Predictive data analysis for litigation

Predictive analysis in legal departments means making data-driven decisions that empower you to forecast and enable your team to get the best possible outcomes. In litigation, we find that many lawyers and businesses base decisions on law firm marketing telling them they're the best firm to do the work in that area of law. We need in-house legal to shift to basing decisions on cold, hard facts rather than marketing-induced emotions. 

Let’s talk about litigation and data. At the most basic level, we would tell our clients to start tracking things such as the type of work they've been sending to law firms, how long it took to get there, how long it took to get resolved (six months or two years)? How much did it cost with that law firm and was it settled? How much did the settlement cost? Did they win? Did they lose? Do they usually lose or win with certain judges? Do they usually lose or win with certain lawyers? 

Once you have harnessed and begun analyzing this data, you will be able to see patterns of how matters will play out with particular firms, judges, and types of disputes. Let’s look at some potential scenarios where this could be valuable: 

Your firm has an employment dispute and potential liability of $1 million. Data shows that if you go to court with Judge X, she always goes with the plaintiff. Therefore, the best option is to offer them $50,000 right now and if it does go to court you’ve made every reasonable offer to that client. You're basing your decisions on solid data, not just playing things out and seeing how it goes at the time.

Perhaps a consumer approaches your company with a dispute. You can see that on average when you send similar disputes to X law firm things tend to take X amount of time and cost X in legal fees, usually resulting in a settlement. These insights allow you to decide to settle right now as it will cost less than a six-month dispute process and subsequent legal fees, dealing with the drama, and paying the settlement in the end. 

Maybe an employee has decided to log a complaint with the Fair Work Ombudsman. Your data from previous complaints shows what types of issues the Ombudsman usually makes you pay the consumer, allowing you to decide if you should settle now or let things play out. 

In all of these scenarios, you end up with a happier consumer and a happier business. Predictive data analysis is a win-win. 

Most businesses don’t capture potential liability or potential litigation data, yet these insights can save the business significant costs. You may have heard of the ‘incarceration of legal data’.  The reality is that if businesses kept a close eye on the costs of legal fees in conjunction with settlement payouts, they would see more often than not that proceeding with litigation is losing the business money. 

Everything in-house legal teams do is trying to save the business money. Therefore, it often doesn’t matter if you settle when you potentially aren’t liable. The bottom line is that companies don’t want to spend money on unnecessary external legal fees, they want the most economic outcome possible.Predictive data analytics informs this. 

How many months or years of data should you collate before you can start making informed data-driven decisions? This is dependent on what you are using the data for. If you are looking to predict complex litigation, you will need years of data from similar matters. If you are dealing with low-level consumer disputes or ‘mum-and-dad’ complaints, having a few months of data will put you in a position to begin to make decisions. Generally speaking, we’re talking about using data for BAU issues, consumer disputes and employment disputes. 

There will be outliers of course, humans are complex, and judges and attorneys will sometimes make decisions we don’t expect. Yet there is much to gain from predictive analysis, it is in many senses the untouched pot of gold in every in-house team. It’s a common misconception that legal is too complex to analyze or predict. The reality is quite different. It produces data - and forecasting is simply data science.