Structured interviews: better decisions, less bias

A structured interview asks every candidate for a role the same core questions, in the same order, scored against the same written criteria immediately after the conversation. It is the single most evidence-backed improvement available in hiring: decades of selection research consistently find structured interviews predict job performance substantially better than unstructured conversations — which, despite feeling insightful, perform closer to coin flips than their users believe. Structure works because it forces comparison on the job’s criteria instead of on rapport.

Building the question set

• Derive questions from the role’s real demands — take the three to five outcomes the job must deliver and write questions that probe each one. A cashier role probes accuracy under pressure; a manager role probes feedback and conflict.
• Prefer behavioral questions — "tell me about a time you..." — because past behavior in real situations predicts better than hypotheticals, which measure imagination and coaching.
• Follow the STAR shape when probing answers: the Situation, the Task, the candidate’s specific Actions, and the Result. Vague answers collapse under "what did you do, specifically?"
• Keep the set small — six to eight core questions fill a 45–60 minute interview once follow-ups are counted. Depth on few beats coverage of many.

Scoring: where structure actually lives

The question set is the visible half; the rubric is the half that changes outcomes. For each question, write in advance what a strong, adequate, and weak answer looks like, and score each candidate against that scale immediately after the interview — not at the end of the week, when memory has already reranked everyone by likability. When multiple interviewers are involved, they score independently before discussing; the discussion then reconciles evidence instead of harmonizing first impressions. Keep the scores: they are your record of why the decision was made, and — reviewed against actual job performance a year later — the only way your interview process itself ever improves.

The common objections, answered

• "It feels robotic." Structure fixes the questions, not the humanity — warmth, follow-ups, and genuine conversation all remain. Candidates broadly experience structured interviews as fairer, because they are.
• "I trust my gut." Everyone does; the research is precisely that this trust is misplaced. Gut feel gets its say — inside the rubric, as evidence, where it can be compared.
• "It is slower to prepare." Building the question set costs an afternoon once per role, and repays it in every interview afterward — and in every mis-hire avoided.

What interviews cannot do

Even perfect interviews test the present — how a person thinks, communicates, and reasons about work. They cannot verify the past. The interview layer should therefore sit on top of verified facts, not substitute for them: confirm the employment history and credentials of finalists (candidates with EmployDB verified histories make this a minutes-long consent grant), then spend the interview on what only interviews can measure. Facts first, judgment second — each tool doing the job it is actually good at.