The Reading Habit That Changed How We Hire
We started asking candidates what they had changed their mind about. The answers told us more than any whiteboard exercise.
For a long time our interviews measured how well someone performed under artificial pressure. We asked people to solve puzzles at a whiteboard while four strangers watched, then wondered why the process kept selecting for confidence rather than judgement.
One question, asked properly
The change was small. We now ask every candidate to tell us about something in their field they used to believe and no longer do, and what changed their mind.
It sounds soft. It is the most informative twenty minutes in our process. The answer reveals whether someone reads, whether they hold their views loosely enough to update them, and whether they can describe a technical idea to someone who does not already share their assumptions.
People who cannot name a single thing they were wrong about have either not been paying attention or are not telling you the truth.
What good answers have in common
The best answers are specific and slightly uncomfortable. Someone describes a position they argued for publicly, what the evidence looked like when it turned, and what it cost to change course. They are usually generous about the people who disagreed with them at the time.
The weaker answers are abstract, or are really stories about someone else being wrong.
- Ask for the belief, not the buzzword.
- Follow up on what specifically changed their mind.
- Notice whether they can explain the old view fairly.
- Let silence sit. The first answer is rarely the interesting one.
It changed the team, not just the process
Two years on, the difference shows up in code review and in planning. Disagreements resolve faster, because people are practised at saying what would change their position. New ideas get a fair hearing, because nobody has staked their identity on last year's architecture.
A note on fairness
This question favours people who have had the chance to be wrong in public, so we are careful with candidates early in their careers. A first-year engineer who describes changing their mind about how to ask for help is telling us exactly what we need to know. The scope of the belief matters less than the honesty of the account.
