There is a certain kind of interview question that makes the interviewer feel clever and tells them almost nothing. How many golf balls fit in a bus. Why are manhole covers round. Sell me this pen. These have mostly gone out of fashion, and good riddance, but they have often been replaced by something only slightly better: broad, abstract questions like "what is your greatest weakness" that any candidate with an afternoon of preparation can answer in their sleep. Both kinds share a flaw. They measure how well someone performs in an interview, not how well they do the actual job.
The questions that genuinely reveal how a person works have a different shape. They are specific, they ask for a real example from the candidate's own experience, and they are built to be followed. You ask about something that actually happened, and then you keep pulling the thread until you reach the part that could not have been rehearsed. What follows are twelve prompts organised into three areas, along with what each one is really trying to surface. The questions matter, but as you will see by the end, how you handle the answer matters even more.
Why behavioral beats clever
The principle underneath all of these is simple. The best predictor of how someone will behave is how they have behaved before, described in concrete detail. A hypothetical question, "what would you do if a project were falling behind," invites a candidate to describe their best self in theory. A behavioral question, "tell me about a time a project fell behind and what you did," forces them onto the record of what they actually did, with all the mess and compromise real situations involve. The gap between the polished hypothetical and the honest specific is exactly the information you are looking for.
That is why every question below asks for a real instance, and why the interviewer's job is to keep asking for more detail. A rehearsed answer is smooth and general. Reality is textured and specific. The moment a candidate stops reciting and starts remembering, you are finally learning something true about how they operate.
On the work itself
These four get at how a person actually does their craft: how they think, where their standards sit, and how honest they are about their own record.
- Walk me through a project you are proud of, and be specific about which parts were yours. Pride reveals what someone values, and the ownership question separates people who did the work from people who were nearby when it happened.
- Tell me about a decision you got wrong. How did you realise, and what did you do next. This surfaces self-awareness and how someone recovers, which matters far more than never being wrong.
- What is something you have changed your mind about in your field in the last year. A person who cannot answer has either stopped learning or is not being candid.
- Describe a time the requirements were unclear or contradictory. What did you actually do. Real work is ambiguous, and how someone acts without a clear brief tells you how they will handle most of their days.
On working with others
Very little valuable work happens alone. These four look at how a person handles the friction and the collaboration that fills most jobs.
- Tell me about a time you disagreed with a teammate. How did it play out. You are listening for whether they can hold a position and still keep the relationship, or whether it was all heat or all surrender.
- Who is the best person you have worked with, and what specifically made them good. This quietly reveals the candidate's own standards, because people describe the qualities they admire and aspire to.
- Tell me about a piece of feedback that was hard to hear. What did you do with it. How someone metabolises criticism predicts how they will grow on your team.
- How do you keep people informed when a project is slipping. Handling good news is easy. How someone communicates when things are going badly is where you learn whether they can be trusted with something important.
On depth and growth
The last four probe how deep a person's knowledge really goes, and where they are trying to take it.
- Teach me something you know well, in five minutes, as if I am new to it. Genuine depth shows up in how clearly someone can explain the thing they know, not in how many terms they can name.
- Looking back at your last significant project, what would you do differently. This tests reflection. A thoughtful, specific answer shows someone who learns from their own work rather than just finishing it.
- What is a common belief in your field that you think is wrong, and why. This reveals whether someone has opinions earned through experience or is simply repeating the consensus.
- What do you want to be doing that you cannot do in your current role yet. This tells you where their ambition points, and whether the role you are hiring for actually feeds it.
The follow-up is the real question
Here is the part most interviewers miss. The twelve questions above are not the point. They are openers. The signal lives in what you do after the first answer, and the tool for extracting it is almost embarrassingly simple: some version of "tell me more." Ask what exactly they did, not what the team did. Ask what they would do differently now. Ask what the hardest part was and how they got through it. Ask for the detail behind the summary.
A rehearsed answer holds up for one round and starts to thin out under the second and third. The candidate who really led the project can go deeper on request, describing the trade-off they agonised over and the thing that nearly went wrong. The candidate who was merely present runs out of specifics quickly and drifts back to generalities. You do not catch that by asking a cleverer opening question. You catch it by refusing to accept the first, smooth answer and gently asking for the layer underneath. Silence helps too. Let a pause sit, and people will often fill it with the honest detail they were about to leave out.
What a good follow-up sounds like
It helps to see the difference in action. Suppose you ask a candidate to walk you through a project they are proud of, and they say they led the launch of a new checkout flow that lifted conversion. That is a fine headline and tells you almost nothing yet, so you follow. What was your specific part, as opposed to the team's. They led it, they say, and coordinated three engineers. Good. Now go sideways: what was the hardest trade-off you had to make. Here is where it gets real, because a genuine owner remembers the trade-off vividly, the feature they cut to hit the date, the disagreement with design they had to resolve, the risk they took that nearly did not pay off. Someone who was merely adjacent to the work starts to thin out here, drifting to generalities about teamwork and communication.
One more layer: knowing what you know now, what would you do differently. The strong candidate has already thought about this and answers with something specific and a little self-critical. The weaker one says they would not change much, which is almost never true of real work and usually means they were not close enough to it to see the flaws. Three plain follow-ups turned a rehearsed headline into a clear picture of how much of that success actually belonged to the person in front of you. No clever opening question could have done that. The most ordinary openers, followed persistently, did.
Adapt the set, and know your red flags
Treat the twelve as a starting bank, not scripture. A role that is mostly independent work needs more weight on the questions about craft and depth, and less on collaboration. A role that lives at the centre of a team needs the reverse. Pick the handful that matter most for the specific job and let the rest go. What you should not do is change the set candidate by candidate on instinct, because then you are back to comparing people on different tests, which is the whole thing structure exists to prevent.
It also helps to know in advance what a worrying answer looks like, so you are not swayed by smooth delivery. Be wary of the candidate who cannot name a single decision they got wrong, or who blames every past difficulty entirely on other people, or whose examples stay vague no matter how many times you ask for specifics. None of these is automatically disqualifying, but each is a signal worth probing rather than glossing over because the person is likeable and speaks well. The whole reason you are asking behavioral questions is to see past the polish, so do not let the polish win at the final moment.
Make it a system, not a performance
Finally, do not treat these as a grab-bag to improvise from differently with each candidate. That drops you straight back into the unstructured trap, where everyone gets asked different things and nobody can be fairly compared. Choose the questions that fit the role, ask the same core set of every candidate, and score the answers against a scorecard right after each conversation while the detail is fresh. Then compare like with like.
Good questions and a fair process reinforce each other. The questions get you honest, specific material about how a person actually works. The structure lets you weigh that material fairly across everyone you talk to. Neither is much use without the other. Ask for real examples, follow every claim with "tell me more," write down what you heard, and compare candidates on the same ground. Do that, and the interview stops being a test of who performs best in a room, which is a real but largely useless talent, and becomes what it was always supposed to be: an honest read on who will do the best work once the room is empty and the job is real. The candidate who interviews beautifully and the candidate who does the work well are sometimes the same person and often are not, and the whole purpose of a specific question followed by a patient one is to tell them apart. Get the questions specific, keep the follow-ups patient, and you will hire the second kind far more often than luck alone would allow. And because every judgment rests on a specific example rather than a general impression, you will be able to say precisely why you chose the person you chose, which over a run of hires is worth almost as much as choosing well in any single one.