(If you subscribe to our newsletter, you've already seen this! And if you don't - Subscribe Here)
I spent much of last week talking new clients through our approach to interviewing candidates, which meant lots of time sharing accumulated anecdotes of the things that have worked in the past - as well as the things that have bombed.
After several conversations, here are the 4 Don'ts that stood out the most. What they all share is low predictive value (though for very different reasons). In other words, using any of these tactics won't really increase your likelihood of correctly identifying the person you need most. They may even decrease your chances.
So without further ado, here are 4 practices to eliminate right now if you want to increase your odds of finding great people.
1. Judge the candidate's behavior
You do want to know what a candidate is really like.
You don't want to determine this by scrutinizing their behavior in the interview.
The Fundamental Attribution Error is the tendency to see our own behavior as a result of external circumstances, but to see others' behavior as a result of their inner character.
I heard a recruiter this week say that if a candidate doesn't ask "why" about a specific software decision, the candidate is "not a curious person."
Sure, that's possible. But I don't know that I'd say it's likely. And there are much better ways to explore someone's curiosity, including validated personality assessments and behavioral interview questions.
Speaking of personality assessments....
2. Use personality assessments however you want
Personality assessments are gaining traction as candidate evaluation tools, and they have a lot to offer if used correctly.
If used correctly.
Valid assessments require the same precision in administering them as you'd need running any scientific experiment.
Moreover, they only tell you what they tell you.
9 out of 10 times I see personality assessment used, they're being used wrong.
Sometimes, people are using assessments no longer considered valid for hiring (cough Myers-Briggs cough).
Sometimes, people decide they'll interpret the results however they like, instead of the way the assessment creators said to. (You should stop using DiSC until you read that.)
Using assessments in a valid way often means using specially trained proctors and can involve a good bit of cost. If you want to use a "personality test," then, make sure you thoroughly explore where and how it is to be used, and what the results do and don't mean.
3. Go by what a candidate says they would do, not by what they've done in the past
Hypothetical questions are used way, way too often in interviews.
"What would you do to fix our website?"
Everyone knows there's a "right" way to answer that question, and now the game is whether the candidate can read the interviewer's mind and deliver the "good" answer.
This is one of the ways you hire someone who interviews well but fails on the job.
I've seen marketing candidates asked to give a "3-month plan," then criticized if they said they'd need more information. ("I'd need more information" is another way of saying "This question is silly and does not replicate a real-world scenario.")
Much better is to find examples from someone's past work that show how they utilized skills or achieved results similar to what you need.
4. Go by what a candidate's done in the past without exploring context
Of course, past behavior does not always predict future behavior - context matters. A lot.
When asking about past behavior, always make sure to note contextual info like team size, management style, deadline pressure, etc. If it's not clear, ask! I'll often ask candidates questions about their exact team size and the interpersonal dynamics, and I'll often explain why I'm asking:
"I'm asking because our team is only 4 people and pretty flat, which means we need people to be especially proactive about offering solutions.")
It doesn't even matter that this is a leading question, because we're going to listen for the quality of the past stories of proactively offering solutions. If the best the candidate has is a story about googling a fix for the Keurig machine, it's probably not enough for us.