How Brad Pitt Can Help Improve Your Hiring Decisions
Dec 2, 2013
By now, most people have likely heard of Moneyball, whether because it is a management approach that revolutionized professional baseball, or because it was made into a movie starring Brad Pitt. In fact, Moneyball was an innovative approach to building a baseball team with a relatively small budget. It was first, and most notably, implemented by the Oakland Athletics in 2002. The A’s, with roughly $40 million, had the third-smallest budget of any club that year and were competing against teams who had as much as three times the money to spend on their rosters.
The system’s success has been widely documented, with many teams incorporating it into their personnel strategies in the decade since. Moneyball, however, is just an example of a phenomenon that is emerging outside of baseball and throughout the business world. It is called “workforce analytics”: the use of data and statistical analysis by employers to make personnel decisions.
Baseball seems to lend itself ideally to workforce analytics. Every time a baseball player goes to work (likely, even since before becoming a professional), his production is recorded and made publically available. And, while electronic databases and almost infinite computing power are recent additions to baseball’s talent acquisition process, detailed records of player performance have been kept for decades. (In fact, Bill James, who is credited with creating the strategy underlying Moneyball, has been collecting data and publishing non-traditional measures of baseball players’ performance since the 1980s.)
The success of Moneyball (the management strategy, not the movie) has raised a number of interesting questions for practitioners of data analytics in the business world. Specifically, given the limited resources and information available to recruiters, how can companies best hire and retain talent? Professional baseball teams have relatively large budgets and hire relatively few people. They are also able to perfectly observe metrics of prospective employees’ performance. And even then, poor personnel decisions are common. Is there any hope for firms that have substantially less to spend for each hire they make?
In fact, many companies already collect substantial data about their applicants, and some (like Google) are trying to develop their own Moneyball strategies. The challenges, however, are substantial. It is fairly easy to see what makes a good baseball player: hitters getting on base and producing runs, pitchers keeping the other team off the scoreboard. But what makes someone a productive programmer? What should a manager look for when hiring a person to sell a specific product? When comparing qualifications of two similar candidates for a retail manager position, what should be the deciding factor? These are the types of questions workforce analytics tries to answer using data and empirical analysis.
The Atlantic has recently published an article that chronicles a history of talent acquisition, from the free-for-alls of the late 19th century, to extensive personality testing of candidates at all levels 50 years later, to the gradual movement toward more “ad hoc” interviews used today. The common theme across this history is the lack of scientific rigor in matching candidates to job openings. The article, for example, cites a recent study which indicated that even of candidates who successfully went through the recruiting process and were hired, nearly a quarter leave their company within a year of their start date, and that “hiring managers wish they’d never extended an offer to one out of every five members on their team.”
There are basically two pieces of information on which a candidate’s hiring decision is made: the resume and the interview(s) with company staff (though some firms may also ask for a sample of the candidate’s prior work). Both of these steps may be prone to subjectivity, particularly given the interviewer’s own biases. Maybe the candidate went to the same university as the interviewer, or had previously interned at a “prestigious” firm in a different industry. Or maybe the candidate indicated being a supporter of the Boston Red Sox, which had caught the hiring manager’s eye.
Interviewer judgment and intuition alone, unsurprisingly, can be insufficient in assessing talent, particularly over a large number of hires. The Atlantic article, for example, cites Xerox Services, which analyzed thousands of its hires into customer care positions and determined that distance between home and work was a much better predictor of employee engagement and retention than previous experience. (Similarly, Google, one of the leaders in the area of workforce analytics, announced earlier this year that its infamous interview brain teasers were a “waste of time.”)
Companies spend substantial resources and time on talent acquisition. However, matching applicants to openings remains much more art than science (from both the perspectives of employers and candidates). Drivers of what will make employees successful, productive, and engaged in a specific role are generally not well understood, particularly when deciding between many similarly qualified candidates. The intent of workforce analytics is to add empirical rigor to practices characterized by subjectivity and bias. Large volumes of data, more so than ever before, are being gathered. The next step is to turn it into information.