This article was originally published in the Behavioral Scientist.
A lot of ink has been spilled and consultant hours racked up talking about the ins and outs of hiring heuristics. What deeply nuanced psychological traits should we be looking for? What tactics can you use to see them? What’s the one surprising trick that will help you identify the exceptional among the average?
According to some of the so-called savviest of executives, the secret to identifying future stars lies in the following interview questions:
1. “On a scale of one to ten, how weird are you?” (Tony Hsieh, Zappos founder)
2. “What’s your spirit animal?” (Ryan Holmes, Hootsuite)
3. “What would you do in the event of a zombie apocalypse?” (Ashley Morris, Capriotti’s)
4. “What was the last costume you wore?” (David Gilboa, Warby Parker)
While these prompts might lead to delightful conversations, no candidate’s response will indicate how well they’ll be able to do a particular job. Not to mention: How does one stack-rank applicants who most recently dressed up as an ice cream sundae or a box of crayons?
The truth is, the best way to hire isn’t outrageous, groundbreaking, or clickbaity. It’s incredibly structured and boring. And that’s why no one does it.
Too many people see hiring as an instinct art form, honed by years of their own experience: when asked, three-fourths of people involved in the interview process at elite law, banking, and consulting firms admitted to making hiring decisions based on their gut.
But our instincts are rarely, if ever, scientifically valid. Humans are error-prone creatures. We get excited about people who have impressive zombie apocalypse strategies or send thank you notes—and then we hire them over candidates who might have become our best-performing employees.
To test how helpful interviews really are, Yale professor Jason Dana and his colleagues had two groups of students predict their classmates’ GPAs. One group had access only to past GPAs and current course lists, while the other was also allowed to conduct free-for-all interviews. The students who interviewed their classmates were over 50 percent worse at predicting future GPA.
To fix the hiring process, we have to replace hubris with heuristics. Here are five steps to identifying the very best candidates. **
**Warning: the rest of this article might seem dry, but I promise it will improve the quality of your hires.
#1: Define job attributes
The goal of any interview process should be to predict how candidates will perform once they
join the team. Start by clearly outlining what missing skills your team needs to succeed, and identifying themes across people you did hire, as well as those you didn’t. At Google, we found four distinct attributes that predicted whether or not someone would be successful on the job: general cognitive ability, leadership, “Googleyness” (e.g., a dose of humility and comfort with ambiguity), and role-related knowledge. Once we identified these attributes, we required all interview feedback to comment specifically on each one.
#2: Ask for a work sample
This entails giving candidates a sample piece of work, identical to that which they would do in the job, and assessing their performance on it. But even this can’t predict performance perfectly, since actual performance also depends on other skills, and worse, many jobs don’t have nice, neat pieces of work that you can hand to a candidate.
#3: Ask behavioral questions
Which brings us to interviews. Asking behavioral questions (something like, “Give me an example of a time when you solved an analytically difficult problem”) will get you two kinds of information: one is you get to see how the candidate actually interacted in a real-world situation, and the valuable “meta” information you get about the candidate is a sense of what they consider to be difficult.
Ask candidates a consistent set of questions, in the same order, with clear criteria to assess the quality of responses. This ensures that any variation in responses is a result of the candidate’s performance, not because an interviewer has particularly high or low standards, or asked harder or easier questions.
#4: Average scores and make a decision
The “wisdom of crowds” works for hiring decisions, too. An average score has the virtue of eliminating the ability of a single person to veto or politic for a candidate. Equally weight each interviewer’s rating—a subordinate’s feedback is at least as valuable, if not more so, than a hiring manager’s feedback. And finally, when you sit down to compare candidates, look at and rank all the scores on the first attribute or question. Then look at and rank all the responses on the second one, and so on. This prevents you from focusing on a single candidate for too long, which is when bias creeps in.
#5: Constantly check that your hiring process actually works
Most organizations never go back a year or two after hiring someone to find out if that thank you note really did predict their performance—which is why they keep letting jokes and notes shape their staff. Continue to identify themes across your top performers, and then work your findings into the interview process.
Following these five steps might not feel exciting or entertaining, and in fact that’s why most companies don’t bother. Sometimes doing something right means doing it in a staid, measured, repeatable way. But by doing so you’ll discover that structure and science (not sentiment or social cues) are the keys to a successful—and unbiased—hiring process.