Structured interviews: a practical guide

What a structured interview is

The U.S. Office of Personnel Management describes a structured interview as an assessment method designed to measure job-related competencies, where candidates are asked the same predetermined questions in the same order and their answers are judged on the same rating scale and standards for acceptable answers. In practice that means three things are fixed before the first candidate is interviewed:

1. The competencies. What the job actually needs, taken from the work, not from a generic list.
2. The questions. Each question targets one or two of those competencies, and everyone for the role gets the same ones.
3. The scoring standards. A rating scale with written anchors that say what a strong, middling and weak answer contains.

An unstructured interview has none of these. It is a conversation that goes wherever it goes, and the score at the end is a general impression. It feels natural, which is why it is so common.

Why it is worth the effort

Research on hiring methods has compared how well different selection procedures predict job performance for decades. A 2022 re-analysis by Sackett, Zhang, Berry and Lievens in the Journal of Applied Psychology corrected what the authors argued were systematic overestimates in earlier work. Many procedures came out with lower validity than previously reported, and structured interviews came out as the top-ranked procedure among those reviewed.

There is a fairness reason too. When every candidate answers the same questions against the same standards, it is much easier to show that decisions were based on the job and to spot where a step is screening out one group more than another. U.S. agencies use the four-fifths rule as a first check for that kind of adverse impact; our compliance guide explains how it works.

Step 1: pick the competencies

Start with the work. Ask the hiring manager what someone in the role does in a normal week, what separates the best person they have had in the role from an average one, and what went wrong with past hires. Turn the answers into four to six competencies, each described in a sentence.

Keep them observable. "Culture fit" is not a competency; "explains decisions clearly to people outside the team" is. If you cannot describe what a good answer about a competency would contain, it is not ready to interview for.

Step 2: write the questions

There are two main kinds of structured question, and a good interview uses both:

• Past behavior: "Tell me about a time you had to tell a customer no. What did you say, and what happened?" These ask for real evidence.
• Situational: "A customer reports what sounds like a security issue while you are handling five chats. What do you do?" These test judgment, including for people without direct experience.

Map every question to a competency, and give every competency at least two questions so one weak answer does not decide it. For a first round, five to eight questions is usually enough. Avoid questions that invite rehearsed answers ("What is your greatest weakness?") or that touch protected characteristics, such as age, family plans, religion or national origin.

Step 3: write the rating scale

This is the step most teams skip, and it is the one that makes the difference. For each competency, write down what a 1, 3 and 5 answer contains. These are called behavioral anchors.

Score each answer against its anchors independently, and write one line on why. Only combine the scores after every answer has been scored, so a great first answer does not color the rest.

Step 4: run it the same way every time

• Same questions, same order for every candidate for the role.
• Planned follow-ups only. "What was the result?" after any example without an outcome is fine if everyone gets it. Improvised probing is where structure leaks away.
• Notes on what was said, not impressions. Recordings help.
• Calibrate. Have two people score the same few interviews and compare. Where they disagree, the anchors usually need tightening.
• Review after each hiring round. Drop questions that do not separate candidates; rewrite anchors that people interpret differently.

How Infyva supports structured interviews

Infyva was built around this method. Questions are generated from your job description so they stay job-related. Generated questions are related but not identical between candidates, so for a strictly structured interview, lock the job to a fixed question bank (Pro and Enterprise) and every candidate hears the same ones. Nova asks each question aloud without improvising follow-ups. Each answer is scored on its own against the job's competencies, with a written explanation, and a competency score cannot sit more than 15 points above the answer it cites. Default scorecards ship with weighted competencies for each Nova profile, and you can edit them. Every answer is recorded, so calibration is a matter of listening to the same recording.

To get the competencies right in the first place, read writing job descriptions for AI screening, and for examples by role see customer support or software engineering.