Interview Scorecard Template (With a Filled-In Example)
A copyable interview scorecard template with 1-4 behavioural anchors, an evidence column and must-pass flags, plus one scored example and a retention note.
An interview scorecard is a one-page form that every interviewer fills in against the same four to six competencies for a role, using a written 1-4 scale, and that records the evidence behind each score. Below is a template you can copy as a markdown table, one filled-in example for a fictional role and candidate, the research behind the format, and a note on how long to keep the sheet.
I run a recruitment agency, ScalingPPL, and I built Pickr, an ATS. This is the sheet I would hand to a hiring manager who has never used one.
Why another interview scorecard template?
Because the ones on the first page of results are blank. I opened three on 2026-10-14. AIHR offers an Excel file and recommends a 1-5 or 1-10 scale but shows no scored candidate. OmniHR asks for your details before the download and shows no scored candidate either. Journeyfront offers an Excel workbook with several formats and no worked example. A blank template answers "what are the columns", not "what does a 3 look like", which is the question every first-time interviewer actually has.
What goes on an interview scorecard?
Six things, in the order the interviewer fills them in:
- Header: role, stage, candidate, interviewer, date. Without the stage you cannot compare a screen scorecard with a final-round one.
- Competencies: four to six, each taken from a requirement in the job brief; more than six and there is no time to collect evidence on each.
- A 1-4 scale with anchors: one sentence per level per competency describing observable behaviour. Google's re:Work guide calls this a standardised rubric "so that all reviewers have a shared understanding" and documents example answers at each level, from poor to outstanding (re:Work structured interviewing guide, March 2026 version, accessed 2026-10-14). I use four levels rather than five because an even scale has no safe middle; from our work at ScalingPPL, a 1-5 scale turns into a 3-5 scale within a week.
- Evidence column: what the candidate said or did, in enough detail that the next interviewer can act on it.
- Must-pass flag: the one or two criteria where a 1 ends the process on its own.
- Overall recommendation with reasoning and a concerns line, filled in last.
The interview scorecard template (copy this)
Header
| Field | Entry |
|---|---|
| Role | |
| Interview stage | screen / hiring-manager interview / panel / final |
| Candidate | |
| Interviewer | |
| Date and duration | |
| Submitted before the debrief? | yes / no |
Scale
| Score | Meaning |
|---|---|
| 1 | Below the bar. No evidence, or evidence of the opposite. |
| 2 | Partly there. Some evidence, with gaps the role cannot absorb. |
| 3 | At the bar. Clear evidence of doing this at the level the role needs. |
| 4 | Above the bar. Evidence of doing this at a higher level or in a harder context. |
Anchors (one row per competency; replace the placeholders with the behaviour this role needs)
| Competency (from the job brief) | Must-pass? | 1 looks like | 2 looks like | 3 looks like | 4 looks like |
|---|---|---|---|---|---|
| Competency A | yes / no | ||||
| Competency B | yes / no | ||||
| Competency C | yes / no | ||||
| Competency D | yes / no | ||||
| Competency E (optional) | yes / no |
Scoring sheet
| Competency | Score (1-4) | Evidence: what the candidate said or did | Question asked | Probe in the next round |
|---|---|---|---|---|
| Competency A | ||||
| Competency B | ||||
| Competency C | ||||
| Competency D | ||||
| Competency E |
Overall
| Field | Entry |
|---|---|
| All must-pass criteria at 2 or above? | yes / no |
| Recommendation | strong hire / hire / no hire / strong no hire |
| Reasoning (two or three sentences referring to the rows above) | |
| Concerns, even if recommending hire | |
| What the next interviewer should probe |
Three rules. Write the evidence before the score, because a score written first gets rationalised afterwards. Submit before the debrief and before reading anyone else's sheet; the reasons are in the hiring manager's guide to interview scorecards. Keep the anchors fixed for the whole search, or the early scores become worthless.
Why does a structured scorecard beat an unstructured interview?
Because the published validity numbers say so, and the most recent re-analysis widened the gap.
