Interview Feedback Template: Write Notes That Hold Up in a Debrief
Interview Feedback Template: Write Notes That Hold Up in a Debrief
Most interview feedback gets written to justify a decision the interviewer already made in the first ten minutes. The score comes first, the reasons get reverse-engineered to fit. That feedback reads fine alone and falls apart the moment a colleague asks, "what did the candidate actually say?" The fix is not a better adjective. It's the debrief test: feedback only counts if it survives challenge from someone who wasn't in the room. This piece gives you an interview feedback template you can paste into your notes, plus three rules that make every note hold up when the panel reconciles scores.
The debrief test is a simple standard. For each thing you wrote, a teammate should be able to point to a quoted moment, tie it to a stated criterion, and tell your observation apart from your opinion. Feedback that fails that check gets overruled by whoever talked last. That gap, between notes that hold and notes that wash out, is what the rest of this closes.
Why Most Interview Feedback Fails the Debrief Test
Interview feedback fails because it records a verdict, not evidence, and a verdict has nothing to defend itself with. "Strong communicator, great energy, would hire" is a conclusion wearing the costume of a note. The debrief test asks the one question that conclusion can't answer: based on what? When the panel meets and two people disagree, the interviewer with quotes wins and the interviewer with adjectives folds. So the adjective-writer learns to fold faster next time, and the debrief drifts toward whoever sounds most certain rather than whoever has the receipts.
That drift has a name your panel already knows. The halo effect lets one good moment inflate every unrelated score on the card, and the debrief test catches it only if your notes separate that moment from the rest. Recency bias does the same trick with time, tilting the room toward the last candidate discussed. Both feed on memory, and memory is a terrible witness. Short-term memory holds raw detail for roughly 20 seconds, so the exact phrasing that would have settled an argument is gone before you've finished nodding. By the debrief, you're not recalling what the candidate said. You're recalling what felt important, which is a far less reliable record than the transcript.
The research backs the cost of getting this wrong. Schmidt and Hunter's 1998 meta-analysis put the predictive validity of structured interviews at 0.51 against 0.38 for unstructured ones, and later meta-analytic work by Wiesner and Cronshaw found structured interviews roughly twice as valid (Test Partnership summary). Validity lives in the evidence, not the verdict, and the evidence is exactly what unstructured notes throw away. Hire on what they said, not how you felt, and you inherit that validity. Skip the evidence and you inherit the coin flip. The structure that recovers it starts with one artifact: a template that forces a quote and a criterion onto the page before you're allowed to write a score.
The Interview Feedback Template
A good interview feedback template does one job: it makes you capture evidence during the interview, when the words are still in the room, instead of reconstructing a story afterward. Score the signal, not the vibe, and the template is what holds the signal still long enough to score it. Google's re:Work structured interviewing framework names recording comprehensive feedback of candidate answers as one of its four pillars, alongside vetted questions, standardized rubrics, and interviewer calibration (Google re:Work). The template below is the recording pillar, made small enough to use live.
Paste this into your notes doc before the call. One block per criterion, three or four criteria per interview, no more:
## Candidate: [name] | Role: [role] | Date: [date]
### Criterion 1: [e.g. Debugging under ambiguity]
- Quote: "[exact words the candidate used]"
- Observation: [what they did, plainly stated]
- Interpretation: [what you think it means for the role]
- Score: [1 poor / 2 borderline / 3 solid / 4 outstanding]
### Criterion 2: [e.g. Stakeholder communication]
- Quote: "..."
- Observation: ...
- Interpretation: ...
- Score: [1-4]
### Criterion 3: [...]
- Quote: "..."
- Observation: ...
- Interpretation: ...
- Score: [1-4]
### Overall recommendation: [hire / no hire / lean] Reason: [one criterion that decided it]
The four-point scale is deliberate. Google documents what a poor, borderline, solid, and outstanding answer covers for each attribute so all reviewers share the same scale, which kills the "my 7 versus your 7" problem that wrecks debriefs run on a 1-to-10 spread (Google re:Work). Four levels force a side. There's no safe middle to park an undecided candidate in, so borderline means borderline and you have to say why. The scale gives the debrief a shared ruler, and the three rules below tell you what to write next to it.
Three Rules That Make a Note Hold Up
Three rules turn a template from a form into a defense. Each rule answers one way feedback dies in the debrief, and each maps to a line you already have on the card.
- Tie every note to a criterion. A note floating free of a criterion is an opinion looking for a home. Before you write anything, name which of your three or four criteria it serves. If it serves none, it's noise, the kind of off-axis impression that lets bias in through the side door under cover of a vague label. Tying notes to criteria is also what keeps the debrief test answerable: a teammate can check whether your evidence matches the thing you claimed to be measuring.
