Updated 2026-09-16
An interview scorecard is a structured evaluation form where interviewers record scores and notes for each job-relevant question against predefined criteria — so every candidate for the same role is assessed on the same dimensions rather than remembered impressions after the call.
A useful scorecard starts with the role and requisition identifier, then lists each scored question with the competency it measures — communication clarity, safety awareness, sales closing instinct, clinical judgment under protocol constraints. Beside each question sits a rating scale with behavioural anchors: what a '4' sounds like versus a '2', not just 'strong / weak' labels that mean different things to different interviewers.
Space for verbatim evidence matters. 'Good communicator' is not auditable; 'Explained escalation path clearly when customer refused script' is. Scorecards that force a one-line evidence note per rating create documentation hiring managers can revisit when two candidates tie or when a candidate disputes a rejection.
Aggregate sections — overall recommendation, flags for follow-up, knockout criteria — sit at the bottom. Knockouts should be explicit and job-related: cannot work night shifts, lacks required credential, failed mandatory safety scenario — not vague culture concerns.
| Artifact | Purpose |
|---|---|
| Interview rubric | Defines what each score level means for a competency |
| Interview scorecard | Captures per-candidate scores and evidence during/after the interview |
| Unstructured notes | Freeform impressions — useful context but poor for comparison |
| ATS evaluation form | Often a digital scorecard synced to pipeline stage |
Paper scorecards still appear in field hiring — store managers interviewing retail associates, floor supervisors screening warehouse temps. They work when volume is low and interviewers sit in the same room for calibration. They fail at scale: data entry lag, lost forms, inconsistent storage, no analytics across requisitions.
Digital scorecards inside an ATS or interview platform enforce required fields, timestamp submissions, and aggregate scores automatically. AI interview engines generate scorecard-like output programmatically — each rubric dimension populated with a score and transcript citation — which recruiters treat as the first-round scorecard before human stages.
The design principle stays the same regardless of medium: score job behaviours, not likability; record evidence, not vibes; keep criteria stable for the life of the requisition.
First-round scorecards typically score four to six competencies. More than that dilutes interviewer attention and lengthens the interview without improving signal.
Policies vary. Some employers share summary feedback; many do not share internal scoring forms. Regulatory contexts may require explanation of automated decision factors — check applicable law.
No. Scorecards should be role-specific. Reusing a generic 'communication + teamwork' card across nursing, sales, and engineering roles produces scores that do not predict performance.
Four- or five-point scales with defined anchors outperform ten-point scales in inter-rater reliability studies for behavioural interviews — interviewers struggle to distinguish a 6 from a 7 consistently.
Head back to Hiring glossary or start now.
· 3 free credits · pay per interview · nothing recurring