In 2022, Sackett, Zhang, Berry and Lievens revisited the validity estimates that personnel selection had relied on since Schmidt and Hunter's 1998 summary, argued that the standard corrections for range restriction had systematically overcorrected, and re-estimated the validity of 25 selection procedures (Sackett et al., 2022, Journal of Applied Psychology, 107(11), 2040-2068, doi 10.1037/apl0000994). Table 3 of the paper gives, among others:
| Selection procedure | Schmidt and Hunter (1998) | Sackett et al. (2022) | Source |
|---|---|---|---|
| Employment interviews, structured | .51 | .42 | Sackett et al. 2022, Table 3 |
| Job knowledge tests | .48 | .40 | same |
| Empirically keyed biodata | .35 | .38 | same |
| Work sample tests | .54 | .33 | same |
| Cognitive ability tests | .51 | .31 | same |
| Assessment centers | .37 | .29 | same |
| Employment interviews, unstructured | .38 | .19 | same |
The abstract sums it up: mean validity estimates "reduced by .10-.20 points", and "structured interviews emerged as the top-ranked selection procedure". The structured estimate has a standard deviation of .19 and a lower 80% credibility value of .18, so the spread is wide; the unstructured estimate's lower credibility value sits just below zero.
The operational side comes from Google. Its re:Work guide reports that pre-made questions, guides and rubrics save "on average 40 minutes per interview", and that rejected candidates who had a structured interview were 35% happier than those who did not (re:Work, accessed 2026-10-14). Both are Google's internal figures without a published method; read them as one large employer's experience.
A scorecard is the part of a structured interview that survives the meeting: same questions, same rubric, written evidence. Drop one and you are back on the unstructured row. The longer argument is in the structured interview: theory right, practice wrong.
The filled-in example: Candidate A, mid-level Account Manager
Everything below is fictional. The role and the company are invented for this article; the candidate is named only as Candidate A.
Header
| Field | Entry |
|---|---|
| Role | Account Manager, mid-level, B2B software, book of about 40 existing customers, remote within the EU |
| Interview stage | Hiring-manager interview (stage 2 of 3) |
| Candidate | Candidate A |
| Interviewer | Head of Customer Success |
| Date and duration | 2026-10-20, 45 minutes |
| Submitted before the debrief? | yes |
Anchors for this role
| Competency | Must-pass? | 1 | 2 | 3 | 4 |
|---|---|---|---|---|---|
| Owning renewals and expansion | yes | Never carried a renewal number | Supported renewals; someone else owned the number | Owned a renewal number; can explain a win and a loss | Owned renewals and expansion on a larger book, with a repeatable method |
| Discovery questioning | no | Pitches instead of asking | Accepts the first answer | Follows up until the real problem is on the table | Reframes the stated problem into the underlying one |
| Handling a hard customer conversation | yes | Avoids it or escalates at once | Cannot say what was said | Names a specific conversation, what they said, the outcome | Same, plus what they would do differently |
| Working with product and support | no | Treats other teams as blockers | Hands issues over without context | Brings the customer's context; tracks to closure | Changes how the other team works with customers |
| Clarity in writing and speech | no | Rambles; needs interrupting | Clear when prompted, long when not | Answers first, context second | Answers the listener can repeat back |
Scoring sheet
| Competency | Score | Evidence: what Candidate A said or did | Question asked | Probe in the next round |
|---|---|---|---|---|
| Owning renewals and expansion | 3 | Carried the renewal number for about 30 mid-market customers for two years. Described one lost renewal (champion left, the new contact had used a competing tool, A found out six weeks before the date) and the change afterwards (quarterly exec check-ins for larger accounts). Expansion stayed at "we upsold when it came up". | "Walk me through the last renewal you lost and what you changed." | One expansion A initiated. |
| Discovery questioning | 4 | In the role-play (customer complains about slow reporting), A asked who reads the reports, how often and what decision they drive before offering anything. Reframed "slow reports" as "finance cannot close month-end"; the mock customer agreed. | Role-play, 10 minutes. | None. |
| Handling a hard customer conversation | 2 | Described telling a customer a promised feature was delayed. Gave the outcome (customer stayed) but not what was said; two follow-ups stayed at "I was honest and explained the situation". | "Tell me about delivering bad news to a customer. What did you actually say?" | Must-pass at 2: scripted bad-news role-play, scored on the same anchors. |
