Blog Viewer

How to Present Data to Lawyers

By Nathan Cemenska posted 07-09-2020 09:32


When Presenting Data to Lawyers, Keep Things at a 10th Grade Level
They say the New York Times, the most respected news source of all time, is written at an 10th grade level.  That’s the way legal analytics should be, because you don’t have to be fancy to be highly influential.

In fact, being fancy hurts you:  On average, a “no-frills” approach will be more influential.  The more you try to be fancy and show how “advanced” you are with your cool-looking visuals, the less effective they will be.

How Simple? Very Simple
I didn’t totally understand simplicity when I started out in legal analytics.  Sure, I knew to keep things simple, but I didn’t get how simple:  Really simple.

I would create visuals like the combined bar and line chart below, on the left.  I thought they were simple and anyone could grasp them, but it turned out that some people found them confusing.  This small but vocal minority could ruin an entire meeting by getting the conversation off track.  Better to avoid such confusion by using the simplest, most straightforward visuals possible—even if you have to use more than one to do it.  See the example on the right, which is slightly simpler because it only requires the viewer interpret one kind of chart (the bar chart).  The same old, same old chart types may be boring, but they are familiar and understood effortlessly.  You can’t beat that.

Another misadventure was my brief but stormy love affair with violin plots.  I won’t bore you by explaining what violin plots are, but read this article if you care.  Violin plots are cool because they cram a ton of info into a very elegant visual.  Unfortunately, they confuse the heck out of people.  Luckily, there is a simple way to avoid the confusion:  Be like the New York Times.  Avoid violin plots, and keep things at a 10th-grade level.


The Simplicity Police
If you follow the advice above and try to keep thing simple, you may encounter others who seem to oppose that agenda.  Certain lawyers or other people you work with will push you to make things more complicated than they need to be.  When this happens, somebody (you) needs to speak up and educate people about good design, lest you end up with the dashboard version of a Jacuzzi tub:  Something that everybody wants, but nobody uses.

Simple Doesn’t Mean Simplistic
I think it goes without saying that just because the way data is presented is simple, that doesn’t mean the message or what went into creating the visuals is.  The two concepts are entirely different.  My company operates LegalVIEW Data Warehouse, the largest body of legal performance data in the world.  In forms the basis of LegalVIEW Predictive Insights, a first-of-its-kind product that uses artificial intelligence and machine learning to predict the cost, cycle time and payout of litigation matters.  The “guts” of how those predictions are produced is incredibly involved and painful to describe, but the final output is very simple:  A list ranking what law firms our AI predicts are the best ones to assign to a new matter based on weighted criteria (cost, cycle time, outcomes) provided by the client.  The list isn’t much to look at, frankly, but I don’t think it matters.  What matters is it is easy for attorneys to understand and it can help you save millions of dollars.

Obviously, the lawyers you work with are far from dumb.  They could easily understand complicated visuals if they wanted to put in the effort.  However, they are preoccupied with all the emails they have to return and the fires they have to put out, not your chart.  If you waste their energy by presenting data in a way that creates unnecessary work for them, the message won’t get through and some people will stop listening altogether, which is easier than ever in a post-COVID world where more and more of us work remotely.  If you are not careful, instead of influencing them you will train them to avoid analytics altogether.  So instead of trying to get them to speak your language, speak theirs by getting them the information they need in the simplest, cleanest, most familiar and easy-to-understand format possible.

Good visualizations, like a haiku, are beautiful in themselves, without any need for ornamentation.  While it is a good idea to apply standard formatting conventions like data labels, borders around the edges of bars in bar charts, etc., I wouldn’t go far beyond that.  And unless you have a special reason to go beyond basic, tried-and-true visualizations that everybody understands (think line charts, bar charts, maybe a scattergram), I wouldn’t do it.  Keep it simple.

Easy to say, hard to do.