1. Blog
  2. Interviewing

Interviewing

Interview scorecard: a template, worked example and hiring decision rules

Alfie Whattam, Co-founder and CEO, Alfa AI ·

Two interviewers at a meeting-room table in a city office, each filling in a printed scorecard with a pen after an interview

An interview scorecard is a shared rubric for assessing candidates against job-related criteria using consistent questions, scoring anchors and evidence. Build it before interviewing, score independently, then use must-have requirements and weighted totals to guide the hiring decision.

A 2022 meta-analysis by Sackett and colleagues ranked structured interviews as the strongest selection procedure for predicting job performance. The point is to structure the assessment, not simply attach numbers to impressions.

What a good interview scorecard includes

Use this checklist before giving the scorecard to your panel:

  • Role outcome: what the person needs to achieve.
  • Job-related criteria: the capabilities needed to deliver that outcome.
  • Must-have flags: requirements that a high score elsewhere cannot compensate for.
  • Consistent questions: a core question for each criterion.
  • Behavioral anchors: observable evidence describing different scoring levels.
  • Weights: the relative importance of each criterion.
  • Evidence notes: what the candidate said or demonstrated.
  • Overall recommendation: completed last, after the evidence and scores.

I'd remove any criterion the hiring manager cannot connect to actual work. “Executive presence” is not useful until you define the behavior the job requires.

How to build a structured interview scorecard

  1. Start with the outcome. Describe what the person must achieve in their first six to twelve months. For a customer success manager, that might mean owning renewal risks and helping customers adopt the product.
  2. Choose four to six criteria. Work backward from that outcome. Mark genuine must-haves before seeing candidates. Keep them consistent with your candidate screening rules, while using interviews to investigate evidence more deeply.
  3. Write a core question for each criterion. Ask for a specific situation, the candidate's actions and the result. Agree neutral follow-ups, such as asking what they personally owned.
  4. Write anchors for scores of 1, 3 and 5. Describe actions, reasoning or work demonstrated. Avoid labels such as “confident,” “likable” or “good communicator.” Define 3 as meeting the role's requirement, not as an average candidate.
  5. Set weights before interviewing. Give greater weight to work that matters most to the outcome. Make the weights total 100%, and test the rubric by having panelists score the same practice answer independently.

The scale, weights and decision rules below are proposed working rules, not research-backed hiring cutoffs.

Copyable interview scorecard template

Replace every bracket before using this job interview scorecard template. The completed example below shows how specific the anchors and notes should be.

INTERVIEW SCORECARD
Role:
Candidate:
Interviewer:
Interview stage:
Role outcome: [What this person must achieve]

Criterion A: [Capability taken from the job]
Must-have: [Yes/No] | Weight: [%]
Question: [Core question asked of every candidate]
1: [Observed actions below the requirement]
3: [Observed actions meeting the requirement]
5: [Observed actions demonstrating advanced capability]
Score: [1–5 / Not assessed]
Evidence notes: [Actions, reasoning, result, missing evidence]

Criterion B: [Capability taken from the job]
Must-have: [Yes/No] | Weight: [%]
Question: [Core question asked of every candidate]
1: [Observed actions below the requirement]
3: [Observed actions meeting the requirement]
5: [Observed actions demonstrating advanced capability]
Score: [1–5 / Not assessed]
Evidence notes: [Actions, reasoning, result, missing evidence]

Criterion C: [Capability taken from the job]
Must-have: [Yes/No] | Weight: [%]
Question: [Core question asked of every candidate]
1: [Observed actions below the requirement]
3: [Observed actions meeting the requirement]
5: [Observed actions demonstrating advanced capability]
Score: [1–5 / Not assessed]
Evidence notes: [Actions, reasoning, result, missing evidence]

Criterion D: [Capability taken from the job]
Must-have: [Yes/No] | Weight: [%]
Question: [Core question asked of every candidate]
1: [Observed actions below the requirement]
3: [Observed actions meeting the requirement]
5: [Observed actions demonstrating advanced capability]
Score: [1–5 / Not assessed]
Evidence notes: [Actions, reasoning, result, missing evidence]

Must-have floor met: [Yes/No/Unresolved]
Weighted total:
Outstanding evidence gaps:
Overall recommendation, complete last: [Hire/No hire/Hold]
Reason tied to the role outcome:

Use 2 or 4 when the evidence falls between adjacent anchors, and explain why. Use Not assessed when the interview did not produce enough evidence. Missing evidence is not automatically evidence of poor capability.

