Back at Waterloo Engineering, a classmate of mine was asked in a job interview: “What’s the one question you like to be asked?”
“That question,” he said.
The interviewer learned more in those two words than most interviews reveal in an hour: the candidate was quick and comfortable under pressure. Good questions do that. Bad ones produce long answers and no information, and most interview questions are bad.
The Economist made this argument recently in its Bartleby column. It sent me back to what I learned interviewing engineers. Three lessons stuck.
First, don’t wing it. Frank Bernieri and his colleagues once showed untrained students clips of job interviews and asked them to guess how the trained interviewers had rated each candidate. The students matched the professionals on nine of eleven traits. They had watched only the first fifteen seconds: the knock, the handshake, the hello. Everything after the greeting mostly confirmed a first impression.
An unstructured interview is an expensive way of validating a snap judgment. I used the Who method, from Geoff Smart and Randy Street: a scorecard written in advance, a chronological walk through each career, the same questions in the same order for every candidate. It sounds bureaucratic, I know. But when every candidate gets a different conversation you cannot compare them, so you hire whoever you clicked with, which usually means whoever reminds you of yourself. Structure is what lets the next hour overrule the first fifteen seconds.
Second, watch who enjoys it. I expected good candidates to tolerate a demanding process and weak ones to resent it. The opposite happened. The best engineers lit up. Nobody had ever asked them to walk through a career decision in real detail, and they found it fun, the way an athlete prefers a match to a warm-up. Several told me afterwards that they had learned something about themselves in the process. Weak candidates preferred vague questions. Enjoyment of a hard interview turned out to be one of the best predictors of performance I ever found.
Third, the interview is your product demo. Some candidates noticed the preparation and said so, unprompted. That told me they valued craft, which was what I was hiring for. It also told me they were reading the process correctly: a company that puts hours into one interview probably puts the same care into its code and its people.
A candidate may already know your product. What they cannot see from the outside is how you work, and the interview is their one look at that before they sign. It cannot be faked, because faking it means doing it. Nor does it need a big company behind it, which is why it matters more for a startup, not less. Google can run a sloppy interview and let its brand do the work. You can’t.
So write the questions down, ask everyone the same ones, and pay attention to who enjoys answering them.
And if a candidate ever answers “what question do you like to be asked?” with “that question” — hire them.
Photo by Emily Morter on Unsplash.