Spend any time talking to people who work in the field of “people analytics” about what they do, and before long someone is bound to reference Michael Lewis’ book “Moneyball.” The book chronicles the story of a baseball team that managed to sign division-winning baseball players on a shoestring budget by leveraging insights gleaned from data analysis.
People analytics is an emerging industry populated by companies that think they can do something similar for traditional businesses—harness the power of big data in order to make more efficient hiring decisions. Thus far, people analytics has mostly been embraced by companies that hire in bulk to staff big operations such as call centers. But some analytics companies think that their techniques could also be used to improve the way law firms do their hiring.
The general principle behind people analytics is straightforward enough. Analytics firms study the existing employees doing a particular type of work for a company and look for attributes that correlate with the most successful workers—and also with the least successful. Firms then analyze job applicants to assess those attributes and predict whether the applicant would fare well, or not, in that position.
Law firms already attempt to do this, of course, by looking at applicants’ grades and the quality of their law schools when hiring new lawyers fresh out of law school. The problem, people analytics experts argue, is that these factors are actually only weakly correlated with success when compared against other factors that are typically given less attention.
“Those are proxy measures for something that’s underlying that,” said Greta Roberts, CEO of Talent Analytics. “What we do, and what they did in ‘Moneyball,’ is to look not at the proxy measure, but to figure out what are the attributes underlying that.”
There is no magic number
The legal industry, as a whole, is not known for embracing change at a rapid pace. (Nor are baseball scouts, a tension that drove much of the conflict in “Moneyball.”) Evan Parker, the director of analytics for Lawyer Metrics, said that right now the law firms taking an interest in people analytics are largely what he called “early adopters”—ones that are unusually willing to try new ideas. But Parker said he expects that people analytics will eventually move beyond early adopters and into wider acceptance.
Lawyer Metrics offers to perform for law firms the sort of correlational analysis described above and then boil job candidates’ qualifications down to a single composite score that estimates their likelihood of success with the firm. Parker conceded that some lawyers are leery of reducing a candidate’s prospects to a single number, but said the numbers were less of a crystal ball and more of a way to hedge against too heavy a reliance on academic pedigree.
“Once you start assigning numbers to people, there’s a tendency to think of that as absolute truth. But the question is whether that aggregate number is better than you would have done otherwise,” Parker said. “What this is really intended to do is help you screen out the tails. It’s going to identify people that are going to be high performers, and it’s also going to single out people that you should really have questions about.”
Parker said that things that job candidates had done in the past turned out to be the best predictors of future performance. Whether a candidate was, for instance, a leader in a high-profile organization in college often ended up being much more predictive than whether they attended an especially prestigious law school.
Hey, you look familiar
Efforts to make hiring decisions with mathematical precision long predate the age of big data. In the period following World War II, administering aptitude and psychological tests to job applicants was very much in vogue. Those early efforts at data-driven hiring decisions eventually fell out of favor, however, because the tests often turned out to be poorly correlated with job success, and concerns were raised about whether the tests were implicitly biased against certain groups.
But Roberts argued that most companies’ hiring processes, which typically rely on screening interviews, can introduce significant amounts of bias to the hiring process in their own ways. Employers will often home in on candidates who are similar to existing employees, but in regard to factors that are unrelated to how well they can perform the job.
“The whole process today, absent analytics, is an extremely biased process. A lot of times, if you look at the hiring manager, you can clearly see that they’re hiring people exactly like themselves,” Roberts said. “Legally speaking, an analytics approach helps you scrub the bias out of the process. It will point out people that you hadn’t thought before would be a good candidate.”
Deducing the factors that correlate with success obviously raises the knotty question of how exactly to define success. In industries that frequently make high-volume hiring decisions, people analytics is often used to predict how long candidates might stay in the position, in order to help reduce the expensive cost of constant re-hiring. Measuring the value generated by high-knowledge workers, like attorneys, is more complicated in comparison, Roberts said.
Who’ll make it rain?
Additionally, the smaller the pool of existing employees there is to analyze, the more difficult it is to identify attributes that will predict success. Each law firm’s needs will differ slightly, so smaller firms would potentially present more of a challenge for an analytical approach than larger ones.
Parker said that Lawyer Metrics typically measures success by analyzing performance evaluations, but another way to sift for exceptional candidates is to look at correlations across the broader industry. To that end, his company studied a tier of attorneys that it deemed “rainmakers”—partners who routinely generate new business for their law firms—and contrasted their traits with those of non-rainmakers.
The study, Parker said, suggests that potential rainmakers differ from other attorneys in ways that can be recognized even before they’ve built a track record of developing clients. Some attributes, like an eagerness to take on leadership roles, are hardly unexpected. But some of the other predictive traits were more surprising.
“One example was rainmakers tend to score more highly on a propensity to take risks. That surprised us at first because attorneys as an industry are known to be risk-averse,” Parker said. “But that was one of the key factors separating them, and from interviews we learned that this risk-taking propensity wasn’t about doing wild and crazy things, but they push themselves out there in a way that most of us are not inclined to do.”
“Risk-taker” is not a label that job candidates often affix to themselves in cover letters, however. To deduce which candidates are made of such stuff, Parker advocated that law firms establish structured interview processes with the goal of eliciting answers from candidates that will signal how likely they are to succeed at the firm.
“What people analytics allows you do to do is make an inference about whether this person is going to pay off in the end,” Parker said. “You don’t want to bring in someone and pay them a large salary for several years without them ever making a long-term contribution to the firm.”
Follow David Donovan on Twitter @NCLWDonovan