How to run an interview debrief

The agenda

1. Confirm every scorecard is submitted. If one is missing, reschedule rather than proceed.
2. Each interviewer states their recommendation in one word. No discussion, no reasons yet.
3. Start with the dissenting view, not the majority.
4. Work attribute by attribute, citing evidence rather than impressions.
5. Name what nobody assessed.
6. Decide: hire, do not hire, or one specific additional assessment with a stated question.

Collect scorecards before the room opens

This is the mechanism the whole debrief rests on. If scorecards arrive during or after the discussion, they record the discussion rather than the interview. Make submission a precondition for the meeting, and be willing to reschedule — once — when someone has not submitted. It happens once per team.

Open with recommendations, not reasons

Go around the room and have each person say only “strong yes”, “yes”, “no” or “strong no”. No justification yet. This gives you the shape of the disagreement before anyone has had a chance to influence it, and it prevents the most senior or most confident voice from framing the conversation. If everyone says the same word, the debrief can be short — and often should be.

Start with the dissent

When opinions differ, ask the dissenting interviewer to speak first. Three reasons: the information that changes a decision is almost always held by the minority, dissent voiced after a majority consensus takes far more courage than most people have in a meeting, and the dissenter is often the person who asked the question nobody else asked. If the dissent does not survive the evidence, it will not survive going first either.

Argue about evidence, not impressions

• “What did they say that led you there?” is the only question that reliably moves a debrief forward. Ask it of agreement as well as disagreement.
• Distinguish ‘I did not see it’ from ‘I saw the opposite.’ These feel identical in a meeting and mean entirely different things. The first is a coverage gap; the second is a genuine conflict.
• Watch for the candidate's best moment doing all the work. One excellent answer is evidence about one attribute, not about the person.
• Watch for seniority collapsing the room. If the hiring manager speaks first and last, you have one opinion with witnesses.

Name what nobody assessed

Before deciding, ask explicitly: which of our attributes did nobody actually observe? Loops routinely assess the same two things four times and never test a third. If a required attribute has no evidence behind it, that is the only legitimate reason to add a round — and the added round should have a single stated question, not be a general second opinion.

Decide in the room

• Hire — move to offer, and agree who closes the candidate and by when.
• Do not hire — agree who sends the rejection and whether feedback is offered. Within a few days, not weeks.
• One more assessment — only with a specific question written down, a named interviewer and a date. “Another conversation to be sure” is how loops reach seven rounds and lose the candidate anyway.

How long a debrief should take

Twenty to thirty minutes for a standard loop. Unanimous outcomes should take ten. If a debrief routinely runs an hour, the cause is almost always that scorecards were not submitted beforehand, so the meeting is being used to do the assessment rather than to reconcile it. Book thirty minutes and end early rather than booking an hour and filling it — long debriefs drift toward consensus, and consensus is not the same thing as agreement.

A debrief that keeps ending in uncertainty is a design problem

If your debriefs routinely cannot reach a decision, the issue is upstream: the loop is assessing the wrong things, the scorecard attributes are too vague to disagree about precisely, or nobody owns the requirement that keeps going untested. Fix the loop rather than adding rounds. See how to design an interview loop and how many interview rounds is too many.

Where Interviewer Mode fits

Disclosure: we make CoPilot Interview, which includes Interviewer Mode for the hiring side and a real-time assistant used by candidates. We have an interest in you using the first, and our view of the second is that interviews should be designed to measure something an assistant cannot fake. Weigh the advice below accordingly.

Frequently asked questions

How do you run an interview debrief?

Collect every scorecard before the meeting, have each interviewer state a one-word recommendation with no discussion, then start the discussion with whoever disagrees with the majority. Work attribute by attribute citing evidence, name anything nobody assessed, and end with a decision rather than an extra round.

Who should speak first in a hiring debrief?

Whoever disagrees with the majority. The information that changes a decision is usually held by the minority, and dissent is far harder to voice once a consensus has formed. The hiring manager should speak last, because seniority anchors the room.

Should interviewers submit scores before the debrief?

Yes, and it should be a precondition for the meeting. Scorecards written during or after the discussion record the discussion rather than the interview, which removes the independence that made multiple interviewers worth having.

What if the interview panel cannot agree?

Work out whether the disagreement is “I did not see it” or “I saw the opposite.” The first is a coverage gap and may justify one more assessment with a specific written question. The second is a real conflict about evidence and should be resolved by the hiring manager, who owns the decision.

Is it okay to add another interview round after a debrief?

Only when a required attribute genuinely went unassessed, and only with one stated question, a named interviewer and a date. Adding a general second-opinion round is how loops drift to six or seven stages and lose strong candidates to faster processes.