Interview Evaluation Templates: Free Forms, Sheets and a Scoring Matrix You Can Copy Today

TL;DR

• An interview evaluation template is a form each interviewer fills in on their own. It scores a candidate’s answers against the same job-related criteria, on a defined scale, so you can compare candidates on evidence rather than impressions.
• This page has six free templates, and there’s nothing to download. You get a full interview evaluation form, a one-page evaluation sheet, a panel debrief record and a weighted interview evaluation matrix with working spreadsheet formulas. There are also swap-in rows for four role types and a bank of example evidence notes. Every template copies and pastes straight into Word, Google Docs, Excel or your applicant tracking system (ATS).
• Three rules do most of the work. Score every answer on a scale where each point has a written description. Write down evidence, not impressions. Submit your scores before the debrief.
• At high volume, the form isn’t the problem – applying it consistently is. That’s where structured interview tools, ATS scorecards and AI notetakers come in.

Most interview feedback is a short paragraph of impressions, written after the debrief: “great energy”, “strong communicator”, “not sure about fit”. That makes candidates hard to compare and hiring decisions hard to defend.

Structure fixes most of it. In the most recent large meta-analysis of selection methods, structured interviews were the top-ranked predictor of job performance. Their corrected validity was .42, against .19 for unstructured interviews – roughly double (Sackett, Zhang, Berry & Lievens, Journal of Applied Psychology, 2022). An interview evaluation template is the cheapest way to build a structured interview process.

Every template below:

• is visible on this page, with no email form or download;
• pastes into Word, Google Docs, Excel or your ATS;
• follows published structured-interview research and public-sector guidance from the US Office of Personnel Management (OPM) and Acas.

A disclosure: Sapia.ai makes one of the tools mentioned later in this guide. We run structured interviews at scale and have assessed 10M+ candidates (vendor-reported). These templates apply the same rules we apply to our own interviews.

How we built these templates: six design rules

The tags in each template, such as [rule 3], show which rules it applies.

1. Start from the job, not a generic list

Pick four to six competencies from a job analysis of what the role really requires. These become your evaluation criteria. Generic rows like “attitude” invite gut ratings, and soft skills are where they creep in most: “good communication skills” or “strong interpersonal skills” mean something different to every interviewer. Define the behaviour you’d expect to see instead – for example, “explains a complex point clearly to a non-specialist”. Job analysis is the first of the 15 structured-interview components identified by Campion, Palmer and Campion (1997).

2. Ask the same questions, and score them one at a time

Every candidate gets the same core questions and follow-up prompts. Score each answer straight after it, not one overall score at the end. Keep the CV closed while you score, so the rating reflects the candidate’s interview performance rather than the employer names or educational background on the page – “control of ancillary information”, in Campion’s terms.

Scoring each question also limits first impressions. Rapport-building in the first few minutes inflates scores on early questions (Swider, Barrick & Harris, Journal of Applied Psychology, 2016), so start with one or two unscored warm-up questions.

3. Anchor every point on the scale

Your scoring system should be a 1–5 scale with a written behavioural description at every point, not just at 1, 3 and 5. A major review of structured-interview research found that anchored rating scales “consistently demonstrate” better reliability and validity (Levashina, Hartwell, Morgeson & Campion, Personnel Psychology, 2014).

OPM’s structured interview guidance also recommends writing an example answer, or “benchmark”, for each level. We found no credible interview-specific evidence that 1–7 or 1–10 scales work better, so a fully described 5-point scale is the sensible default.

4. Write evidence, not impressions

Notes should record specific examples of what the candidate said and did – the situation, their action and the result – not adjectives or the interviewer’s impressions. OPM’s guidance says notes should never refer to demographic characteristics. And in the UK, candidates can request their interview notes through a subject access request, so write every note as if the candidate will read it.

5. Rate independently, then discuss

Each interviewer submits scores before the debrief. Disagreements are settled by pointing to notes, and any changed score is recorded with a reason – the sequence OPM sets out. With multiple interviewers, pooled ratings are far more reliable than one person’s: in a meta-analysis by Huffcutt and colleagues (2013), panel interview ratings had a mean inter-rater reliability of .74, against .44 for separate interviews. Some research suggests structured panels benefit from a consensus discussion, so the rule isn’t “never discuss” – it’s “score on your own first”.

