Panel Interview: How to Conduct One as an Employer
How to conduct a panel interview: who belongs on the panel, how to divide the questions, how to score independently, and the mistakes to avoid.
Panel Interview
How to run one from the employer side: who earns a seat and who should not have one, how to split the questions so panelists stop asking the same thing three ways, a question bank organized by panelist role, why every panelist scores alone before anybody speaks, how long the session should actually run, and the specific mistakes that make a panel worse at picking people than a single well-prepared interviewer
The first panel interview I ran was four of us and one candidate, and it was a disaster nobody in the room noticed at the time. Three of us asked versions of the same question. The candidate answered the loudest person and largely ignored the other two. Afterwards we talked for twenty minutes and agreed she was great, which is a sentence I now recognize as the sound of four opinions turning into one.
We hired her. It worked out, and that is the part that worried me later, because nothing about the process deserved the outcome. What we had built was an audience, not a panel.
This is the employer side of the panel interview: who earns a seat, how to divide the questions so panelists stop covering the same ground, a question bank organized by panelist role, why everyone scores alone before anybody speaks, how long the session should run, and the mistakes that make a panel worse than a single well-prepared interviewer. I build the hiring and people tooling for businesses without a dedicated HR department at FirstHR. This is general information rather than legal advice.
What a Panel Interview Is
A panel interview is one session in which several interviewers meet a single candidate at the same time, each assessing the candidate against competencies agreed in advance. It is not a group interview, where several candidates are assessed together, and it is not a loop of separate meetings held back to back.
The distinction that matters operationally is between a panel and a sequential loop. A panel compresses several assessments into one hour of the candidate’s time and yours, and every panelist hears the same answer to the same question. A loop spreads the same assessments across a morning and gives you several independent conversations at the cost of repeating the basics several times.
| Format | What it is | Best for | The cost |
|---|---|---|---|
| Panel interview | Several interviewers, one candidate, one session | Final rounds, roles with several genuine stakeholders, tight calendars | Contamination between raters if scoring is not independent |
| Sequential loop | The same interviewers, one candidate, separate sessions | Senior roles, technical depth that needs its own hour | Half a day of candidate time and repeated introductions |
| One-on-one interview | One interviewer, one candidate | Early rounds, screening, small teams | One person’s judgment carries the whole decision |
| Group interview | One or more interviewers, several candidates together | High-volume hourly hiring where interaction is the signal | Weak individual evidence and a poor candidate experience |
Most small businesses do not need a panel at every stage. The screen and the first conversation work better one to one, which is why the phone screen stays with a single interviewer and the panel arrives at the point where several people genuinely have to live with the decision.
What the Research Actually Shows About Panels
Panels do not raise the accuracy of a hiring decision on their own. Structure does. That is the single most important thing to understand before you spend four calendars on one candidate.
That result surprises people who assume more raters must mean better judgment. The likely explanation is the one you can watch happening in any unstructured panel: the ratings are not independent. Panelists hear each other’s follow-ups, read each other’s reactions, and converge before anyone writes a number down.
The same meta-analysis is worth reading for what it says about question content. Situational interviews reached .50 for job performance criteria, ahead of job-related interviews at .39 and psychologically oriented interviews at .29. The earlier work by Wiesner and Cronshaw that McDaniel and colleagues summarize found the highest corrected validity of all, .64, in structured board interviews using consensus ratings.
Read those two findings together and the instruction is clear. A panel is not a substitute for a structured interview, it is a delivery mechanism for one. If you are choosing between adding people to the room and adding structure to the questions, the evidence says fix the questions first. The gap between structured and unstructured interviewing is larger than the gap between formats.
Who Belongs on the Panel
Every seat on the panel exists to assess a specific competency, and anybody who cannot name theirs does not have a seat. That single test removes most of the people who end up in panel interviews at small companies.
The Office of Personnel Management guide to structured interviews says a panel of two or three interviewers may be better able to document and interpret what a candidate says, and that a panel also reduces the risk of bias in ratings and allows a more diverse set of interviewers. Three is the working default. Four is defensible when the role sits between functions. Five is a room, not a panel.
