Interesting article in the NY Times about what Google has learned about hiring over the years. The highlights:

1) GPA & standardized test scores were uncorrelated with employee performance.

2) Ability to answer brainteasers was uncorrelated with employee performance.

3) Behavioral interviews were effective in predicting employee performance.

In 2007, I shared with Google my SAT scores, transcripts, and sat through multiple brainteasers. Given that experience, I was intrigued to learn they’ve since abandoned this approach. Three personal reactions:

1) Behavioral interviews have long been the gold-standard. My alma-mater, Procter & Gamble, used the behavioral framework as their interview centerpieces. “Tell me about a time when you…” I think it’s great that Google has come around to this method. I agree w/the Googler’s quote that they are a great view inside the candidate’s thinking process, standards, and past behavior. More so than pointing out that Google apparently eschewed a well-known practice in their early years, I wanted to tip my cap to P&G for getting this right since back in the day.

2) Interviews often reveal more about the company than they do the candidate. When Google asked for SAT scores, transcript, and brainteasers, they revealed a lot about their culture. In the 6 years since I interviewed at the big G, I’ve come to know dozens of ex-Googlers. They’ve pretty consistently shared stories of a culture that valued intelligence above all else. At the time I thought it was kind of myopic/snobby how much they focused on where I went to school, my high school class rank (!), etc.  But based on what I’ve heard since, I guess that was part of the culture of the place; it reflected their view that they only wanted the smartest folks, and that those folks got the highest SAT scores, best grades, etc. By contrast, in interviews P&G obsessively looked for strong predictors of leadership and teamwork. Naturally, P&G paid attention to where you went to school and how you did,  but only as a minor consideration compared to what you had done. That very much reflected how the company worked. P&G cared much less about who the smartest person was, and much more about teamwork and running effective processes. (Of course, this rigorous focus on process is why P&G can also feel suffocating to some.)

3) Interviews should focus on asymmetric topics. “[Brainteasers] don’t predict anything. They serve primarily to make the interviewer feel smart.”  This is a key plot point in the article. Many interviewers focus on topics where there’s a large knowledge asymmetry. It’s natural for interviewers, like all humans, to prefer to feel smart. Ironically, this is something that I thought Google did well in my experience. In my phone screen, I was asked the classic Google question: “Teach me something.” I saw it as a good window to the candidates communication skills, how difficult of a task they chose to explain, etc. Personally, I like a hybrid approach: a lightweight case study merged with the candidate’s area of expertise. Interviewing a college student for P&G, I might ask: “If you were the Brand Manager of Cascade and wanted to target college market, how would you think about that?” Instead of using interviews as an opportunity to get confirmation of knowing more than the candidate, use as an opportunity to learn from them while assessing how they go about it.

Final thought: Square peg, round hole. The interview process is really just the endgame. When a company doesn’t really know what it’s looking for, it’s not going to know how or when they find it. Understanding that it needs a 2” round peg is the same as understanding what the desired employee needs to accomplish, the kinds of skills they’ll need, the personality traits to work effectively in that organization, etc. Else, the company ends up picking the shiniest square peg, in the coolest color, that has the right “feel” to it. Hiring can be random without a strong grasp of what’s needed. That’s probably why it’s so easy to cheat and focus on the person’s likability, or some other random criteria.

One reason P&G is so good at hiring is because it’s a completely known quantity: for close to 200 years they’ve looked for talented, driven folks to plug into a system. Since they know exactly what they need, they’ve close to perfected that process. In hyper-growth companies in rapidly changing industries, it’s a lot harder. That’s why Google is doing this analysis and why hiring seems like such a crapshoot in Silicon Valley. This makes Google hiring such strong talent, at massive scale, even more impressive. Always room for improvement, and the referenced article shows that Google agrees/is getting even better.