6. Weight competencies equally unless the job says otherwise

OPM’s guidance is that “using equal weights is generally the most effective and defensible course of action”. If you do weight one competency more heavily, write down why.

Drop rows for “culture fit” and “likeability” – they’re one of the most common routes for hiring bias into an evaluation form. Research shows interviewers’ fit judgements track how similar a candidate seems to them more than any actual match in values (Cable & Judge, Journal of Applied Psychology, 1997). In elite firms, “fit” often came down to shared hobbies and backgrounds (Rivera, American Sociological Review, 2012). If values matter for the role, define them and anchor them like any other competency. Our guide to culture fit interview questions shows how to keep those questions job-related.

The templates: copy, paste, edit

Template 1: Interview evaluation form (full)

Best for: any structured, competency-based interview – in person, on video or by phone – where each interviewer needs to score the same questions against the same criteria.

Design rules applied: [rule 1] [rule 2] [rule 3] [rule 4] [rule 5]

Part A – Interview details

Part B – Instructions for interviewers

1. Score each answer straight after it, from your notes.
2. Keep the CV closed while you score.
3. Submit this form before the debrief. Don’t read colleagues’ scores first.

Part C – Rating scale

These are generic anchors. Replace them with role-specific anchors for each competency, and ask the hiring manager to write an example answer for each level.

Part D – Competency blocks (repeat for each of your four to six competencies)

Here’s one competency, fully worked, so you can see what anchoring looks like in practice.

Part E – Summary

The recommendation scale has no “maybe”, so interviewers have to commit either way. There’s deliberately no box for an overall impression: the recommendation should follow from the scores.

How to use it:

• Fill in the competency rows from your job analysis before interviewing starts.
• Run a 15-minute calibration in which every interviewer scores the same sample answer and compares results.
• Keep the form identical for every candidate for that role, so each person is compared with other candidates on the same evidence.

When not to use it:

• Short phone screens. It’s too long for a 15-minute call. Use Template 2.
• Hard skills. It doesn’t test them. For coding, data or design roles, pair it with a work sample and the work-sample form in our candidate evaluation form templates.
• Hundreds of applicants per role. At that volume, interviewer drift and slow turnaround become the weak points (see the tools section below).

Paste note: each part is its own table, so it holds its formatting in Word and Google Docs. Most ATSs let you rebuild the same structure as a scorecard or interview kit. Paste the questions and anchors into the scorecard’s guidance fields so interviewers see them while they score.

Template 2: One-page interview evaluation sheet

Best for: phone screens, early-careers first rounds and short video calls where you need a simple evaluation form and a fast, consistent record.

Design rules applied: [rule 2] [rule 3] [rule 4]

How to use it: ask the same three or four questions of everyone at this stage, so you can compare applicants on the same evidence. Score during the call rather than reconstructing it afterwards.

When not to use it:

• Final decisions. The compressed cues are less precise than full anchors, so use this sheet to screen, not to decide.
• Must-have checks that aren’t must-haves. These checks are only as fair as the requirements behind them. Remove anything that isn’t required for the job. That also keeps you on the right side of discrimination law.

Paste note: fits on one A4 page in Word. It also rebuilds easily in Google Forms or Microsoft Forms, though neither keeps an audit trail or stops interviewers seeing each other’s responses by default.

Template 3: Panel debrief record

Best for: turning several interviewers’ individual forms into one defensible decision.

Design rules applied: [rule 4] [rule 5] [rule 6]

How to use it:

1. The chair collects every completed form (Template 1) before the meeting and copies in the scores.
2. Open with the competencies where scores diverge most.
3. Ask the most junior interviewer to speak first, so seniority doesn’t anchor the room.
4. Change an agreed score only when someone points to evidence in their notes.

When not to use it:

• A badly designed interview. The record won’t fix one. If every competency shows a wide spread, the anchors are the problem, so recalibrate.
• Single-interviewer stages. It adds admin where Template 1 alone is enough.

For a weighted panel interview evaluation form that combines scores in a single document, see Template 4 in our candidate evaluation form templates.

Template 4: Interview evaluation matrix (weighted)

Best for: comparing three or more finalists side by side once the debriefs are done – the closest you’ll get to an objective comparison, provided the scores feeding it are sound.

Design rules applied: [rule 6]

Use the agreed scores from each candidate’s panel debrief record (Template 3). Gate on must-haves first: anyone who doesn’t meet a genuine requirement comes off the matrix before you score.