Two seats are more often harmful than useful. The first is the person invited because they would be offended not to be asked, which is a political decision dressed as an assessment. The second is the future direct report of the candidate, on a management hire, placed in the same room as the owner. Subordinates almost never contradict the boss in front of the boss, so if you want their read, give them their own conversation instead.
Consistency across candidates matters more than the exact composition. The same OPM guidance recommends using the same interviewers across all candidates wherever feasible, because ratings are only comparable if the raters are. A panel that changes between the first candidate and the fourth produces four scores that cannot be laid side by side, which is the same failure mode that bias reduction work is trying to prevent everywhere else in the process.
How to Divide the Questions
Assign each panelist a named block of competencies and a fixed set of opening questions, in writing, before the day. Overlap is the most common reason a panel produces less evidence than a one-to-one interview would have, and it is entirely preventable.
Start from the core competencies for the role rather than from the list of people you invited. Write down the four or five things the interview has to establish, then map each one to exactly one panelist. If a competency has no owner, either assign it or accept that you are not testing it. If a panelist owns nothing, remove them.
| Panelist | Competency block | Question type | Airtime |
|---|---|---|---|
| Hiring manager | Core job capability, prioritization, ramp | Behavioral plus one situational | The largest block |
| Future peer | Collaboration, conflict, resilience under pressure | Behavioral | A full block |
| Internal customer | Handoffs, communication across functions | Behavioral plus a real scenario | A short block |
| Craft assessor | Technical or professional depth | Work sample or situational | A narrow, deep block |
| Owner or senior leader | Motivation, trajectory, deal-breakers | Open, forward-looking | The shortest block |
Fix the opening question of each block and let the probes move. That is the compromise that keeps the interview comparable without making it robotic: every candidate faces the same starting questions in the same order, and each panelist chases whatever was thin in the answer they received. Our behavioral interview question templates are a reasonable starting bank to cut down, and the STAR method gives the panel a shared way to tell a complete answer from a rehearsed one.
A Panel Question Bank by Panelist Role
These are questions organized by seat rather than by topic, so you can hand each panelist their block and stop worrying about collisions. Cut them down to the three or four per block you will actually use, and ask the same ones of every candidate for the role.
Situational questions of the kind in the craft assessor block are worth over-weighting, given that situational content tested strongest in the meta-analytic data. The most useful version uses a real problem from your own business in the last quarter, because candidates cannot have prepared an answer and you already know what good looks like.
How to Run the Room
One panelist chairs, and it is usually the hiring manager. The chair opens the session, introduces every panelist by name and role, explains the format to the candidate, keeps time, hands cleanly between blocks, and closes with what happens next and by when.
OPM guidance describes exactly this coordinator role: one panel member designated as chairperson, responsible for the logistics and for documenting the process, who welcomes the candidate, introduces each panel member, describes the job, explains the interview process in a standardized way, tells the candidate notes will be taken, and asks whether they have questions before starting.
The panel format is harder on the candidate than a one-to-one, and SHRM notes the trade-off directly: panels bring multiple viewpoints simultaneously and enhance reliability and objectivity, but they can intimidate a candidate and limit rapport, which affects the depth of the answers you get (SHRM interviewing toolkit). Telling candidates in advance who will be in the room and what each person covers costs nothing and recovers most of that loss, which is the cheapest candidate experience win in the whole process.
Score Independently, Then Discuss
Every panelist scores every competency alone, in silence, before anybody says a word about the candidate. This is the rule that separates a panel that adds information from a panel that adds noise, and it is the one most often broken.
The OPM guide is unusually specific about the sequence. Each panel member reviews their notes immediately after the candidate leaves the room and rates the candidate, forming an independent evaluation without discussion with other panel members. Ratings should be specific, defensible, and supported by behavioral examples. Only after that do panelists compare notes and ratings, explore the basis for discrepancies, and reach a consensus recorded on a group rating form (OPM, Structured Interviews: A Practical Guide).
Use one scale with defined levels and use it everywhere. It does not matter greatly whether it runs to four points or five, but it matters enormously that a three means the same thing to the peer as it does to the owner, and that each level is anchored to observable behavior rather than to a feeling. An interview evaluation form that carries the anchors on the page is the simplest way to keep panelists honest under time pressure.
