Structured Interviews: The Highest-Return Change Most Teams Skip

Unstructured interviews are the most trusted and least reliable step in most hiring processes. Interviewers routinely report high confidence in judgements formed in the first few minutes, and those judgements predict job performance poorly.

Structure is the fix, and it is cheap. The reason teams skip it is not cost — it is that structure feels like it removes the part of interviewing that experienced interviewers value most.

What structure actually means

Four things:

1. The same questions, in the same order, for every candidate for that role.
2. A defined rubric — written before the interviews, describing what a weak, adequate and strong answer contains.
3. Independent scoring — each interviewer scores before discussing with the others.
4. Notes on evidence, not impressions. "Described rolling back a migration under load, walked through the decision" rather than "seemed technical."

That is it. Nothing about tone. A structured interview can be relaxed and conversational — the structure is in what you ask and how you score, not in how it feels.

Why unstructured interviews fail

First impressions dominate. Judgements form in the first minutes and the rest of the interview gets spent confirming them. Interviewers reliably experience this as "getting a read on someone quickly."

Questions vary with rapport. When an interviewer likes a candidate, the questions get easier and more open. When they do not, the questions get sharper. The candidate is then evaluated on a different exam.

Similarity gets mistaken for quality. Shared background, shared vocabulary, shared references — all produce a sense of ease that is genuinely pleasant and has no relationship to whether someone can do the job.

Nothing is comparable afterwards. Two interviewers who asked different questions cannot meaningfully compare candidates. The debrief becomes a contest of conviction, usually won by whoever is most senior or most certain.

Building one

1. Start from the scorecard

The interview should test what the screening criteria could not. If you have already built a hiring scorecard, the resume-verifiable items are done — the interview is for the rest.

This is also where "culture fit" and "detail-oriented" go to become real. They cannot be read off a resume, but they can be probed with a specific question and a real rubric.

2. Write four to six questions

Fewer than you think. A good behavioural question plus genuine follow-up takes ten to fifteen minutes.

Behavioural — past behaviour, specific instance:

Tell me about a time you disagreed with a technical decision your team had already committed to. What did you do?

Situational — hypothetical, relevant:

You join and find the deploy process takes four hours and everyone has quietly accepted it. What are your first two weeks?

Avoid: brainteasers (predict nothing), "what's your greatest weakness" (tests interview practice), and anything answerable with a rehearsed narrative.

3. Write the rubric before you interview

For each question, describe what the three levels contain. Concretely:

Question: Time you disagreed with a committed technical decision.

Weak (1): Either never disagreed, or escalated immediately without attempting to resolve it directly.

Adequate (2): Raised it with the decision-maker, presented a case, accepted the outcome.

Strong (3): Raised it with evidence, sought to understand the original reasoning, changed their own position or the decision on the merits, and can describe what they would do differently.

Written before, this is a rubric. Written after, it is a rationalisation of a decision already made.

4. Score independently, then discuss

The single most important procedural detail. If interviewers discuss before scoring, the most confident voice anchors the room and you have converted several independent signals into one.

Everyone submits scores, then you discuss — and discuss the disagreements specifically. A candidate scored 1 and 3 on the same question is the most informative thing in the debrief.

5. Keep the notes about evidence

"Good communicator" is not a note. "Explained the caching strategy to me without jargon and checked I was following" is a note. The first is a conclusion, the second is the thing the conclusion should rest on.

The objection worth taking seriously

Experienced interviewers often say structure removes their ability to follow an interesting thread.

It does not, and this is worth being clear about: structure fixes the question set and the rubric, not the conversation. Follow-ups are encouraged — the requirement is that every candidate gets the same starting questions and is scored on the same scale. Probe as deeply as you like within that.

What structure removes is the ability to ask candidate A about their side projects and candidate B about their weaknesses, and then compare the two. That comparison was never valid.

What it costs

Real, but front-loaded: a couple of hours per role to write questions and rubrics, reusable across every subsequent hire for that role. Roughly fifteen minutes per interviewer per interview for independent scoring.

Against that: interviews get shorter and more decisive, debriefs stop being arguments, and rejected candidates can be given a real reason — which matters increasingly for compliance as well as for basic decency.

Where it sits

Structured screening and structured interviewing are the same idea applied at two stages: decide what you are measuring before you look at anyone, and measure everyone the same way.

Automated screening without explicit criteria produces confident nonsense. Interviewing without a rubric does exactly the same thing — it just feels better, because a human produced it.