Spreadsheet version. Copy the block below and paste it into cell A1 of a blank Excel or Google Sheets file. The formulas calculate the weighted score and rank. Replace the competency names and example scores with your own, and change a weight only if your job analysis supports it.

Candidate Problem solving Communication Teamwork Planning Technical knowledge Weighted score Rank Deciding evidence

Weight 1 1 1 1 1

Candidate A 4 3 4 5 3 =SUMPRODUCT(B2:F2,B3:F3)/SUM(B2:F2) =RANK(G3,G3:G12)

Candidate B 3 4 4 3 4 =SUMPRODUCT(B2:F2,B4:F4)/SUM(B2:F2) =RANK(G4,G3:G12)

Candidate C 3 3 3 4 3 =SUMPRODUCT(B2:F2,B5:F5)/SUM(B2:F2) =RANK(G5,G3:G12)

The weighted score formula is =SUMPRODUCT(B2:F2,B3:F3)/SUM(B2:F2). With equal weights it’s a simple average. Doubling a competency’s weight to 2 counts it twice. To add candidates (up to ten), copy the last row’s formulas down.

How to use it:

• Leave the weights equal unless the job analysis shows one competency matters more. If you change them, write the reason next to the weights row.
• Treat small gaps – under half a point – as a tie. Decide those on the strongest evidence, not the decimal.

When not to use it:

• When the scores feeding it are weak. A matrix makes any score look precise, but it’s only as good as the anchored, independent scores that go into it.
• When candidates had different interviews. Don’t rank candidates interviewed with different questions or by different panels on one matrix.
• As the decision itself. The top score shows who gave the strongest evidence, but choosing the best candidate is still a human call. Record the reasoning in the last column.

Template 5: Role swap-in rows

Swap these rows into Template 1 for four common role types, and expand each 1 / 3 / 5 cue into five full anchors. They’re starting points, not substitutes for your own job analysis.

For frontline retail and contact-centre roles, use the ready-made scorecards in our guide to interview score sheets.

Early careers and graduates

Experienced professionals

The first row asks about experience directly – what they did, at what scale, with what result – rather than counting years from the CV. Years of experience on its own is a weak predictor of job performance. In the same 2022 meta-analysis, its validity was .07 (Sackett et al., 2022).

People managers

Technical roles

This interview doesn’t test technical skills directly, such as whether someone’s code is correct. Pair it with a work sample or a technical assessment.

Template 6: Evidence-note bank (how to write interview evaluation comments)

Evidence notes are the qualitative feedback that makes a score defensible. Each one should be a concrete example, tied to a question and job-related. Copy the pattern on the right.

Three rules:

1. Quote or paraphrase the answer.
2. Tie every note to a question number.
3. Leave out appearance, accent, age cues, family and anything else that isn’t job-related.

How to copy these templates into Word, Google Docs, Excel or your ATS

• Word or Google Docs: select the table, copy it and paste. In Word, choose “Keep source formatting”. Then edit the competency rows.
• Excel or Google Sheets: copy the spreadsheet block from Template 4 and paste it into cell A1 of a blank sheet. The formulas work as soon as they land.
• Your ATS: rebuild Template 1 as a scorecard or interview kit. Paste the questions and anchors into the guidance or description fields, so interviewers see them while they score.
• Then lock it: save your edited version as the team’s master copy, and keep the questions fixed for each role. Changing questions halfway through evaluating candidates for a role breaks the comparison.

When a template stops scaling: tools that run interview evaluation

For most hiring teams and most later-round interviews, a well-built template plus your ATS scorecard is enough. Templates come under strain in three situations:

• Volume. Dozens of interviewers can’t apply the same form quickly or consistently across hundreds of applicants.
• Distributed panels. Interviewers drift apart without regular calibration.
• Note-taking load. Interviewers who are busy writing listen less well.

Tools for each problem fall into three lanes. The table compares them on what the candidate produces, what any score is built from and whether there’s a published independent bias audit. Details are correct as of October 2026.

Lane 1: Structured first-round interviews at volume

Sapia.ai is built for high-volume and early-careers hiring.

• The interview: every applicant takes the same untimed, mobile, text-first chat interview at the point of application (why candidates prefer chat).
• The scoring: answers are scored against the role’s competencies, without using the CV. Results write back to your ATS.
• Before: Job Analysis Studio builds the competency-mapped interview from the job.
• After: the Talent Intelligence Assistant lets recruiters query everyone they’ve interviewed with Sapia – for example, to compare a shortlist against the competencies. Each answer traces back to what candidates said.

