Law firms try self-analysis, “Moneyball” style

Internal data detect future problems

Financial Times, December 2017


When law firm DLA Piper began grappling with a sudden loss in client work from some of its usual base around four years ago, the marketing team started an initiative with a consultant to figure out what was going wrong.

DLA’s chief marketing officer, Barbara Taylor, theorised that the firm could take a “Moneyball” approach of analysing data, following the lead of Michael Lewis’s 2003 book of the same name about the art of winning in baseball by using statistics.

The consultancy hired by DLA, Axiom Consulting Partners, built a predictive learning model, called a support vector machine. It could identify which clients were likely to turn to other law firms for advice and how to stem the tide of departures. “It predicts what we call ‘at-risk’ clients, across all clients, and it got that right about 80 per cent of the time,” says Ms Taylor. “We’re obviously really excited and inspired.”

DLA identified 22 predictive markers that would indicate clients were likely to stray. It found that a number of changes could make a big difference: reducing the overall size of its legal team on a matter, while increasing the amount of time spent on a client per team member, introducing one new professional to the relationship, and adding an industry expert to the team.

If the firm took none of those actions, there was an average 48 per cent revenue loss. If it did one, revenue increased 14 per cent, and if two or more were accomplished, monthly fees rose 20 per cent. She calculates that the programme has added nearly $38m in new revenue per year.

The firm is one of a growing number of those that, after years of advising clients on how to capitalise on the reams of data they had in their possession, are now looking inwards to do the same themselves.

How firms have sought to take advantage of their data has varied. Kirkland & Ellis was one of the earliest among law firms in this move. Its CTRAN database, which all its lawyers can access, includes figures from deals where the firm represented a party in the transaction, including information on financing terms and ways that buyers allocate the risk of financing failure.

Now the firm has a real-time dashboard featuring its own analytics on every Kirkland attorney’s desk, says Norbert Knapke, the partner who has spearheaded the project.

The third iteration of this database, CTRAN 3.0, is in the works and will incorporate more data from its international offices that include details gleaned from deals where the target is a private company and information is not necessarily publicly available.

It has also allowed the firm’s attorneys to spot trends such as increased use of representation and warranty insurance, says Mr Knapke. Clients, he adds, “want to make sure they’re competitive but they’re not going too far — they need reassurance that what they are doing is the right thing”. The firm’s evolving proprietary database helps ensure clients and advisers more fully understand “what’s going on in the market”, he argues.

Littler Mendelson, by comparison, is focusing on using data analytics tools to help its clients to detect employment risks.

The employment law specialist briefed its global director of data analytics, Zev Eigen, to help clients review operational concerns. These ranged from ensuring disadvantaged minorities were properly welcomed into the workplace, calculating and examining the justifications for pay discrepancies between men and women, and identifying other dangers of employment lawsuits.

“Data are messy, data are siloed, data are not designed for innovation when they were originally collected or gathered,” says Mr Eigen. “There is voluminous work that has to happen and often a big investment of time and resources before true benefits are gleaned.”

Mr Eigen’s own company Syndio has worked on issues outside the legal space using data. In one instance, it measured how teams in neonatal intensive care units worked together in order to predict which employees should be grouped together or kept apart. That analysis can be similarly applied to Littler’s clients, he says.

In work for a Littler client, Mr Eigen helped a tech company analyse whe­ther diverse staff were being integrated into the divisions in which they had been hired.

To their surprise, they found there was no correlation bet­ween the breakdown in how many diverse staff there were versus non-diverse and whether they were being integrated. “It’s called ‘diversity and inclusion’ — if you’re just focused on headcount, you’re not focused enough on inclusion,” he says.

Companies “are nervous about the legal risk of running these kinds of analytics”, he says. “We can usually assuage it by it running through a law firm.”

Originally published in the Financial Times, December 2017

Data Analytics
Data Science
Law Firms
Professional Services
Revenue Growth
Related Insights