Organization Design

Structured Interviews and What Actually Predicts Performance

Turn the interview from a conversation into a measuring instrument, and find out what the evidence actually says about which methods predict performance — including the correction that cut the famous numbers roughly in half.

  • Intermediate
  • 14 min total
  • 14 chapters

What decision this helps you make: Which parts of your hiring process are producing signal, which are producing confident noise, and what a realistic improvement in hit rate is worth paying for.

What this topic is

A structured interview is one where the questions, their order, the rating scale, and the definition of what each rating means are all fixed before any candidate is seen, every candidate gets the same questions, each interviewer scores independently, and the scores are combined by a rule rather than by discussion. Structure is a property of the process, not a property of the questions. Having a list of good questions and then talking it over afterwards is an unstructured interview with a preamble.

Why it matters

Interviewing is the most expensive decision most managers make and the one they get the least feedback on — you never observe how the rejected candidates would have performed, so the process cannot correct itself from experience. Confidence in an interview judgement is not evidence about its accuracy; the research on that point is unusually blunt. Structure is the only intervention with a consistent, replicated effect, and it is cheap. It is also the intervention that fails most often in practice, because most teams adopt the questions and abandon the scoring.

Who should learn it

Any manager who runs hiring loops, founders designing a process for the first time, and anyone who has sat in a debrief where the loudest person in the room decided the outcome.

What you will understand

  • What structure actually consists of, component by component, and which components carry the effect
  • What the corrected validity evidence says, and why the numbers most people quote are too high
  • How to convert a validity coefficient into the only thing you care about — how often you are right
  • Why the confidence you feel after an interview tells you nothing about whether you were right

Prerequisites

Common misconception

"I can tell within the first few minutes." Almost everyone believes this, and the belief is untouched by evidence because the feedback loop is broken: you find out how the people you hired did, and never how the people you rejected would have done. In one study, interviewers who were told some answers might be random still formed confident impressions from answers that genuinely were random, and rated those interviews as informative. The second misconception is that structure means rigidity, or that it stops you liking a candidate. Structure does not tell you what to think about someone. It stops five people who each thought something different from resolving it by argument, seniority, and whoever spoke first.