In effect, Sapia applies rules 2–5 at a volume no interview panel could manage.

• Validity: we publish our science, but SHL’s published validity documentation is deeper than ours.
• AI-written answers: Sapia detects AI-generated answers and reports 98% detection with a 1% false-positive rate (vendor-reported).

Limitations:

• The audit is dated. It’s from late 2022 and covers our US and Canada models only. Ask us, and every vendor, for newer role-level results.
• Detection may be less fair for some writers. AI-text detectors have been shown to over-flag non-native English writers (Liang et al., Patterns, 2023), so ask for that group’s false-positive rate.
• It doesn’t suit every candidate or role. A text format won’t fit everyone.
• It isn’t an ATS. It doesn’t replace your later human rounds.

HireVue suits large enterprises that want video and voice AI interviews alongside an assessment library. Its voice AI Interviewer launched in June 2026 and scores answers against what HireVue calls “IO-validated rubrics”. It reports 1,150+ customers and 180M+ assessments (vendor-reported). DCI Consulting has audited its algorithms since 2023, but no audit has been published for the voice product. Camera and voice formats also add friction for some candidates.

SHL Smart Interview comes in two versions:

• Smart Interview Professional (launched February 2025) structures human interviews, with question libraries, scoring guides and AI summaries of interviewers’ ratings. People do the scoring.
• Smart Interview On Demand AI-scores async video, but doesn’t disclose what the score is built from.

SHL’s wider validity research is the deepest in this list. Neither product has a published independent audit.

Lane 2: ATS scorecards – where most teams put Template 1

Greenhouse offers interview kits and scorecards, plus an AI Notetaker that drafts notes against scorecard questions. Greenhouse’s own wording is that “the scorecard remains the source of truth”. Since completing its acquisition of Ezra AI Labs on 27 May 2026, it also has a native voice AI interviewer that scores answers against your own rubric. Warden AI audits its Talent Matching feature monthly, and it has held ISO 42001 certification since February 2026. We couldn’t confirm that audit results for the voice interviewer are published.

Ashby has configurable feedback forms and an AI Notetaker (a paid add-on) that drafts feedback. Ashby explicitly warns admins against prompts that ask the AI to score candidates – a useful guardrail. Its published bias audit (FairNow, August 2024) covers its CV-scoring application review, pooled across roles, rather than interviews. An entry pricing tier is published.

Workable includes interview kits and scorecards on all plans, with published pricing. Its AI agent (launched March 2026) screens CVs rather than evaluating interviews, and it has no published bias audit.

Lever (Employ Inc.) and Teamtailor also support structured interview feedback and AI interview notes. Neither publishes a bias audit.

Lane 3: Interview assistants and notetakers

Metaview writes interview notes and fills in ATS scorecards. It says it captures “objective facts” and leaves the judgements to you.

BrightHire, owned by Zoom since December 2025, adds interview plans, AI notes and an AI screening interviewer called BrightHire Screen.

Neither publishes a bias audit. With either tool, score first and read the AI notes second, so a pre-filled scorecard doesn’t anchor your judgement. For more on this category, see our guide to interview intelligence platforms.

Using interview evaluation templates with AI notetakers and AI interviews

AI is now part of many interviews, on both sides of the table. Four rules keep your evaluation sound:

1. Get consent first. Tell candidates when an interview will be recorded or transcribed, and get their consent. Acas advises consent before recording, and the ICO expects you to be transparent.
2. Score before you read the AI summary. AI-drafted notes and autofilled scorecards can anchor your rating. Submit your own scores, then use the transcript to check facts.
3. Know what any AI score is built from. Ask whether it rests on the candidate’s answers alone or also uses CV and profile data. Then ask whether there’s an independent, published bias audit of that specific product, broken down by role rather than pooled.
4. Plan for AI-assisted answers. Some candidates use AI tools during remote interviews, and no template can detect that. Follow-up questions that probe for specifics help. If you use an AI interview vendor, ask for its published detection and false-positive figures.

Check the rules where you hire. If you use automated tools to score candidates, local law may apply:

• New York City: Local Law 144 requires a bias audit and candidate notice.
• Illinois: HB 3773 has applied since January 2026.
• Colorado: SB 26-189 applies from January 2027.
• EU: the AI Act’s high-risk obligations for employment apply from 2 December 2027.

