Interview Questions That Actually Predict Performance
Hypothetical questions predict how well someone tells a story, not how well they'll do the job.
Key takeaways
- Ask what someone actually did, not what they would hypothetically do – behavior predicts behavior.
- Push past the rehearsed story with a specific follow-up they couldn't have prepared for.
- Test the real skill directly – a writing sample, a real decision, a live walkthrough – instead of asking about experience with it.
- Use the same core questions and a written rubric for every candidate in a role, not an open conversation.
- Pay attention to the questions they ask you – they reveal judgment too.
"Where do you see yourself in five years?" and "What is your biggest weakness?" persist in interviews because they are easy to ask, not because they tell you anything useful. The candidates who are best at interviews are often best at answering interview questions – a skill that has a weak relationship with the skill you are actually hiring for.
Ask about what actually happened, not what might
Hypothetical questions ("What would you do if...") let candidates describe their best intentions. Behavioral questions ("Tell me about a time...") force them to describe what they actually did, which is a much better predictor of what they will actually do again.
- Tell me about a project that failed, and what you did next – reveals accountability, not just competence
- Walk me through a decision you made with incomplete information – reveals judgment under real conditions
- Tell me about a time you disagreed with your manager – reveals how they handle conflict, not whether they claim to be collaborative
- Describe the worst feedback you have received and what you changed – reveals whether they actually adapt or just say they do
Push past the rehearsed answer
Most candidates arrive with two or three polished stories ready to deploy. The signal is not in the story itself – it is in what happens when you ask a specific follow-up they did not prepare for. "What was the actual number?" "What did your manager say when you told them?" "What would you do differently now?" Vague or evasive answers to specific follow-ups are one of the most reliable tells in an interview.
Test the actual skill, not a proxy for it
- For a role built around writing, ask for a writing sample or a short paid exercise – not a conversation about writing
- For a role built around judgment calls, walk through a real, anonymized decision your company is facing and see how they reason through it
- For a role built around a tool or process, watch them use it briefly rather than asking them to describe their experience with it
- Score the exercise against a rubric you wrote down before the interview, not your gut reaction afterward
A structured process – the same core questions and the same scoring rubric for every candidate in a given role – consistently outperforms an unstructured conversation, even though it feels less personal. It is harder to be fooled by charisma, and easier to compare candidates against each other honestly instead of against your last impression.
The best predictor of future behavior is documented past behavior, not a hypothetical answer polished for the interview.
Finally, leave room in every interview for the candidate to ask you hard questions back. How they interrogate the role – what they push on, what worries them – tells you as much about their judgment as anything you asked them.