Structured interviews: the complete guide, with templates

What a structured interview is

A structured interview is one where the questions, their order and the scoring standard are set before any candidate is seen. The US Office of Personnel Management (OPM) describes it as a method "designed to measure job-related competencies", in which all responses are "evaluated using the same rating scale and standards for acceptable answers."

Four things separate it from the usual interview conversation:

What the research says

Two meta-analyses anchor the evidence. Schmidt and Hunter (1998) pooled 85 years of selection research. In 2022, Sackett, Zhang, Berry and Lievens showed that earlier work had over-corrected for range restriction, inflating many estimates. Re-run, most validities fell, and structured interviews moved to the top.

On the current estimate, a structured interview carries a little over twice the predictive signal of an unstructured one (SIOP summary). Practitioner data agrees. Google's re:Work guide reports that pre-made questions, guides and rubrics save its interviewers an average of 40 minutes per interview, and that rejected candidates who had a structured interview were 35% happier than those who did not.

The five parts of a structured interview

OPM's practical guide breaks development into eight steps. These five decide whether the interview works.

1. Job analysis and competencies

Start from the work, not from a list of virtues. Write down what the person will actually do in their first year, then name the competencies that work requires. OPM's guide notes that a structured interview typically assesses four to six competencies. Each needs a definition an interviewer can test: "explains a technical cause to a non-technical customer", not "good communicator".

2. Behavioral and situational questions

Behavioral questions ask what the candidate did in a real past situation. Situational questions pose a realistic dilemma from the job. Many interviews use both.

Good questions ask for one episode ("the last time", "the hardest"), not a habit ("how do you usually..."). OPM's guide suggests words such as most, last or worst. Plan two or three probes per question, so every interviewer digs the same way.

3. Anchored rating scales

A scale works only if two interviewers mean the same thing by a 4. A behaviorally anchored rating scale (BARS) fixes that: for each competency, describe what a weak, an adequate and a strong answer sounds like. OPM's guide recommends at least three levels and suggests five to seven. A 1 to 5 scale with written anchors at 1, 3 and 5 is the practical minimum.

4. Independent scoring

Each interviewer scores alone, straight after the interview, before hearing anyone else. OPM's guide is explicit that at this stage each panelist forms "an independent evaluation without discussion with other panel members." Scores collected after a group conversation measure the conversation, not the candidate.

5. A structured debrief

The panel then meets to compare evidence, one competency at a time, and a named person decides. Our debrief agenda covers those 30 minutes in detail.

Build one, step by step: a customer support associate

Here is the whole process for one common role. The steps are the same for any role; only the content changes.

1. List the work. Resolve tickets by chat and email, calm upset customers, find the cause rather than the symptom, and pass clear bug reports to the product team.
2. Name four competencies. Problem diagnosis, clear writing, composure under pressure and ownership.
3. Write two questions per competency, one behavioral and one situational. Eight questions fit a 45-minute interview with room for probes and the candidate's own questions.
4. Anchor each scale. Describe a 1, a 3 and a 5 in one sentence each, in the language of the job.
5. Write the interviewer guide. Question order, timings, probes, and what to tell the candidate at the start.
6. Pilot it on a colleague who does the job well. If they score a 2, the anchor is wrong, not the colleague.
7. Score alone, within 24 hours, then debrief. The scorecard guide covers how to make that happen.

A structured interview template you can copy

Copy these three tables into a document and replace the content for your role. A fuller version for this role is in the customer support associate kit.

Competencies and questions

An anchored scale for one competency

The interview plan

Common objections, answered

"It is too rigid."

Structure fixes the questions you score, not the conversation around them. Interviewers still warm up, follow up and chase an interesting answer. What they cannot do is ask one candidate about system design and the next about their weekend, then compare the two.

"Candidates hate it."

The evidence points the other way. Google reports that rejected candidates who had a structured interview were 35% happier than those who did not. What candidates dislike is silence: in a Greenhouse survey, 61% of job seekers said they had been ghosted after an interview. Structure makes a timely, specific reply easier to write.

"We are too small."

Small teams need it most. With two interviewers and no recruiter, nothing else catches a bad call. A rubric for one role is a few hours of work, and it is reused for every hire into that role.

When structure is overkill

Structure is for the conversations that feed the decision. A scorecard on the others adds paperwork, not evidence.

• Sell and closing calls. Once you have decided to make an offer, the conversation is for the candidate. Do not score it.
• Informal meet-the-team chats, as long as they carry no vote. If an opinion from the chat will be raised in the debrief, it was an interview and needed a rubric.
• Roles nobody can yet define. If you cannot write four competencies, you are not ready to interview. Do the job analysis first.

One test settles most cases: if anything said in the conversation could change the decision, structure it.

Keeping structure up across a whole team

Designing a structured interview is the easy part. Keeping it up is harder: the same questions from every interviewer, scorecards in on time, a debrief that sticks to evidence. That is the part Itya is built for. Each job has one rubric, and Itya can draft the interview kit for your team to edit. Its notetaker joins Google Meet, Microsoft Teams and Zoom, transcribes with speaker labels, and drafts scorecards that cite the moment each answer was given. By default, Itya's AI recommends and a named person confirms; it never rejects a candidate automatically. If you only want the templates, the interview kits are free to copy.