Write the feedback while the notes are fresh. The panel discussion is where the substance of interview feedback exists, and a week later all that survives is the verdict. Small businesses that keep the scoring sheets against the requisition also get something more valuable over time: a record of which panelists predict well and which ones simply have opinions.
How Long a Panel Interview Should Run
Forty-five to seventy-five minutes covers most small business roles, and the length should follow the number of competency blocks rather than the number of people in the room. Budget roughly ten to fifteen minutes per block plus the opening and the candidate’s own questions.
| Segment | Time | Owner | What it produces |
|---|---|---|---|
| Panel briefing before the candidate arrives | 10 minutes | Chair | Agreed blocks, agreed scale, no surprises |
| Opening and introductions | 5 minutes | Chair | A candidate who understands the format and relaxes |
| Competency blocks | 10 to 15 minutes each | Each panelist | The behavioral evidence you will score |
| Candidate questions and close | 10 minutes | Chair | Their read on the job, plus a dated next step |
| Silent independent scoring | 10 minutes | Every panelist | Uncontaminated ratings with evidence attached |
| Score reveal and discussion | 15 minutes | Chair | A recorded decision and a written rationale |
Book the scoring time in the same calendar invitation as the interview. Panels that leave scoring for later reliably do it from memory two days on, and a decision made from memory is a decision made by whoever remembers most confidently. Sessions running beyond ninety minutes rarely add information: answers flatten, panelists start improvising, and the candidate stops being at their best in a way that tells you nothing about the job.
The Legal Exposure Multiplies With the Panel
Every additional interviewer is another person who can ask a question you would never have asked, and in a panel it is asked in front of witnesses and recorded in several sets of notes. This is the strongest practical argument for briefing the panel before the session.
The EEOC states that information regarding race, sex, national origin, age, and religion is irrelevant to determining job qualifications, that pre-offer inquiries about disability are prohibited, and that questions which would reveal an applicant’s race, sex, national origin, disability status, age, religion, color, or ancestry should generally be avoided (EEOC, prohibited employment policies and practices). A panelist making friendly conversation about kids, church, or how long somebody has been in the country is creating a record, not building rapport.
The documentation side of a panel is an asset rather than a burden. OPM lists what to retain: the interview questions, scores, and notes for each candidate, the training provided to interviewers, and the interview guides and rating scales used. A structured panel with independent score sheets is the best-documented selection decision a small business is likely to make, provided the paperwork survives past the offer. That is a good reason to keep the whole hiring process and its records in one system rather than in six inboxes, and it pairs with the same discipline you should be applying to the reference check at the end.
Where Panel Interviews Go Wrong
Six failure patterns account for nearly everything I have watched go wrong, and the first two do most of the damage.
Discussing before scoring is first and worst. It converts several independent assessments into one senior opinion at several times the cost, and it is invisible when it happens because the conversation feels collaborative. If your panel does exactly one thing differently after reading this, make it the silent ten minutes.
Overlapping questions are second. Three panelists asking about teamwork in slightly different words spend most of the session gathering the same evidence, and the competency nobody was assigned goes untested. This is a planning failure, not an interviewing failure, and it is fixed on paper the day before.
Panels assembled for political reasons are third. A seat given to somebody so they feel included produces a rating from a person with no accountability for the outcome, and that rating gets averaged in with the hiring manager’s. Inclusion in the decision is not the same as a seat in the room, and there are cheaper ways to give it.
Inconsistent panels across candidates are fourth. Swapping a panelist between the second and fourth candidate means the scores are no longer comparable, and the candidate who happened to face the tougher rater looks weaker for a reason that has nothing to do with them.
Untrained panelists are fifth. The peer or internal customer sitting on their first panel usually has no idea what a behavioral question is, what the scale means, or what they cannot ask. That is a thirty-minute problem to solve and a lawsuit-shaped problem to ignore, which is why training everyone who interviews is worth more than any single process change.
Treating the panel as a performance is sixth. A room of people taking turns to be impressive at a candidate is not an assessment, and candidates read it accurately. The best panels are quiet, orderly, and slightly boring, and they produce evidence that is still usable a week later.
Frequently Asked Questions
What is a panel interview?
