Running an interview debrief that is not just a vote

The standard debrief: five people in a room, someone asks "so, thoughts?", the most senior person answers first, and the meeting is effectively over. Everything after that is people locating reasons for a conclusion that has already formed.

This is not a culture problem. It is a predictable consequence of the format — first speaker and highest status both anchor a group, and no amount of good intent overrides that. You have to change the format.

Independent input before any discussion

Non-negotiable, and it is most of the benefit. Every interviewer submits their assessment before the debrief and before seeing anyone else's. If someone has not submitted, the debrief does not start.

The moment an interviewer knows the hiring manager was impressed, their own memory of the interview begins to reorganise around that. Independence is not about fairness in the abstract; it is about preserving information that will otherwise be destroyed in the first ninety seconds of the meeting.

Open on disagreement, not verdict

Do not start with a round of overall recommendations. Start with the single competency where the submitted scores are furthest apart, and ask the two people who disagree to describe what they heard.

Almost always, one of three things is true, and all three are worth discovering:

• They asked different questions and saw genuinely different behaviour — the candidate is inconsistent, which is itself the finding.
• They heard the same answer and applied different standards — your rubric is underspecified, and this will keep happening until you fix it.
• One of them was not really listening for that competency — a coverage problem in your interview loop.

Evidence, not adjectives

Enforce one rule hard: claims about a candidate must be supported by something the candidate said or did. "He seemed junior" is not admissible. "When I asked who disagreed with the migration, he could not name anyone, and he had described it as his decision" is.

This rule does most of the work in a debrief, because the vast majority of bias enters through unsupported impressions. "Culture fit", "not hungry enough", "seemed nervous" — none of these survive a requirement to cite the moment.

It is also why verbatim capture matters. Debriefs collapse into adjectives largely because nobody can remember what was actually said three days and four interviews later.

Decide against the bar, not against the pool

"Best of the three we saw" is not a hiring standard, it is a ranking. If the bar for the role is written down per competency, the question is whether this person clears it — and the honest answer is sometimes that the strongest candidate you interviewed still does not.

Teams that skip this step hire to fill the slot and then spend a year managing the consequence.

Write down why

Record the reasoning, not the outcome. Two or three sentences: which competencies cleared the bar, which did not, what trade-off you accepted, and what you will watch for in onboarding if you hired.

This does three things:

1. It lets you calibrate later — six months in, you can compare the reasoning against how the hire actually performed. This is the only real feedback loop hiring has.
2. It protects you if a rejected candidate challenges the decision.
3. It makes the next debrief faster, because the standard is written rather than re-argued.

The practical obstacle

All of this depends on interviewers arriving with accurate, specific, evidence-backed assessments. In reality they arrive with a memory of a conversation from Tuesday, half-notes typed while trying to maintain eye contact, and a general feeling.

That is the gap worth closing with tooling rather than exhortation. When the transcript, the competency scores and the candidate's actual words are captured during the interview, the debrief starts from a shared record instead of five divergent memories — and the argument can be about the evidence rather than about who remembers it best.