You can follow her on twitter at @HLSPaola.)

On to this month’s question…

Ask Dr. Paola: Looking at how my firm uses data analytics, it seems we’re able to pull a lot of data around our firm’s past performance. So, we can get a pretty good descriptive idea of what we’ve done, but what else can we determine from that data?

Dr. Cecchi-Dimeglio: I think what you are talking about is the difference between descriptive and diagnostic analytics, which have been central to the performance of management and reporting systems of organizations for many years. In fact, there is a habit of organizations to use those kinds of analytics almost naturally, compared to predictive and prescriptive analytics, which we discussed last time.

So, descriptive analytics tells you what happened, and this is the kind of information that you’re going to get. It’s an insight, and it’s easy to determine. Diagnostic analytics will tell you why it happened.

Just as we described before where predictive analytics will tell what could happen; and prescriptive analytics describes how you can best try to make a certain outcome happen — it’s important to keep those four questions in mind and how they relate to each other. What happened is descriptive, why did it happen is diagnostic, what will happen is predictive, and how can we make it happen is prescriptive.

And as I said, descriptive and diagnostic analytics are used over and over by many organizations. However, it is very important to know that both descriptive and diagnostic — and I want to make this very clear — is not about causation. What you find in descriptive and diagnostic analytics is about correlation, and correlation does not imply causation — that has to be clear for people in regard to this data.

The important factor in the use of descriptive analytics, of course, is knowing how descriptive your statistics are. 

The descriptive data is really showing you what is happening within your organization. It’s there to help you to uncover different patterns and get a better picture of where you are as an organization — and you are basing that knowledge on incoming data. What generally happens in organizations, both in professional service firms and other legal entities, is you get either real-time dashboard data or a report, depending on where you are within the organization. So, you’ll use descriptive analysis to describe and summarize the raw data, making all the data tangible so a human being can understand it.

The important factor in the use of descriptive analytics, of course, is knowing how descriptive your statistics are. Do you have statistics that are easily available to you on a minute-by-minute basis with which you can run a report? Or are they available on a monthly basis? Or on a quarterly, annual or five-year basis? It obviously makes a big difference.

If you are trying to look at a pattern and you have analyzed five years of data, that’s great, but you also need to be able to look at the data on a year-by-year or even month-by-month basis. Otherwise, you may find certain things when all the data is analyzed together, but if you’d look on an annual basis, those patterns or trends do not hold true. You need to have that kind of precision when you look at data.

Dr. Paola Cecchi-Dimeglio

As for the diagnostic part, you still look at past performance, but you’re really trying to understand the question of why a particular situation occurred. What happened there?

Mostly, diagnostic analysis has been used for finance, for example, to uncover the cash flow a firm has, or speeding up the invoicing process, or to determine the overall profitability of the firm. That diagnostic analysis is a very important part of an organization and of a firm, because that knowledge will help them to decide where to invest next year in technology or new practices or new partners. In fact, diagnostic analytics can really help a firm or organization to determine the flow of the workforce they may need going forward and where best to make those investments.

Overall, firms need to understand the return on investment in using diagnostic and descriptive analytics — what is the firm getting out of it? What changes or investment were they able to make using this data and what dividends did that pay?

The biggest victory for any firm is their ability to show a measurable return on investment by the use of this data. And that is translated into increasing the number of new clients that the firm has, getting more business from their existing clients, and producing work more efficiently and at a lower cost to benefit their clients and the firm itself.