The Structured Interviewing Playbook (and Why Unstructured Costs You 22%)

Decades of meta-analyses show structured interviews are 2x more predictive than unstructured ones. The implementation is simple — the discipline is what most teams fail at.

The evidence base

Schmidt & Hunter's century-spanning meta-analysis pegs structured interview validity at 0.51 vs 0.38 for unstructured. The 22% performance gap between hires made under the two approaches is the most replicated finding in IO psychology.

What 'structured' actually means

Same questions, same order, anchored scoring rubric, independent ratings before debrief. The last point matters most — group debriefs without independent scoring drift toward whoever spoke first or loudest.

Common failure modes

Templates that nobody reads. Rubrics that score 1–5 with no anchors ('what's a 4 vs a 3?'). Debriefs that happen days later with no notes. Interviewers freelancing 'just one more question'.

The 90-minute implementation

Pick a rubric template per role family. Train interviewers in one 45-min session. Lock the scorecard inputs so reviewers must rate before seeing peer scores. Audit four sessions in week one. That's the whole programme.