Using Legal AI to Eliminate Cost Overrun Risk

The world of AI-driven legal billing analysis is growing fast, with several companies exploring how they can use natural language processing (NLP) and machine learning to provide insights into the heart of any legal business: how fees are calculated.

One of the companies staring into the inky darkness of this legal black box is Holland-based Clocktimizer. Artificial Lawyer talked with founder and former DLA Piper lawyer, Pieter van der Hoeven, about what his company was now doing to make this murky world a little more transparent and predictable.

The first thing that CEO, van der Hoeven, says is that Clocktimizer is at present mainly focused on law firms, rather than corporates, with about 90% of clients in the former group.

Yet, one might think that any AI system that helps to bring additional clarity to the arcane discipline of creating legal bills could potentially work against a law firm’s ability to increase revenues and conversely help a client to cut its bills. The unpalatable reality is that law firms benefit from a lack of transparency, as do all businesses with what one can call ‘information asymmetry’ in relation to their clients. 

So, first of all, how come van der Hoeven chose to focus initially on shedding light on law firm billing practices….for law firms rather than the clients? Isn’t that counter-intuitive? Why would a law firm want to invest in technology that could cut its own bills? That’s illogical, right? No, says, van der Hoeven.

‘When I was an associate I used to get stomach aches from having to give clients estimates for what a piece of work would cost. Would it be the right amount?’ Van der Hoeven explains.

He had felt the pain of having to guess what a piece of work should cost based on insufficient data to make a truly objective estimate. This mattered because, as he says, clients might then right off many hours of work, which the firm would have effectively paid the price for in terms of production costs. In short, a double loss: wasted hours and hours that could have been paid for, rather than written off, if they’d been utilised more effectively.

In a time when clients didn’t push back on legal bills this issue may have been unimportant, but as the founder explains, since the financial crisis of 2008/9 client attitudes have changed. There is now a real risk of upsetting valuable clients by appearing to over-bill. In short, helping a law firm to suggest a more accurate estimate for a piece of work helps to remove right offs, which in turn reduces wasted time and cost inside the firm.

To Artificial Lawyer this makes a lot of sense. Time spent on wasted, unbillable work is time associates could have spent on remunerative matters, or simply more remunerative matters in general. I.e. making sure your bills are the right level before you start work could increase the overall revenue and profit of a firm.

It may sound illogical to cut a proposed bill today to make more money longer-term, but in the world of law firm economics that is precisely what the outcome may be.

In fact, one of Clocktimizer’s benefits is that by using its software to read through previous examples of similar matters it can help a firm to better predict the profit margins that are possible.

It can do this because it looks at all the other pieces of work that have the same outputs, e.g. contract creation, time spent on negotiations and due diligence. It can then work backward and say to itself: ‘What else has this firm done that looks like it had the same type of work product outcome?’ I.e. one deal can be compared with another not by practice group, but on the actual output the client receives.

In short, the question is: what inputs does the law firm require to create these outputs for the client?And, in doing so, Clocktimizer becomes a system that helps law firms to eliminate the risk of cost overrun. That’s it, in a nutshell.

‘This has a forward planning capability,’ says van der Hoeven. ‘And this is done by analysing historical data.’

‘Clocktimizer helps you to build a budget and can use ‘reference matters’ that are similar so the firm can compare them [in terms of cost, fee earners used and profit margin,]’ he adds.

Overall this seems like a very positive step forward and certainly one that General Counsel will likely welcome. They are busy people and having to push back on bills is a waste of their time and energy.

After all, dealing with the myriad legal risks is a full time job for inhouse lawyers. Who wants to have to fight with external advisers because they keep handing what look like inflated bills, or bills where it cannot easily be explained how the firm came to the decision to bill that amount?

And, one can conclude, if a law firm keeps annoying a client in this way there is one very simple and powerful thing a client can do: stop using them.

The loss of a client due to inefficient billing is a silly way to lose a client, especially given the ‘cost of sales’ required to win a new major client in the first place. In which case, use of AI billing software could be seen as a client retention tool as much as a financial analysis tool.

It seems likely that this AI tool and others (e.g. Legal Decoder and Bright Flag), and the IBM Watson OCI system (which was covered last week), are going to make an increasing impact on the legal market both when used inside law firms and among inhouse legal teams.