Interview scorecard example: customer success manager

This is an illustrative candidate, not a customer case study. The questions, evidence and scores are invented to demonstrate the rubric.

Role: Mid-level customer success manager.

Role outcome: Independently manage an account portfolio by identifying renewal risks, building adoption plans, coordinating stakeholders and prioritizing customer work.

Renewal risk diagnosis

Weight: 35%. Must-have: Yes.

Question: Walk us through an account you believed might not renew. How did you diagnose the risk and decide what to do?

  • 1: Relies on the relationship alone; cannot identify risk signals or explain a recovery action.
  • 3: Identifies relevant signals, investigates causes and agrees a recovery plan with the customer.
  • 5: Tests competing explanations using account data and stakeholder input, monitors the recovery plan and adjusts it as evidence changes.

Score: 4. Evidence: Compared falling usage with a sponsor's departure, consulted users and agreed a recovery plan with the new sponsor. Explained monitoring adoption, but did not show how monitoring changed the plan.

Adoption planning

Weight: 30%. Must-have: Yes.

Question: Describe an onboarding or adoption plan you owned. How did you decide whether it was working?

  • 1: Lists training activities without connecting them to customer goals or changed usage.
  • 3: Connects a customer goal to adoption milestones, named actions and a relevant usage measure.
  • 5: Differentiates user groups, investigates adoption barriers and changes the plan based on observed behavior.

Score: 3. Evidence: Connected the customer's reporting goal to setup, training and a usage review. Named responsibilities and the measure to inspect. Did not demonstrate adapting the plan for different user groups.

Stakeholder alignment

Weight: 20%. Must-have: No.

Question: Tell us about conflicting requests from stakeholders within a customer account. What did you do?

  • 1: Accepts incompatible requests or avoids explaining the conflict.
  • 3: Clarifies competing needs, explains constraints and agrees a next step.
  • 5: Tests alternative options with stakeholders, secures an agreement and records ownership of commitments.

Score: 4. Evidence: Separated a sponsor's rollout needs from procurement's price concern. Compared scope options and agreed an escalation route. Did not explain how final commitments were recorded or assigned.

Portfolio prioritization

Weight: 15%. Must-have: No.

Question: Describe competing demands across your accounts. How did you decide what to handle first?

  • 1: Responds to the loudest request without explaining customer impact or risk.
  • 3: Weighs renewal timing, customer impact and workload, then explains the sequence of work.
  • 5: Tests trade-offs, reallocates effort deliberately and defines triggers for revisiting priorities.

Score: 3. Evidence: Prioritized a near-renewal account with blocked users over routine training. Explained the workload trade-off, but gave no trigger for revisiting that priority.

Completed score summary

CriterionWeightScoreWeighted contribution
Renewal risk diagnosis35%41.40
Adoption planning30%30.90
Stakeholder alignment20%40.80
Portfolio prioritization15%30.45
Total100%3.55 out of 5

Overall recommendation, entered last: Hire. The evidence supports the stated role outcome and clears both must-have floors. Adapting adoption plans is a development area, not a waived requirement. This is the interviewer's recommendation, not an automatic panel decision.

Turn interview scores into a hiring decision

Use the debrief to reconcile evidence, not to collect opening statements about who everyone liked.

