Law departments of the world, unite. You have nothing to lose and everything to gain with data analytics!

Your business is awash in numbers, along with software tools you can use to learn from them. You also have an important cadre of allies: Your key law firms have for years been stockpiling operational numbers about your work that will help you increase and demonstrate your department’s value to your company. But those firms need prodding to encourage them to mine their data for your benefit. That’s why you need to stay informed and involved.

The five imperatives below are the kinds of things you may want to say to your key firms. If you do, this manifesto gives you an easy way to deliver the message.

1. Collect and curate data that helps our law department manage our legal services.

Your firm should catalogue what numbers it currently tracks or could reasonably track to improve our management decisions (and your firm’s, to be sure). Examples of such data include the cost and duration of our matters (and similar matters of your other clients), the number and level of timekeepers on our matters, and the types of matters you handle (with descriptions meaningful to my business operations and legal risks). You should also consider tracking and compiling data on who in our law department or among our clients asks you to work on matters.

Let us know what information you are collecting, your methodologies and the definitions you are using for the data because we may want to suggest improvements or priorities. Also, we may have some data we can share with you to supplement your own.

In addition to storing your raw data in spreadsheets or databases, you should curate it. Look for oddities in the data (possible typos or outliers). Make sure that you use standard forms of numbers and words (1,000, not “1K”; always “Real Estate” rather than sometimes mixing in “RE”). You should also try to fill in missing data, if you can, by adding informed estimates or by letting software do what is called “imputation.” We will be impressed if you supplement the data you collect with external data. In short, we want you with consistency and thoughtfulness to pull together numbers that will help you help us do our jobs better.


2. Learn from the management-oriented data you accumulate.

As a starting point, statistical tools abound that can describe what you collect. On their own, such statistics can benefit both of us in understanding workloads and resources needed. For example, if we learn the average amount of time it takes to obtain a certain license in a certain state, we might change how we proceed or what we tell our clients. If we know the variance between the number of hours different levels of paralegals and lawyers record on matters, we might decide to adjust core staff. All kinds of descriptive statistics could prove useful.

Your firm either has or can obtain software that can do the analyses. Even Excel is quite capable. But other programs, such as Tableau, SPSS and the open-source powerhouses of R and Python, can go far beyond it. For your level of needs and talent, these are straightforward tools, not hard-to-learn programming languages.

We urge you to apply more sophisticated statistics, especially regression. Multiple regression can predict such crucial estimates as total cost and duration of matters, and it can also pick out and prioritize drivers of cost and delay.

Machine learning algorithms are the next step. We would be excited to try them out. With sufficient data, algorithms such as neural nets and Naive Bayes can weave the straw of data into the gold of insights. Text mining algorithms also beckon to extract meaning from our documents and text repositories.

By all means include us in your journey of learning; we want to keep abreast of legal data analytics to help us in the law department serve our company better. For our part, we will see if corporate IT staff can team with you. These are early days in the analysis of legal management data, and everyone involved has much to learn about the strengths of different methodologies as well as the constraints on them.


3. Explain your insights with graphs.

We want to be able to read the stories in the data you compile. When graphs effectively visualize data, someone who reviews them should be able to understand your conclusions and draw their own. Graphs are far better than lengthy text, dense tables or long lists at conveying insights, and your firm would do well to master the techniques of “data viz.” Alternatively, if graphs are unable to make a clear point, it may be that more or different data needs to be collected or that refined analytic or visualization techniques need to be employed. Note, however, that we don’t need fancy infographics or the glitzy output of Photoshop or GIMP.

Start simply when you put our data in graphs. Scatter plots are particularly effective. All the data can be present on the plot and the data points can be colored, shaped or sized to pass on even more information. Imagine if you showed the number of claims in employment discrimination charges on one axis, the cost of resolution on the other axis and colored the points by the state where the charges were filed. If the scatter plot boasts a trend line, even better. Box plots are also a steady standby when there are factor variables such as matter type or level of timekeeper. So-called panel or facet plots can clarify complicated data by breaking subsets into multiple views.

Compelling graphics go a long way to inform, but explanations of how you arrived at them make them even more valuable. In general, we strongly urge you to be as transparent as possible about how you make use of the data generated by the work we have given you. Reproducible research is the goal.


4. Propose actions based on your analysis and visualization of data.

Unless we both give thought to what we might do differently, based on the data, your efforts won’t realize as much as they could. Let the light of analytics illuminate dark areas of our operations. We encourage you to propose actions and changes based on what you see. To that end, as we progress we will develop a common understanding about data that has been probed and portrayed effectively. Our joint learning will create a shared vocabulary for considering improvements.

For example, when your numbers tell their story, we might alter the staffing of matters; we might train our in-house lawyers to instruct you sooner or more effectively; we might explore different billing arrangements based on the data; we might change how we train our clients or draft our agreements; and we might send more work your way.


5. Educate and encourage your lawyers to embrace metrics.

As you gain analytic skills, you will be able to differentiate your firm in the marketplace. But more important, you will forge stronger ties with us and our executives. As you continue to incorporate metrics into your decision-making processes and day-to-day practices, this management tool will gain leverage. Managers who can astutely draw on metrics become better (and more profitable) managers. To nudge the mindsets of partners, associates and paralegals so that they recognize and value quantitative resources combined with the powers of statistical software is to move your firm toward the future.

We appreciate that numbers can go only so far in attempting to tackle the conceptually complex arena of legal counsel and drafting, but analyzed practice data has been an underexploited resource for realigning operational management and efficiency. Many aspects of the craft of lawyering can improve as more lawyers quantify attributes and integrate the learning that comes from doing so into the traditional practice of law.

Our internal business clients increasingly embrace and rely on analytics. We in the law department also want to exploit good practices grounded in data. To that end, we are considering how to incorporate into our outside counsel guidelines more analytics and, indeed, how to make data analytic skills part of how we evaluate your performance and the value you deliver. Better data is a fulcrum for better management.

Our department wants to ride the big waves of data analysis and visualization. We strongly encourage you to dive in and play an active and creative role in their development.