| Working with product and support | 3 | A recurring bug hit three accounts: wrote a one-page summary with revenue at risk and reproduction steps, got it prioritised, reported back to each customer when it shipped. | "Tell me about a time you needed product or support to act for a customer." | None. |
| Clarity in writing and speech | 3 | Most answers led with the result; two ran long and needed a prompt. The written exercise (reply to an unhappy customer) was short and specific. | Observed throughout, plus the written exercise. | None. |
Overall
| Field | Entry |
|---|---|
| All must-pass criteria at 2 or above? | yes (renewals 3, hard conversations 2) |
| Recommendation | hire |
| Reasoning | Clear evidence of owning a renewal number and of real discovery skill, the two things this book needs most. The gap is hard conversations, where the evidence stayed vague across three attempts; a 2 rather than a 1 because the outcome was good and A did not avoid the conversation. |
| Concerns | Cannot tell whether A has a method for difficult conversations or has had easy customers. Expansion ownership is thinner than renewal ownership. |
| What the next interviewer should probe | The bad-news role-play. One expansion A initiated. |
Notice what the sheet does not contain: nothing about how Candidate A seemed, nothing about energy or fit, nothing about age, family, accent or nationality. Every cell is something the candidate said or did in response to a named question; that is what makes the 2 defensible and the 4 believable. In the debrief, start from the must-pass rows and the rows where interviewers disagree, and let the evidence columns settle "what did each of you hear".
What should never go on an interview scorecard?
Anything that is not evidence against a competency.
Impressions dressed as scores: "good energy", "would fit in", "a bit junior". If you cannot point to a question and an answer, it is not evidence. Protected characteristics and anything adjacent: age, family plans, health, nationality, religion, accent. Under the German AGG a rejected applicant has two months from receiving the rejection to assert a claim in writing (§ 15 Abs. 4 AGG, accessed 2026-10-14); a scorecard that mentions any of those is the document the claim is built on. And excessive information: the UK ICO's draft recruitment guidance says employers "must not collect excessive information about candidates at interview", and that a candidate can make a subject access request for what was recorded at interview, including interview notes (ICO, shortlisting, testing and interviewing candidates, draft, accessed 2026-10-14).
Write every scorecard as if the candidate will read it, because in some jurisdictions they can.
How long do you keep a completed scorecard?
This is not legal advice; it is what the public sources say as of 2026-10-14.
| Jurisdiction | What the source says | Source |
|---|---|---|
| United States | Hiring records, application forms included, must be kept for one year from the date the record was made or the personnel action taken, whichever is later; if a discrimination charge is filed, all relevant records until final disposition | 29 CFR 1602.14, accessed 2026-10-14 |
| Germany | A rejected applicant must assert a compensation claim in writing within two months of receiving the rejection, so the scorecard is the employer's evidence for at least that window | § 15 Abs. 4 AGG, accessed 2026-10-14 |
| United Kingdom | Data protection law "does not specify timescales for keeping recruitment records"; set clear retention periods, and do not keep records beyond the statutory period in which a claim can be brought. DRAFT guidance, under review after the Data (Use and Access) Act; the consultation on it closed 5 March 2024 | ICO, keeping recruitment records, draft and ICO consultation page, accessed 2026-10-14 |
Whatever period you choose, apply it to the scorecard and the interview notes together.
How does this work inside an ATS?
An ATS adds enforcement and recall.
In Pickr, every interviewer on the panel scores against the same criteria for the role, and each rating is tied to what the candidate said in the live transcript rather than to an overall impression, so the panel sees at a glance where it disagrees. After the interview, each criterion already carries the relevant passages from the transcript; the interviewer confirms, corrects or disagrees, and a named person confirms every evaluation before it counts. Candidate data runs with EU hosting in Frankfurt, and names and email addresses are redacted from AI prompts by default. Details are on the interview intelligence feature page; the case for drafting from a transcript is in interview intelligence: how AI is changing candidate evaluation.
