Interview debrief template: agenda, speaking order and decision record

An interview debrief is the meeting where the people who interviewed a candidate compare what they saw and reach a hiring decision. A good one takes about 30 minutes, starts only after everyone has submitted an independent scorecard, walks through the evidence competency by competency, hears recommendations with the most senior voice last, and ends with a written decision record.

Below is a timed agenda, the rules for speaking order, how to handle disagreement, a decision record template, a filled example, and the phrases a facilitator actually uses in the room.

Before the debrief: independent scorecards

Most debriefs go wrong before they start. If interviewers have already swapped impressions in the hallway or read each other's feedback, the meeting confirms whatever the first opinion was. The U.S. Office of Personnel Management's structured interview guide describes the same sequence for panels: each member rates independently without discussion, and only then do they compare notes, explore the reasons for differences, and reach a consensus.

• Scorecards are due before the meeting, with a set deadline, such as two hours before. The facilitator checks and chases.
• No discussing the candidate between interviews. Say it in the loop briefing, not after it has happened.
• Hide feedback until you submit. If your ATS can keep others' scorecards hidden until an interviewer submits, turn it on.
• The facilitator prepares one page: a grid of competencies by interviewer, spreads of two or more points highlighted, the status of each must-have, and anything not assessed.
• Hold it within a day or two of the last interview, while details are fresh and before the candidate accepts something else.

The grid can be as plain as this:

Competency                  Must-have  Ana  Ben  Chris  Dee   Spread
Stakeholder management      yes        3    3    -      -     0
Prioritization              yes        4    -    2      -     2  (discuss)
Data analysis               no         -    3    -      3     0
Written communication       no         -    -    3      2     1
Not assessed: none
Recommendations: collected at the end of the meeting, not shown here

The scorecards themselves come from your interview scorecard template. If they arrive as adjectives with no evidence, the debrief has nothing to work with; the fix for that is evidence-based interview feedback.

The 30-minute debrief agenda

DEBRIEF: [Candidate] for [Role]                       30 minutes
Attendees: [interviewers], [hiring manager], [facilitator]

0:00  Facilitator frames it (2 min)
      - The decision we are making today
      - The must-haves and the rating scale, read aloud
      - Rules: evidence first, recommendations last, hiring manager last

0:02  Silent read of the ratings grid (3 min)

0:05  Walk each competency (15 min)
      - The interviewer who owned it reads their evidence and rating
      - Others add evidence only, no verdicts yet
      - Any spread of 2+ points: both read their evidence, then resolve
      - A rating changes only if someone names new evidence

0:20  Must-haves and risks (5 min)
      - Each must-have: met / not met / not assessed
      - Risks we would be accepting
      - What evidence, if any, would change the decision

0:25  Recommendations (3 min)
      - Round the table, most junior first, hiring manager last
      - One sentence each, pointing to the ratings

0:28  Decision and record (2 min)
      - Decision, who decided, next step, owner, date
      - Facilitator writes the decision record before leaving

Walk the meeting by competency, not by interviewer. Going person by person turns the debrief into a sequence of speeches that each end in a verdict. Going competency by competency keeps the evidence side by side, so a 4 and a 2 on the same thing get looked at together.

Speaking order

Once someone senior says "I'd hire them", it becomes hard for a newer interviewer to say they saw a problem. The order of speaking is how you protect the independent views you collected.

1. Evidence before opinions. Nobody gives an overall recommendation until every competency has been walked.
2. The owner of each competency speaks first on it, because they asked the questions.
3. Recommendations go from most junior to most senior, with the hiring manager last.
4. The facilitator speaks last of all, if they interviewed, and otherwise does not give a view.
5. On video calls, reveal together. Ask everyone to type their recommendation in the chat and send it on a count, so nobody adjusts to the first message.

Set these rules at the start of the meeting, every time. Enforcing a rule mid-meeting feels like correcting a person; reading it out beforehand feels like process.

