Recently my friend Jared Spool posted on twitter about why it’s wrong to give interview candidates design exercises:
This split me right down the line.
First, I think most interviews are bullshit. There’s little evidence interviewing identifies better candidates than other methods do, it’s just one we are all familiar with, and it’s easy, so we do it. Any questioning of job interview practices makes me happy, since we put great faith in something we’ve never examined that is filled with bias. Some interviewers are sadists who never have to justify their cruelty since it’s never questioned.
The ideal way to hire is to use referrals and references of people you trust in your line of work to arrive at candidates. And then, to hire a candidate to do a small project for you where they can demonstrate their ability. That is the best interview in the universe since it is based on what they can actually do for you on real work, rather than an absurdly artificial interview loop of mostly cliche conversations. To do this requires more work and is less efficient, but I’m convinced the results are dramatically better.
However, assuming I’m required to interview candidates, there’s value in asking people to show their ability to do tasks I’m going to hire them to do. If they were a painter, I’d ask them to give their opinion of the paintings in my office. If they were a carpenter, I’d ask them to evaluate the furniture in the room. I’d pick something real, in the present, where their abilities and knowledge should apply, and get out of their way. “How would you make this better?” and hand them a whiteboard pen and step back. I’d want to see what questions they ask and what they know about what they need to know before they can design something. I can do this in a friendly way, and at natural points politely disagree with them, just as I would if we worked together.
The idea of “see how they handle the pressure” is a tricky one. Most work environments are not designed around designers. To be successful a designer needs to be willing and able to persuade, convince and persist in selling or defending ideas. While I’m not certain there is a reliable way to test if a person is capable of doing this during an interview, it’s my job as an interviewer to try. I have to find places to challenge and ask tough, but fair, questions, to let them demonstrate their conviction in their ideas. As well as their ability to change their minds and learn from new information.
Or put another way, if I were hiring an Air Traffic Controller, I’d need to know they can handle the stress of being an Air Traffic Controller. It wouldn’t be unethical for me to expect them to demonstrate something of what I need to see to know they are credible. While the means I use to get this information could be unethical, my interest in getting that data isn’t – it can’t be. It’d be unethical to all the passengers of all those airplanes if I hired someone without even attempting to learn about their abilities for a key part of the job.
All of this is obviously more abstract and collaborative than having them build a chair, or paint or a portrait, while I yelled at them through a megaphone while timing them with a stopwatch. That’s a game show, not a job interview. But if done right a design exercise is a chance for them to lead the conversation through a problem. And if leading the way through a problem is something expected of them in the job, it’s reasonable to provide an opportunity for them to do it during the interview process.
Related: my essay on how to use an hour to interview creatives for a job