Most interview feedback says something like “strong communicator, good culture add” and stops there. A hiring manager can’t defend a decision on that, and a candidate who hears “someone else was a better fit” can’t do anything with it.

The fix I’d make is at the moment of writing: every line of interview feedback starts from something the candidate said and ends on a line of your rubric.

Feedback written any other way comes from memory, usually a day late, and the debrief turns into an argument about who remembers the candidate best. The candidate pays for that as well, with a rejection that teaches them nothing.

This guide covers five habits that keep every line tied to the candidate’s words across a whole interview panel, and how to check that they’re working.

Why interview feedback goes vague.

Feedback usually goes vague for one of three reasons, and all three come from how and when it gets written.

Most interviewers go straight from the interview into their next meeting. By the time they open the scorecard, the exact words are gone, and what gets written down is an impression of the candidate.

Two interviewers who never agreed what a three on “communication” looks like will rate the same kind of answer a four and a two, and neither rating explains itself.

Internal notes also get written for the debrief, so when the recruiter has to tell the candidate something, there’s almost nothing in them a candidate could use. That’s how “someone else was a better fit” becomes the standard rejection: it’s the one sentence the notes can support.

Five habits that make interview feedback usable.

Each habit below fixes one of those reasons, and they run in the order the work does, from the panel’s first meeting to the message the candidate reads.

Agree what each rating means before the first interview.

Settle what a rating means before any ratings are in, because afterward every disagreement becomes an argument about the candidate.

Run a 30-minute session with the panel for the role. Take two sample answers for each skill on the scorecard, one strong and one weak, and agree as a group what rating each one earns and why. Then write that agreement into the rubric, so it outlasts the meeting and reaches the interviewer who joins the panel next month.

If the rubric itself needs rebuilding first, this post on interview rubrics walks through how to write one.

Write each observation from what the candidate said.

Start interview feedback from the candidate’s own account: the phrase they used, or the trade-off they walked through. No one else in the debrief can check your impression of them, so put it after the evidence.

“I scoped the project at three weeks, but my manager wanted two” is something a hiring manager can weigh and a candidate can recognize. “Good problem-solving skills” is a label, and the candidate already knew they were being labeled.

Before the panel starts, give every interviewer one line per competency saying what they’re listening for. After the interview, write each observation as what the candidate said, then what you made of it.

Tie every observation to a line of the scorecard.

An observation that doesn’t map to a competency on the scorecard reads as opinion, even when it’s right. Mapped to one, it becomes a reason the team can use.

Compare “she was great” with “she showed four of the five behaviors the rubric sets for the senior level.” The second is a decision the team can defend in the debrief and explain to the candidate afterward.

Add an evidence column to your scorecard template, and ask for one sentence of evidence beside every rating before the debrief opens. If an interviewer can’t fill the column, the rating is probably higher than the evidence supports.

Put the writing time on the calendar.

The longer the gap between the interview and the write-up, the more of the write-up comes from memory. So the write-up needs a slot of its own, booked before anyone’s day fills.

Book half an hour on each interviewer’s calendar straight after the interview, at the same time you book the interview itself. I’d expect a slot that’s already on the calendar to survive the week better than one an interviewer has to find at the end of a day of calls.

Then hold the debrief until every panelist has submitted their scorecard, so no one rewrites their notes after hearing the room.

Give the candidate the short version within 48 hours.

The candidate gets a shorter version of the internal notes, written from the same evidence and for a different reader.

Give them one to three strengths and one to three development areas, strengths first. The cap makes the interviewer pick the observations that carried the decision, where a full list would bury them under everything else they noticed.

A rejected candidate still gets both. “The role needed more experience designing systems at the staff level, and most of your examples were about delivering them” tells a candidate what to build next, and the strengths tell them what to keep leaning on.

Split the writing between two people. The interviewer writes the substance, because they hold the evidence, and the recruiter edits it for tone and sends it.

Send it within 48 hours of the interview. If the panel hasn’t decided by then, send a short note saying when the candidate will hear, and keep that date.

For the wording of the message itself, writing rejection emails has examples you can adapt, including one that carries feedback.

Where Metaview fits in interview feedback.

Of the five habits, I’d expect writing from the candidate’s own words to break first at volume, because it depends on the exact words surviving until someone writes them down. Metaview keeps the words, and the judgment stays with your team.

Capture the interview with the Notetaker.

Metaview’s Notetaker joins the interview as a visible participant, then records and transcribes it into structured notes. It captures every spoken word, so the evidence column on your scorecard starts from the candidate’s own answers, and so does the scorecard draft.

Metaview drafts the scorecard from the conversation, and the interviewer sets the ratings and submits it.

1

2

3

1. 1The summary card opens the notes with a recap of the conversation.
2. 2Each question the interviewer asked becomes a heading.
3. 3The candidate’s answers sit under it as short notes, ready to turn into evidence on the scorecard.

Quora’s recruiting and hiring teams work across the globe, and they brought in Metaview’s notes when feedback was slow to travel from hiring managers to recruiters and on to candidates. Collaboration and calibration between Quora’s recruiters and hiring managers have improved since then, especially with the team working remotely. If your feedback stalls between the hiring manager and the recruiter, that handoff is the place to start.

Read the whole loop in one set of notes.

Multi-source notes combine the conversations and documents you add to them, such as the loop’s interviews and the job description, into one synthesized notes document. Add every interview in the loop, and the interviewer writing the candidate’s short version can read those conversations in one place.

This 53-second clip from Metaview introduces multi-source notes and the kinds of material you can add to them, from the intake call to earlier interviews.

Interviewers often walk into the room thinking they’re making the hiring decision. Reading the loop in one document is a reminder that each of them covered one part of the candidate.

Check coverage in Reports before anyone writes.

Metaview Reports shows what your team covered across the interviews it has captured. In the view below, each row is one candidate’s interview and each column a competency, counted in the questions asked about it.

A competency with no questions against it has no evidence behind any rating it gets. Find those before the debrief, while there’s still a later round to cover them in.

1

2

3

1. 1The filter picks the interviews to count, here backend engineering interviews from the last month.
2. 2Each column is a competency, and each cell counts the questions asked about it in that interview.
3. 3A “0 questions” cell means that interview never covered the competency.

When a debrief turns on a claim about a candidate, use Reports to filter the interviews by that topic and read each candidate’s answer before anyone writes it into the feedback.

1

2

3

1. 1The filter sets the topic and the interviews to search.
2. 2Each row is a candidate and the interview stage the answer came from.
3. 3The answer column sums up what each candidate said on the topic.

How to tell whether your interview feedback is working.

Run these checks on the last month of interviews for one role, and fix the habit each one names when it comes back badly.

• Evidence behind the ratings: open five recent scorecards and find what the candidate said behind each rating. Where you can’t, start with writing from the candidate’s words.
• Empty evidence columns: count the ratings with no evidence sentence beside them. A high count means the column is optional in practice, so make it required before the debrief opens.
• Spread between interviewers: compare two interviewers’ ratings on the same competency for the same candidate. A gap of two points or more means the definitions need another session with sample answers.
• Days to the candidate: count the days from the interview to the feedback the candidate received. Past two days, book the writing time on the calendar before you change anything else.
• The “better fit” count: read the last 10 rejections your team sent and count the ones whose only reason is fit. Trace each one back to the scorecard behind it, and fix the evidence column there.

I’d start with the evidence column, because a rating with a sentence of the candidate’s own words beside it is feedback a hiring team can defend in the debrief and a candidate can act on afterward.