How to Run Structured Interviews: A Step-by-Step Hiring Guide

What is a structured interview?

A structured interview asks every candidate the same predetermined, job-related questions in the same order and scores every answer against the same rating scale. That is the definition the U.S. Office of Personnel Management (OPM) uses in its assessment guidance (2026 page), and it contains three separate commitments: the same questions, the same sequence, and the same scoring standards.

The point is not to make the conversation robotic. The point is that when you compare two candidates, the difference you see comes from the candidates and not from the way each interview happened to unfold. OPM also notes that structured interviews can measure competencies that are hard to test any other way, such as interpersonal skills, and that candidates tend to find them more personal than other assessment methods.

An unstructured interview is the opposite: each interviewer improvises, follow-ups wander, and the final decision rests on a general impression. Most real interviews sit somewhere between the two, which is why it helps to treat structure as a set of dials you can turn up one at a time.

Why does structure produce fairer hiring decisions?

Structure reduces the room for personal preference to decide the outcome. When every candidate faces the same questions and the same scale, an interviewer cannot quietly ask an easier question of someone they like or a harder one of someone who reminds them of a past bad hire.

There is a legal side as well. The U.S. Equal Employment Opportunity Commission (EEOC) explains in its guidance on employment tests and selection procedures (2026 page) that a neutral procedure can still be discriminatory if it has the effect of disproportionately excluding people based on race, color, religion, sex or national origin. Where that happens, the employer must be able to show the procedure is job-related and consistent with business necessity.

The same page says employers may not adjust scores or use different cutoffs by group.

A written question set, a written rubric and saved scorecards give you the evidence trail for that kind of defense. They also let you check your own results over time. The EEOC notes that judging adverse impact ordinarily requires statistical analysis, so keep your interview outcomes in a form you can count.

This guide is general information, not legal advice, and rules differ by country.

How do you build a structured interview, step by step?

Build the interview from the job, not from a list of favorite questions. The order below works for a single role and can be reused as a template for similar roles.

Keep the competency list short. A role with ten required traits cannot be assessed in a one-hour interview, so rank them and test the four to six that matter most. Anything you cannot test in the interview, such as a technical skill, can be covered by a work sample or an assessment instead.

1. Write down what the job requires: four to six competencies or outcomes that separate strong from weak performers in this role.
2. Choose a question type for each competency (see the next section) and write the exact wording.
3. Write a rating scale for each question with a description of what a weak, acceptable and strong answer sounds like.
4. Decide which interviewer covers which questions so that no topic is asked twice and none is missed.
5. Brief every interviewer on the format, the scale and the rules about what not to ask.
6. Run the interviews, take notes during the conversation, and score each answer before the debrief.
7. Compare scores in a debrief, decide, and file the scorecards.

Spend the most effort on the rating scales. A question without a defined good answer is only a prompt for conversation, and two interviewers will read the same answer differently.

Which questions should you ask: behavioral or situational?

Use behavioral questions to learn what someone has done and situational questions to learn what they would do. OPM describes structured interviews as asking about behavior in past experiences, hypothetical situations, or both.

A behavioral question sounds like: "Tell me about a time you had to deliver bad news to a customer. What happened and what did you do?" It suits experienced candidates, because they have a history to draw on. A situational question sounds like: "A teammate hands you work with an error you would have to rely on tomorrow.

What do you do first?" It suits early-career candidates and roles where the situation is predictable.

Whichever type you choose, write follow-up prompts in advance so every candidate gets the same chance to fill in the detail.

• Situation: what was happening and why did it matter?
• Action: what did you personally do, and why that choice?
• Result: what happened, and what would you change?
• Evidence: who else could describe what you did?

Avoid hypothetical questions that are really trivia, and avoid leading questions that reveal the answer you want. Keep the number of questions small enough that you can probe each one properly.

How do you write a rating scale people actually use?

A good rating scale describes observable answers, not feelings about the person. A common design is five points, with written anchors for the low, middle and high ends. Three anchored points also work, and the exact count matters less than having the anchors written down before the first interview.

For a question about handling conflict, the low anchor might read: blames others, no attempt to resolve it. The middle anchor: raises the issue but does not follow up. The high anchor: names the issue directly, agrees a next step with the other person, and checks the result.

Each interviewer then matches the answer to the anchor instead of to a gut feel.

Score each answer immediately after it is given, before the next question. Scoring at the end of the day lets the first and last minutes of the conversation crowd out everything in between.

Pilot the scale on a colleague or a past hire before you use it live. If two interviewers score the same practice answer three points apart, the anchors are too loose and need rewriting.

How should interviewers take notes and score?

Interviewers should write what the candidate said, not what they concluded. Notes such as "led the migration, chose the phased approach, two teams affected" can be checked later. Notes such as "seemed confident" cannot.

Ask each interviewer to score alone before the debrief. If people hear a senior colleague's verdict first, the rest tend to move toward it, and you lose the independent views that a panel is meant to give you.

Keep the scorecards after the hiring decision. They are the record that the process was applied the same way to every candidate, and they are the raw material for the yearly review described later in this guide.

