Goldman Sachs and Google are both thought of as companies that compete for talent. But they have different ways of spotting it.
One of the best-known—and surprising—results of Google’s internal research into hiring and success is that academic track records don’t really matter. In a New York Times interview (paywall), HR chief Laszlo Bock said university grades are “worthless as a criteria [sic] for hiring.” At the People Analytics Conference April 10-11 at the University of Pennsylvania’s Wharton business school, Bock reiterated the point:
We did a bunch of analysis and found that grades are a little predictive your first two years, but for the rest of your career don’t matter at all.
Other companies, however, have found something different. At the same conference, Goldman Sachs managing director and operations data lead Afsheen Afshar spoke about the investment bank’s own data-intensive efforts to see if it’s missing potentially interesting candidates. The result is a very different way of thinking about the grade-point averages that universities use to measure academic achievement: ”GPA isn’t the whole story, but it is part of it,” Afshar said at a panel discussion. “Leadership went into the analysis thinking it might not mean anything, but it does matter.”
Despite their increasing overlap, Goldman has a very different population of people to Google, Afshar emphasized. Wall Street as a whole tends to hire from a narrow set of elite schools, something else Google says is a bad hiring practice.
Johnson & Johnson, which has built a nascent people analytics group in the past few years, has also found a role for grades, according to its head of human resources, Peter Fasolo. The company has been on a mission to hire more young people after its team found that a bias for more experienced hires didn’t result in materially better performance. The team’s analysis also supported a firm minimum level for GPA, which at least in the US is typically measured on a four-point scale.
“A predictor for us, especially functionally, is GPA,” Fasolo said in an interview with Quartz. “We won’t bring someone in who has a GPA of less than 3.0. That’s a screening mechanism. So why 3.0? We’ve been able to demonstrate that it’s a marker that you’ve demonstrated functional capability. We’ve been able to predict that those with higher GPA come in and at least have the capability to get up and running pretty quickly.”
So why is Google the outlier on the predictive value of GPA? It’s possible that Google just has more and better data than other companies on the relationship between academic grades and performance at work. But it’s just as possible that the absence of such a relationship is, like a lot of hiring practices, unique to Google.