Handling disagreement

A spread of two points or more

Ask both interviewers to read the evidence behind their ratings. There are usually three causes: they heard different parts of the story, they read the anchor differently, or one of them was rating something else, such as a strong overall impression. The first is resolved by combining the evidence. The second is a calibration issue to fix after the meeting. The third is resolved by going back to the anchor.

Evidence against impression

When one interviewer has observations and the other has a feeling, the observations set the rating. The feeling is not thrown away: turn it into a specific question that a follow-up conversation or a reference could answer.

Changing a rating

A rating changes only when someone names evidence the interviewer did not have, not because the room leaned on them. Leave the original scorecard as submitted and record the change, with the new evidence, in the decision record.

One strong no

Treat it as information, not an automatic veto and not an outlier to average away. Ask which competency it concerns and what the evidence was. A strong no on a must-have, backed by evidence, is either a no or a specific follow-up. A strong no on a nice-to-have is weighed like any other rating.

Not enough evidence to decide

"Let's think about it" is not a decision. Name the missing evidence, how to get it (a follow-up conversation, a work sample, a reference), who owns it, and the date the team decides. When to add a step, and how to compare finalists once you have it, is covered in how to compare candidates after interviews.

The hiring manager and the panel disagree

The hiring manager usually decides. What makes it defensible is the written reason: which evidence outweighed which, and which risk they are accepting. A decision record that says "HM override" and nothing else helps nobody later.

The decision record template

Write it at the end of the meeting, before anyone leaves. It takes two minutes and it is the only artifact of the debrief that anyone will read in six months.

DEBRIEF DECISION RECORD
Candidate: [name]                Role / req: [title, ID]
Debrief date: [date]             Facilitator: [name]
Interviewers present: [names]    Absent (scorecard used): [names]
All scorecards submitted before the meeting: [yes / no: who was late]

DECISION: [Advance to offer / Reject / More information needed]
Decided by: [name, role]
If more information: [what evidence, how, owner, decide by date]

Must-haves
- [Must-have 1]: [met / not met / not assessed]. [one line of evidence]
- [Must-have 2]: [met / not met / not assessed]. [one line of evidence]

Ratings (by interviewer, then the team's view)
- [Competency]: [A: 3, B: 4], team view [3]
- [Competency]: [C: 2], team view [2]

Main reasons (job-related, pointing to evidence)
1. [reason]
2. [reason]

Risks accepted: [risk, and any plan to manage it]
Ratings changed in the meeting: [competency, old rating, new rating, the new evidence]
Dissent: [who, on which competency, their evidence]
Next step: [action, owner, date]
Stored at: [ATS record or file location]

Keep it with the scorecards. Under 29 CFR 1602.14, employers covered by Title VII, the ADA and GINA must generally keep hiring records for one year from when the record was made or the decision taken, whichever is later, and longer once a discrimination charge is filed. That is a summary, not legal advice; confirm your own retention rules with HR or counsel.

A filled decision record

Invented candidate and team, showing a rating change and a recorded dissent:

What the facilitator says

The "not job-related" line is worth rehearsing, because it comes up: family, age, where someone lives. What does not belong in the conversation also does not belong in the notes; see what not to write in interview notes.

Variations: async debriefs, several finalists, small teams

• Async debrief. For teams across time zones: scorecards submitted, the facilitator posts the grid and flags spreads in one thread, interviewers add evidence (not verdicts) within a set window, and a short call happens only if a must-have is disputed. The hiring manager posts the decision and the facilitator writes the record.
• Several finalists. Debrief each candidate against the anchors first, in separate agenda items. Compare finalists afterwards, as its own discussion. Comparing candidates in the middle of a debrief pulls ratings toward whoever was interviewed most recently.
• Two interviewers. Keep the independent scorecards and the decision record even when the whole debrief is a 15-minute call. Small teams skip process first and regret it when a decision is questioned. A lightweight setup is in structured interviews without an ATS.
• Onsite panels. Agree who owns which competency before the day so the grid has no gaps and no duplicates; the day itself is covered in panel interview tips for interviewers. Interview Signal syncs which questions have been asked across a panel, without sharing anyone's notes or transcript, so a missed must-have shows up before the debrief rather than during it.