• Write down the candidate's actual words and the specific example they used.
• Mark each answer against the anchors, and note which anchor you matched.
• Hold scores back until every interviewer has submitted theirs.
• Discuss large gaps between scores, and ask what evidence each person saw.
• Record the final decision with the reasons that link back to the competencies.

What questions are off limits?

Questions should only cover what the job requires. Questions about age, marital status, children, pregnancy, religion, health, ethnic origin or similar personal topics invite unequal treatment and can create legal risk, even when asked casually.

The EEOC's guidance on selection procedures describes the legal test as job-relatedness and business necessity, which is a useful screen for every question you write: can you explain how this question connects to a duty of the role? If the answer is no, cut it.

Check local law for the countries where you hire, since protected characteristics and permitted questions differ. Add a short list of prohibited topics to the interviewer brief and have a plan for what to do when a candidate volunteers personal information: note nothing, move on, and do not score it.

How do you run the debrief and make the decision?

The debrief is where structure is most often lost, so run it with a fixed agenda. Start with each interviewer's scores per competency, read aloud or shown on a shared sheet, before any discussion of overall impressions.

Then look at the pattern: strong where the role needs strength, weak only where the weakness can be coached or the work moves elsewhere. Combine scores the same way for every candidate, for example an average per competency with agreed weights, and write the weights down before you interview the first person.

If the group wants to override the scores, make them say why in terms of the competencies. Sometimes the scores missed something real, such as a requirement nobody wrote down, and the fix then is to improve the template for the next round. Overrides without a reason are just the unstructured interview coming back in.

Keep the debrief short and time-boxed, and let the person who chairs it speak last. A chair who speaks first sets the frame for everyone else in the room.

Where does software help with structured interviews?

Software helps most with consistency and record keeping: shared scorecards, interview kits attached to each requisition, scheduling and reminders so interviewers receive the questions in time, and reports on how scores are distributed. Applicant tracking systems usually carry the scorecard features, and some video interview platforms add recorded one-way responses that every candidate answers to the same prompts.

Tools do not create structure by themselves. If the questions and anchors are weak, a scorecard form only stores weak scores neatly. Settle the template on paper first, then check that your system can enforce it, for example by requiring a score on each question before a candidate can move to the next stage.

Our independent software pages cover applicant tracking systems, video interview software and pre-employment assessment software, and each lists what a product does badly as well as what it does well. For help weighing a purchase, see how to choose HR software. Vendors can pay for more visibility; paid placements are labeled and explained in our disclosure.

No commission is earned on clicks or purchases, and we do not work as an affiliate.

A practical test before you buy anything: can the system hold your questions and anchors, require a score on every question, hide other interviewers' scores until yours is submitted, and export results in a form you can count?

How do you check that your interviews are working?

Review the process at least once a year with the data you saved. Look at three things: whether interviewers agree with each other (large, regular gaps point to unclear anchors), whether scores relate to later performance in the role, and whether pass rates differ across candidate groups in ways you cannot explain by job requirements.

The EEOC page notes that such analysis is ordinarily statistical, so involve whoever in your organization handles that work, and involve counsel where the numbers look uneven. Small hiring volumes will not give reliable percentages, so use the review to read the scorecards and the debrief notes qualitatively as well.

Finally, ask new hires and the candidates you rejected about their experience. A structured interview should feel consistent and fair to the person on the other side of the table. After the hire, a regular one-on-one and clear feedback from the first weeks help you compare what the interview predicted with what happened.

Frequently asked questions

What is the difference between structured and unstructured interviews?

In a structured interview every candidate gets the same job-related questions in the same order and is scored on the same scale. In an unstructured interview, each conversation goes wherever the interviewer takes it, and the decision rests on overall impression. Many companies sit between the two, using fixed core questions with a short, planned follow-up section.

How many questions should a structured interview have?

There is no fixed number. Choose enough questions to cover each competency you listed for the role, and few enough that you can probe each answer properly. For many roles that means roughly one question per competency, with follow-up prompts written in advance.

If you cannot score every question, you have asked too many.

Can I still have a casual conversation with the candidate?

Yes, but keep it separate from the scored part. A few minutes at the start to put the person at ease and a stretch at the end for the candidate's own questions are normal. Do not let that talk feed the score unless you wrote down in advance what it would be measured against.

Do structured interviews work for senior roles?

They can, with adjustments. Senior candidates have longer histories, so behavioral questions work well, and you may want a panel of more than one interviewer. The rating anchors should describe senior-level scope, such as the size of decisions made and the effect on other teams, instead of task-level execution.

Who should score the candidate?

Each interviewer should score their own questions independently, before the group debrief. Where several people sit in the same interview, they should score separately and then compare. Hearing a senior person's view first pulls other scores toward it, which defeats the purpose of having more than one view.

Is a structured interview legally safer?

It gives you a stronger record that decisions were based on job-related criteria applied consistently, which matters when a procedure is questioned. The EEOC says a procedure with a disproportionate effect must be job-related and consistent with business necessity. This is general information, not legal advice, so check the rules in each country where you hire.

Sources and further reading

• Structured Interviews, U.S. Office of Personnel Management
• Employment Tests and Selection Procedures, U.S. Equal Employment Opportunity Commission
• Structured interview, Wikipedia