Ask most hiring managers which they prefer and they'll say structured interviews are "too rigid" or "miss the human element." Ask them how their last three hires have worked out, and the conversation often shifts.
Decades of industrial-organisational psychology research have produced a clear verdict: structured interviews outperform unstructured ones at predicting job performance. Not by a small margin. By a large and consistent one.
The Validity Numbers
In the most-cited meta-analysis of personnel selection methods, Schmidt and Hunter (1998) measured the validity coefficient of different hiring methods — essentially, how well each method predicts actual job performance. Here's how interviews compare:
- Structured interviews: validity of ~0.51
- Unstructured interviews: validity of ~0.38
- Work sample tests (for reference): ~0.54
An unstructured interview is only marginally better than looking at years of experience (~0.18) or checking references (~0.26). The gap between structured and unstructured interviewing is substantial — and it compounds across every hiring decision you make.
Why Unstructured Interviews Feel Better Than They Are
Unstructured interviews create an illusion of insight. You have a natural conversation, you feel like you've really gotten to know someone, and you form a confident impression. That confidence is the problem.
When there are no standard questions, you're unconsciously influenced by the factors that have nothing to do with job performance:
- Whether the candidate reminds you of yourself (affinity bias)
- Whether they made a strong first impression that colours everything after (halo effect)
- Whether you happened to ask harder questions to some candidates than others
- How you were feeling that day
What Structured Interviews Fix
Structure addresses the core problems of unstructured interviewing by:
- Standardising the questions — every candidate gets the same test
- Anchoring scores — rubrics define what good and poor answers look like
- Forcing evidence — behavioural questions require specific examples, not opinions
- Removing improvisation — interviewers can't wander into territory that opens legal risk
The Case for Keeping Some Flexibility
A fully rigid interview — no follow-up questions, no clarification allowed — isn't optimal either. The evidence supports structured interviews, not robotic ones. The best practice is to have a set of standardised questions with space for standardised probing. You follow up on vague answers with the same probes you'd use for any candidate: "Can you tell me more about your specific role in that?" or "What happened as a result?"
Making the Transition
Most managers resist structured interviews because building them from scratch takes time — identifying competencies, writing questions, creating scoring rubrics. That friction is where the resistance comes from, not any genuine belief that gut-feel is better.
Remove the friction and adoption follows. With a scorecard generated in 30 seconds from a job description, there's no longer a practical argument for the unstructured alternative.
Build your structured interview scorecard in 30 seconds
ScoreDesk generates competency-based questions and scoring rubrics from any job description — no setup required.
Try ScoreDesk Free →