Interview Questions That Actually Reveal Something

You've been here: the candidate who answered every question like they'd read the script. Confident, polished, said all the right words about teamwork and ownership. Three weeks in, you realise they've never actually done the thing you hired them to do — they just knew how to talk about it.

That's the problem with most interview questions. "Tell me about a time you showed leadership" doesn't test leadership. It tests whether someone has rehearsed an answer to that exact question — and by now, most experienced candidates have three of them ready, word for word.

A better question doesn't ask someone to perform a value. It asks them to reconstruct something they actually did, in enough detail that faking it becomes hard work.

Why the standard questions stop working

The usual list — strengths and weaknesses, where do you see yourself in five years, tell me about a challenge — is public. Candidates prepare for it the same way they prepare a CV: polish the story, cut the messy parts, land on the version that sounds best. You're not getting information, you're getting a performance.

The fix isn't cleverer questions. It's questions that ask for detail a rehearsed answer usually doesn't include.

Ask for the parts nobody rehearses

Most prepared stories have a clean beginning, middle and triumphant end. The unrehearsed parts are the boring middle details — what tool they used, who they had to convince, what they got wrong first. Go there directly:

• "Walk me through the actual steps you took, in order."
• "What did you try first that didn't work?"
• "Who disagreed with you, and what did they say?"
• "What would you do differently if you did it again next week?"

Someone who really did the thing can answer these without pausing. Someone who's describing a story they heard, or stretching a small contribution into a big one, starts giving vague answers or repeating the same three sentences with different words.

Ask about failure, specifically

"Tell me about a weakness" gets you a rehearsed non-weakness ("I work too hard"). Ask instead:

• "Tell me about a decision you made that turned out to be wrong. What did you notice first that told you it was wrong?"
• "What's something you used to do that you've stopped doing, because you learned it didn't work?"

The second question is quietly useful because it asks for a change in behaviour over time — hard to invent on the spot, and revealing if someone genuinely can't think of an example. Everyone who has done real work for more than a year has one.

Ask them to reason in front of you, not recall

Recall questions ("tell me about a time...") test memory and storytelling. Reasoning questions test the thing you're actually hiring for. Give them a small, real problem from your world and watch them think:

• For an ops role: "We have 40 orders stuck in a queue and the supplier just told us stock is short. What do you check first?"
• For a support role: "A customer is angry but wrong. Talk me through what you'd say to them, out loud, right now."
• For a finance role: "This month's numbers look fine but cash is tight. What are the first three things you'd look at?"

You're not grading for the "correct" answer — most of these don't have one. You're listening for how they structure a problem they haven't seen before. Do they ask a clarifying question first, or guess? Do they narrate their thinking, or freeze?

Follow up like you actually want to know

The single most useful interview skill isn't the question — it's the second question. Whatever they say, ask one more layer down:

• "Why that, and not the obvious alternative?"
• "What did the other person say when you told them that?"
• "How do you know that worked? What changed?"

A rehearsed answer runs out of detail after the first follow-up. A real one gets more specific, not less — because the person is remembering, not reciting.

What this won't catch

Be honest with yourself about the limits here. A great interview answer proves someone can talk clearly about their work under a bit of pressure. It doesn't prove they can do the work — those are correlated, not the same thing. Someone anxious in interviews but excellent on the job will underperform this method; someone glib and mediocre will do fine.

If the role has a skill you can actually observe — writing, coding, a spreadsheet model, a sales call — the more honest test is to watch them do a small, real version of it. That's a different exercise from the interview and worth doing separately; a short work sample or a structured test (AssessFit runs these adaptively and scores them alongside the interview, for what it's worth) tells you things a conversation never will, no matter how good your questions are.

A short list to bring into your next interview

Pick three or four, not fifteen. Depth beats breadth:

1. Walk me through the actual steps you took, in order, on [a project relevant to this role].
2. What did you try first that didn't work, and how did you find out?
3. Tell me about a decision that turned out wrong. What told you it was wrong?
4. Here's a small problem from how we actually work — talk me through how you'd approach it.
5. Why that approach, and not the obvious alternative?

Write down what they actually said, not your impression of it — impressions blur by the fourth candidate of the day, specifics don't. Then compare notes across candidates on the same questions, not on a general "how did they come across" feeling. That's the difference between an interview that reveals something and one that just confirms whoever performed best on the day.