The other failure mode is access. If the people who sit on panels have no login, the recruiter fills in the evidence column second-hand, and the scorecard records the recruiter's memory of the interviewer's memory. More on that in interviewer ATS access as a data decision.
The columns are the same on paper and in an ATS. What changes is whether they get filled in within 24 hours, by the person who was in the room.
Sources
- Google re:Work, A guide to structured interviewing for better hiring practices (updated March 2026): https://rework.withgoogle.com/intl/en/guides/a-guide-to-structured-interviewing-for-better-hiring-practices (accessed 2026-10-14)
- Sackett, P. R., Zhang, C., Berry, C. M., & Lievens, F. (2022). Revisiting meta-analytic estimates of validity in personnel selection: Addressing systematic overcorrection for restriction of range. Journal of Applied Psychology, 107(11), 2040-2068: https://doi.org/10.1037/apl0000994 (accessed 2026-10-14; the DOI resolves to APA PsycNet)
- Sackett et al., copy of the 2021 advance online version of the same paper (APA typeset, with Table 3): https://www.gwern.net/doc/statistics/meta-analysis/2021-sackett.pdf (accessed 2026-10-14)
- ICO, Employment practices and data protection: recruitment and selection (draft, under review): https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/employment/recruitment-and-selection/ (accessed 2026-10-14)
- ICO, Keeping recruitment records (draft): https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/employment/recruitment-and-selection/keeping-recruitment-records/ (accessed 2026-10-14)
- ICO, Shortlisting, testing and interviewing candidates (draft): https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/employment/recruitment-and-selection/shortlisting-testing-and-interviewing-candidates/ (accessed 2026-10-14)
- ICO, Recruitment and selection consultation (closed 5 March 2024): https://ico.org.uk/about-the-ico/ico-and-stakeholder-consultations/2024/08/recruitment-and-selection-consultation/ (accessed 2026-10-14)
- 29 CFR 1602.14, Preservation of records made or kept (Cornell LII): https://www.law.cornell.edu/cfr/text/29/1602.14 (accessed 2026-10-14)
- § 15 AGG, Entschädigung und Schadensersatz (gesetze-im-internet.de): https://www.gesetze-im-internet.de/agg/__15.html (accessed 2026-10-14)
- AIHR, Interview scorecard template: https://www.aihr.com/blog/interview-scorecard/ (accessed 2026-10-14)
- OmniHR, Interview scorecard template: https://omnihr.co/templates-guides/interview-scorecard-template (accessed 2026-10-14)
- Journeyfront, Interview scoring template: https://www.journeyfront.com/interview_scoring_template (accessed 2026-10-14)
Frequently Asked Questions
What should an interview scorecard template include?
A header (role, stage, candidate, interviewer, date), four to six competencies taken from the job requirements, a 1-4 scale with a written anchor per level, an evidence column for what the candidate actually said or did, a must-pass flag on the criteria that end the process on their own, and one overall recommendation with the reasoning.
Why use a 1-4 scale instead of 1-5 on an interview scorecard?
An even scale has no middle value, so the interviewer has to decide whether the evidence was above or below the bar. On a 1-5 scale the 3 becomes the default for every answer that was not probed. With four levels and written anchors, two interviewers giving the same score are more likely to mean the same thing.
Are structured interviews with scorecards more predictive than unstructured interviews?
Yes, by a wide margin in the published research. The 2022 re-analysis by Sackett, Zhang, Berry and Lievens in the Journal of Applied Psychology puts structured interviews at an operational validity of .42 against .19 for unstructured interviews, and ranks structured interviews first among the procedures it reviewed.
How long should you keep a completed interview scorecard?
It depends on the jurisdiction, and this is not legal advice. US regulations require hiring records to be kept for one year. In Germany a rejected applicant has two months from receiving the rejection to assert a claim under the AGG. The UK ICO's draft guidance sets no fixed period and tells employers to set their own.
Can a candidate request to see their interview scorecard?
Under UK data protection law the ICO's draft recruitment guidance says a candidate can make a subject access request for information recorded at interview, including interview notes. Write every scorecard so you would be comfortable with the candidate reading it: observed evidence only, nothing about protected characteristics.
Written by Andreas Amann, Founder of Pickr and of ScalingPPL