Interview Feedback Examples for Evidence-Based Hiring

Interview feedback becomes useful when another reviewer can answer three questions without asking the interviewer to reconstruct the conversation: What role criterion was assessed? What did the candidate say or do that informed the rating? What remains uncertain?

That sounds simple, yet many feedback records stop at conclusions such as “strong communicator,” “not enough depth,” or “good fit.” Those phrases may capture a reaction, but they do not preserve the interview evidence underneath it. The hiring team then has to debate an adjective instead of reviewing the candidate’s example, scope, action, or result.

The interview feedback examples below use a repeatable pattern: criterion, observation, interpretation, uncertainty, and recommendation. They are written for recruiters, hiring managers, and interviewers who need an internal record that supports a careful human decision.

Start interview feedback with the agreed criteria

Feedback should not introduce a new definition of the role after the interview. Begin with the outcomes and requirements already approved during intake, then assign each interviewer a focused part of that definition. The structured interview questions guide shows how to connect each question to a rating guide before conversations begin.

Current recruiting-system documentation reflects the same sequence. Greenhouse’s scorecard overview, updated May 29, 2026, describes scorecards that use predetermined criteria before interviewers submit feedback and later review completed records as a group. The durable workflow point is not the software field itself. It is that the assessment definition exists before the reviewer writes a recommendation.

Use the same criteria across candidates for one role. Relevant follow-up questions can differ when a candidate gives an incomplete answer, but the underlying standard should remain stable. If an interview reveals that the role definition is wrong or incomplete, record that as a workflow issue and decide how the change will apply to every active candidate.

Use a five-part interview feedback record

A compact feedback record can be detailed without becoming a transcript. Use these five fields for each assigned criterion:

1. Criterion: the approved requirement or outcome being assessed.
2. Observation: the candidate example, action, scope, or result heard in the interview.
3. Interpretation: why that observation supports, partly supports, or does not establish the criterion.
4. Uncertainty: the missing detail that could materially change the assessment.
5. Recommendation: the interviewer’s human-owned next-step view, with a reason tied to the criterion.

Keep the observation close to the source. An Indeed guide updated June 16, 2026 recommends a consistent feedback format and says interviewers should support statements with interview examples or resume details. That source boundary matters: the note should show what the candidate provided and what the interviewer inferred from it.

A candidate scorecard can hold the criteria and ratings. The feedback record adds the evidence and uncertainty that make those ratings reviewable.

Interview feedback examples: vague versus evidence-ready

These examples are deliberately short. Each “evidence-ready” version gives the next reviewer something specific to inspect without pretending that one answer proves more than it does.

Example 1: project ownership

Vague: “Strong project leader.”

Evidence-ready: “For the project-ownership criterion, the candidate described leading a six-person rollout across support and operations. They set the weekly review, maintained the risk log, and escalated a delayed dependency to the sponsor. The example supports coordination ownership; it did not establish responsibility for the final implementation plan. Recommendation: supported for this stage, with implementation ownership still open.”

Example 2: stakeholder communication

Vague: “Communication was not senior enough.”

Evidence-ready: “For the executive-communication criterion, the candidate explained how they summarized customer findings for product and sales leaders, including the tradeoff they recommended. They did not describe presenting the decision to an executive sponsor. The answer supports cross-functional communication but leaves executive-level scope unresolved. Recommendation: ask one focused follow-up if that scope is required at entry.”

Example 3: missing evidence

Vague: “No experience with process improvement.”

Evidence-ready: “The interviewer asked for a process-improvement example. The candidate described following an existing intake workflow but did not identify a change they designed, the reason for it, or a measured result. The interview did not establish process-improvement ownership. This is an evidence gap in this conversation, not a claim that the candidate has never done the work.”

Example 4: a strong answer with a scope limit

Vague: “Excellent analytical skills.”

Evidence-ready: “For the analytical-decision criterion, the candidate compared three support-volume segments, identified that one channel drove most repeat contacts, and changed the triage rule. They reported that repeat contacts fell in the following review period. The example supports operational analysis within one team; the interview did not test work across multiple business units.”

Example 5: conflicting evidence

Vague: “Mixed signals on team leadership.”

Evidence-ready: “The resume states team leadership, while the interview example described coordinating peers without direct responsibility for goals or performance review. The candidate may have led delivery without managing the team. Record the difference rather than choosing one label. Recommendation: clarify the relationship, team size, and decision authority before changing the rating.”

Separate observation, interpretation, and recommendation

The easiest way to improve feedback is to stop those three layers from collapsing into one sentence. “The candidate paused before answering” is an observation. “The candidate lacks confidence” is an interpretation that the pause alone does not establish. “Do not advance” is a recommendation that still needs a role-related reason.

Ask interviewers to label each layer. If a conclusion cannot be connected to a candidate example or another approved source, move it into uncertainty or remove it from the decision record. A precise unknown is more useful than a confident statement without support.

When notes are incomplete, an interview transcript analysis workflow can help locate the original exchange. The transcript is a reference, not a rating. The interviewer still has to verify context and decide what the answer supports.

Keep internal feedback distinct from candidate communication

Internal feedback and a candidate-facing message are related records with different purposes. The internal note can include the criterion, evidence, uncertainty, and recommendation needed by the hiring team. A candidate message should communicate only the confirmed process outcome and any feedback the team has deliberately approved for that audience.

Do not copy an unverified internal summary directly into a message. First confirm that the decision owner has recorded the outcome. Then check that every explanatory statement is supported by the interview record and reflects the criterion used for this role. If the team has not resolved conflicting notes, the message should not invent a single confident reason.

This separation also protects the usefulness of the internal record. Interviewers can document an open question honestly without implying that the candidate has already received or accepted that interpretation.

Write feedback independently before the debrief

Complete the feedback before reading other interviewers’ recommendations. The U.S. Office of Personnel Management’s Structured Interview Guide instructs panel members to record and evaluate responses individually before discussing their ratings and supporting observations. Independent first-pass records give the group real differences to examine rather than a story assembled after one opinion has already framed the candidate.

The later interview debrief has a separate job: compare the completed records, identify whether ratings conflict or cover different criteria, and document the human-owned next step. Feedback is the input. The debrief is the review.

Use AI to organize feedback, never to author the verdict

AI can sort verified notes under the right criteria, flag a rating with no evidence field, identify wording that mixes observation with inference, and prepare a provisional summary for review. It can also help compare whether each interviewer completed the same required fields.

The interviewer remains responsible for the record. Compare every generated statement with the source notes or transcript, correct attribution, restore scope, and remove details the candidate did not provide. AI should not infer personal traits from delivery, turn silence into a negative finding, resolve conflicting accounts, or choose whether a candidate advances.

Interview feedback review checklist

• Does every rating name the approved role criterion?
• Is the supporting candidate example or observable detail recorded?
• Are observation and interviewer interpretation separate?
• Does the note state scope without stretching the example?
• Is missing information described as uncertainty rather than failure?
• Are resume and interview conflicts preserved for review?
• Was the feedback completed before the group discussion?
• Does any follow-up name the exact question it must resolve?
• Is the recommendation tied to evidence instead of a general impression?
• Has a person verified every AI-assisted summary?
• Does a named human own the next decision?

Make interview feedback useful to the next reviewer

Good interview feedback does not need to sound decisive in every line. It needs to preserve what was assessed, what the candidate demonstrated, what remains unknown, and why the interviewer recommends a next step. Write those pieces independently, keep the source boundary visible, and let the hiring team make the final decision from a record it can actually inspect.