  1. Score alone before the debrief. Each interviewer submits evidence and scores independently. This follows the US Office of Personnel Management's individual-rating-then-discussion approach. Keep recommendations private until discrepancies have been examined.
  2. Discuss scoring gaps before sharing hiring views. Where interviewers differ by two or more points on a criterion, compare the evidence against its anchors. Also examine any disagreement that crosses a must-have floor, even if the gap is smaller. Agree a final criterion score only where the evidence supports it.
  3. Apply the must-have floor. A final score of 1 or 2 on any must-have means no hire, regardless of the total. Do not average away a failed requirement. An unresolved evidence gap means hold for a focused follow-up, not pass.
  4. Calculate and compare weighted totals. Multiply each final score by its weight as a decimal, then add the contributions. Compare only candidates who clear the must-have floor and have complete assessments. Do not combine totals from different rubrics.
  5. Make and record the decision. The hiring owner explains whether the evidence supports the role outcome. The highest total is not an obligation to hire. If the panel remains split because evidence is missing, collect that evidence rather than settling it by vote.

Keep original scores and document any revisions. If eligible candidates tie, use a consistent, job-related follow-up rather than inventing a new criterion after meeting them.

Five mistakes that ruin interview scorecards

MistakeWhat to do instead
Criteria copied from a generic listConnect every criterion to a responsibility or outcome in this job.
Anchors describing personalityReplace “polished” with evidence such as explaining a trade-off and checking stakeholder understanding.
Scoring after the debriefSubmit independent scores before hearing colleagues' recommendations.
A halo from a strong answerScore each criterion against its own evidence. A persuasive renewal story does not prove prioritization.
An overall score with no notesRequire evidence for each criterion before accepting a recommendation.

A useful note names an action and its limits. “Strong answer” gives the panel nothing to inspect.

Keep interview scorecards fair and legally careful

Where AI interviews belong in the scorecard

Treat an AI score as an input to review, not a new criterion called “AI fit.” For an AI interview or notetaker workflow, ask for the underlying transcript or recording and evidence attached to each assessment.

A person should read that evidence against the human rubric and make the call. If the first-round conversation never tested portfolio prioritization, leave that criterion unassessed until a later stage. Do not count the same answer again simply because both AI and a person scored it.

Alfa AI shows the AI Interview recording, a speaker-labeled transcript and a Candidate Scorecard containing the AI score and supporting evidence, alongside Screening Checks. A hiring team member reviews each AI Interview recording and decides the next step. I'd keep that AI score beside the human scorecard, outside its weighted total.

Before rollout, use an AI recruitment tool pilot scorecard to assess the workflow. UK teams should also work through the checks for using AI in recruitment. Keep the human decision traceable to the job requirements and the candidate's evidence.

Frequently asked questions

What is the best interview scorecard template?
Choose a format your panel can complete during its actual workflow, whether that is a document, spreadsheet or form. I'd test usability before choosing presentation: can an interviewer find the relevant anchor while taking notes, and can the hiring owner trace a recommendation back to evidence? Avoid designs that hide notes behind a prominent overall rating.
How many criteria should an interview scorecard have?
I'd start with four to six. For a complex role, distribute additional assessment across interview stages rather than extending every panelist's form. Assign each criterion an owner and decide where the strongest evidence should come from. Keep eligibility checks, scheduling details and compensation discussions separate from the scored capabilities.
Should interviewers see each other's scores before the debrief?
Not before submitting their own assessment. Have a facilitator check that forms are complete without circulating colleagues' judgments. At the debrief, release the criterion scores and supporting notes for comparison. If someone revises a score, preserve the original and record what evidence changed their assessment, rather than silently overwriting it.
What if a candidate has no directly comparable experience?
Look for transferable evidence of the required capability rather than an identical job title or customer type. You could use a job-related scenario, provided candidates receive a comparable opportunity to demonstrate the skill. Agree that approach beforehand. Do not lower an essential requirement, but do not confuse an unfamiliar setting with an inability to do the work.
Can I change the scorecard midway through hiring?
If you discover that the rubric misses essential work, pause and correct it rather than knowingly using a poor assessment. Document the change and reassess earlier candidates against the revised criteria. Where their interviews did not collect the necessary evidence, offer a comparable follow-up. Do not compare old totals with totals produced by the revised rubric.
Your AI recruiter

Alfa sources, screens and shortlists candidates, so your team only spends time on the people worth meeting.