Check current requirements with counsel before you rely on any of this.

Keeping interview evaluation records: access and retention

This isn’t legal advice, but it’s worth knowing the basics before you file a stack of evaluation forms.

• UK: candidates, including rejected candidates, can make a subject access request, and that can cover interview notes and outcomes.
  • The ICO doesn’t set a fixed retention period. It expects you to justify the period you choose and to tell candidates what it is (ICO recruitment guidance, still being finalised).
  • From 1 October 2026, the Employment Rights Act 2025 extends most employment tribunal time limits, including for discrimination claims, from three months to six (Lewis Silkin, 2026). If your policy was to keep unsuccessful candidates’ records for six months, review it with your legal team.
• US: the EEOC requires employers to keep hiring records for at least one year, and until any charge is resolved (29 CFR 1602.14).
• Australia: unsuccessful applicants can generally ask for access to the information you hold about them (OAIC).

The practical point is the same everywhere: write notes you’d be comfortable for the candidate, a tribunal or a regulator to read. Template 6 helps with exactly that.

Where Sapia fits (and where it doesn’t)

Our lane: the first-round, structured interview at volume. When a role draws hundreds or thousands of applicants, the weak point isn’t the form. It’s that dozens of interviewers can’t apply it consistently, quickly or without seeing the CV.

Sapia gives every candidate the same structured questions, in an untimed text format that works on any phone, and scores the answers without using the CV. An independent BLDS audit ran 72 tests in late 2022. It covered our US and Canada models and found no evidence of practically significant disparate impact (FAIR framework). Your people then interview the shortlist using templates like the ones above.

What that looks like in practice (customer-reported, via Sapia case studies):

• Candidate experience at scale: at Woolworths, 82.6% of candidates completed the chat interview and rated it 9.2/10 (Woolworths case study).
• Recruiter time: Starbucks saved 1,900 hours a month that had gone on screening CVs (Starbucks case study).
• Fairness through the funnel: LNER saw a consistent 30% ethnic minority representation through all hiring stages (LNER case study).

More on how this works for volume hiring and diversity hiring.

Where we don’t fit, and what to use instead:

• Small hiring volumes → a good template in Docs or your ATS is enough.
• Later-round human interviews → Templates 1–4 and your ATS scorecards.
• Voice or video AI interviews → Greenhouse or HireVue.
• The deepest published validity documentation → SHL.
• Coding and hard-skills tests → a technical assessment platform.
• A system of record → your ATS. Sapia sits on top of it (integrations).

The template principles are the same whether a hiring manager applies them in a Word document or Sapia applies them to 10,000 applicants: the same questions, anchored scores, evidence over impressions and independent ratings. If volume is what’s breaking your hiring process, book a demo.

How to roll out an interview evaluation template: a 5-step process

1. Run a short job analysis and pick four to six competencies. Use the job description and conversations with people who do the role well. Our job description template and Job Analysis Studio can help.
2. Write the interview questions and anchors before you advertise. Write one core question and two or three follow-ups per competency, with five anchored levels each. Ask the hiring manager to write an example answer for each level. Our guide to competency-based interview questions has question ideas.
3. Calibrate in 20 minutes. Every interviewer scores the same two sample answers. Discuss any spread of two points or more, and tighten the anchors.
4. Run the interview process as designed. Ask the same questions of everyone. Have each interviewer submit their form before the debrief, use the panel debrief record (Template 3), then compare finalists on the matrix (Template 4).
5. Review after the first 10 hires. Look for three things:
   • Rater spread and leniency: does one interviewer give all 4s and 5s?
   • Adverse impact: compare pass rates across groups.
   • Prediction: as data builds, check whether interview scores line up with early performance and retention.
6. For a fuller calibration plan, see our guide to interview score sheets.

Conclusion

A good evaluation form has few boxes and detailed anchors. Copy the one that fits, replace the competencies with your own, and protect the rule that everyone scores independently before the debrief. That rule, more than any layout, is what makes the scores worth comparing.

For roles where volume makes even a great template hard to apply consistently, Sapia runs the structured first interview for every applicant, CV-blind and independently audited (BLDS, late 2022, US and Canada models), so your interviewers can spend their time on the shortlist. See how the chat interview works.

FAQs

About Author

Laura Belfield

Head of Marketing