- Quote the candidate. Write their words, in quotation marks, not your summary of their words. "She said the rollback failed because nobody owned the on-call rotation" beats "good incident response." The quote is the part of your note that survives the 20-second memory cliff, and it's the only line a skeptical colleague can independently weigh. Quote first, paraphrase never, because the paraphrase is already an interpretation pretending to be a fact.
- Separate observation from interpretation. Keep what happened on a different line from what you think it means. "Candidate paused 15 seconds, then walked through three failure modes" is an observation. "Strong under pressure" is an interpretation. When you split them, the debrief can accept your observation and challenge your interpretation without the whole note collapsing, which is exactly the kind of stress the debrief test applies.
These three rules feed one mechanic the panel runs at the end. Each interviewer scores independently first, then the group reconciles discrepancies against the evidence rather than the loudest voice in the room (Metaview). That reconciliation is its own structured meeting, and running it from an interview debrief template is what keeps the evidence-first notes from being overwritten when the panel convenes. Independent-first scoring only works if there's evidence to reconcile, and the rules are what put it on the page. The contrast between a card built this way and the usual freeform notes is worth seeing side by side.
Evidence-First Notes vs Verdict-First Notes
| Dimension | Verdict-first notes | Evidence-first notes |
|---|---|---|
| What gets written | A score and a feeling | A quote, an observation, then a score |
| Survives the debrief test | No, nothing to point to | Yes, every score has a cited moment |
| Vulnerable to halo and recency | Highly, memory fills the gaps | Less, the record outlasts the impression |
| Useful weeks later | Fades with memory | Reads the same on day 30 as day 1 |
| Reconciles in a panel | Loudest voice wins | Evidence wins |
The table is not a style preference. It's the difference between feedback that compounds into a defensible hiring decision and feedback that evaporates between the interview and the debrief. Evidence-first notes also carry a fairness dividend: structured interviews increase predictive validity and decrease differences between demographic groups, because the same criteria and the same scale apply to everyone (Google re:Work). For more on how structured evaluation counters the specific biases that creep into freeform interviews, see Hiring Bias: Types, Causes, and How Structured Interviews Reduce It. The catch with evidence-first notes is the one every interviewer hits in practice: capturing a verbatim quote while also listening well enough to ask the next sharp question.
What We See When the Note-Taking Gets Automated
We built Asked to join the call and transcribe live precisely because the quote-capture rule is the one humans drop first under load. In our own use across interview panels, the pattern is consistent: when interviewers stop splitting attention between writing and listening, their follow-up questions get sharper and their scorecards stop arriving days late with the details already faded. That observation is qualitative, not a benchmark, but it tracks the research. A study on note-taking in the employment interview found that taking notes increased recall accuracy and that reviewing those notes increased judgment accuracy (Middendorf and Macan). The mechanism is the same one the debrief test depends on: a record beats a memory.
The harder problem is that good note-taking competes with good interviewing for the same attention. Laszlo Bock, who ran People Operations at Google and built much of its structured hiring, put the underlying standard plainly: "The goal of our interview process is to predict how candidates will perform once they join the team." Prediction needs evidence, and evidence needs to be captured without taxing the very attention that makes the interview good. That's the tension a live transcript resolves, and it's why the last step is operational, not philosophical: pick one role and run the template on a real candidate this week.
FAQ
How many criteria should an interview feedback template have?
Three or four per interview. Fewer and you miss real signal; more and each one gets shallow notes because attention runs out. Assign each interviewer on the panel a different couple of criteria so the whole rubric gets deep coverage instead of every interviewer skimming all of it.
What's the difference between an observation and an interpretation in interview feedback?
An observation is what happened: the candidate's words, actions, and timing, stated plainly. An interpretation is what you think it means for the role. "Paused, then named three failure modes" is observation. "Handles ambiguity well" is interpretation. The debrief can challenge an interpretation without throwing out the observation underneath it, which is why you keep them on separate lines.
Can I write interview feedback after the interview instead of during?
You can, but it fails the debrief test more often. Short-term memory holds detail for about 20 seconds, so writing hours later means recording impressions, not evidence. If you must write after, do it in the first five minutes while quotes are fresh, and flag any score you couldn't tie to a remembered specific.
Does a feedback template slow the interview down?
Capturing a quote and a one-line observation per criterion costs seconds, not minutes, and it removes the after-hours scorecard entirely. The time you appear to lose live, you get back by not reconstructing the interview from memory later, and the panel gets evidence it can actually reconcile.
Do This Next
Pick one role you are actively hiring for and write down its three most important criteria before your next interview. Build the markdown template above into your notes doc, one block per criterion, with a quote line you commit to filling. Use the four-point scale on your next two candidates and tie every score to a quoted moment. Start today: try Asked free, let it transcribe the call live, and let it draft the evidence-first scorecard from the transcript so you can spend the interview listening instead of scrambling to write.