A panel interview is a single interview in which several interviewers meet one candidate at the same time and each rates the candidate against agreed competencies. The US Office of Personnel Management describes it as an interview in which each panel member individually observes, records, and evaluates the candidate’s responses before the panel compares notes. The format differs from a sequential loop, where the same interviewers meet the candidate one after another in separate sessions. Panels are used to shorten the calendar, to expose several stakeholders to the same answers, and to reduce the influence of any single interviewer’s preferences. None of those benefits are automatic. A panel that has not divided the questions and does not score independently is simply an audience watching one person interview, and it produces a worse decision than that person would have produced alone.
How many people should be on an interview panel?
Three is the working default for a small business, and the Office of Personnel Management guide says a panel of two or three interviewers may be better able to document and interpret what is said. The right size is the number of seats the decision genuinely needs: the hiring manager who owns the outcome, a future peer who will work beside the hire, and the internal customer who receives the work. Add a subject matter expert only when technical depth is hard to verify any other way. Every extra panelist costs scheduling time, shrinks the number of questions each person gets to ask, and creates another stakeholder whose view has to be reconciled. If somebody cannot name the competency they are there to assess, they do not have a seat.
How do you conduct a panel interview?
Assign each panelist a named block of competencies and questions in writing before the day, then run the session with a single chair who opens, introduces the panel, explains the format, keeps time, and hands between blocks. Ask the same opening questions of every candidate for the role in the same order, and let follow-up probes vary where an answer is thin. Panelists take notes on what the candidate said rather than what they concluded, and nobody reacts to a colleague during the session. When the candidate leaves, the room stays silent while each panelist scores every competency alone and writes the supporting evidence under the number. Only then do the scores come out together and the discussion begins, starting with the most junior panelist.
Should panelists score before or after discussing the candidate?
Before, without exception. The Office of Personnel Management guide instructs each panel member to review their notes immediately after the candidate leaves the room and rate the candidate while forming an independent evaluation without discussion with other panel members, and only afterwards to compare notes, explore the basis for discrepancies, and reach a consensus. The reason is that the independent judgments are the whole value of the panel. Once a senior person says the candidate was impressive, the other ratings collapse toward that opinion and you have converted several assessments into one, at several times the cost. Independent scoring also gives you a record. Keeping the individual sheets is what later shows whether one panelist consistently rates a particular kind of candidate lower than the rest of the panel does.
How long should a panel interview last?
Between forty-five and seventy-five minutes for most small business roles, with the length driven by the number of competency blocks rather than the number of panelists. Budget roughly ten to fifteen minutes per block, five minutes for the chair to open and set the format, and ten minutes at the end for the candidate’s own questions and a clear statement of what happens next. Then protect fifteen minutes immediately afterwards for silent scoring and discussion, which is part of the interview and not an optional extra. Sessions that run past ninety minutes rarely produce better information. Fatigue flattens the answers, panelists start improvising questions that nobody agreed, and the scoring conversation gets pushed to a later day when everyone is working from memory instead of notes.
Are panel interviews better than one-on-one interviews?
Not automatically, and the meta-analytic evidence is a useful corrective here. McDaniel, Whetzel, Schmidt and Maurer, publishing in the Journal of Applied Psychology in 1994 across 245 coefficients and 86,311 individuals, found that individual interviews predicted job performance at least as well as board interviews and in their overall comparison somewhat better. What did raise validity for job performance in the same analysis was structure, with structured interviews outperforming unstructured ones. The practical reading is that a panel buys you scheduling efficiency, several stakeholders hearing identical answers, and a written record from more than one rater. It does not buy accuracy on its own. A structured one-on-one beats an unstructured panel, so if you only have time to fix one thing, fix the structure rather than adding people to the room.
What questions should each panelist ask in a panel interview?
Each panelist should own a distinct block tied to the competency they are there to assess, so the candidate is never asked the same thing three ways. The hiring manager takes the core work and ramp time. A future peer takes working style, disagreement, and how the candidate behaves when stuck. The internal customer takes handoffs and cross-team friction. A subject matter expert takes verified technical depth using a real situation from your own business. An owner or senior leader takes motivation and trajectory, asks the fewest questions, and speaks last during scoring. Write the blocks down and circulate them before the day. Overlap is the most common panel failure, and it costs the candidate the airtime that would have produced the evidence you actually needed.