Skip to main content

Free template

Hiring decision scorecard

The decision is the moment a structured process either pays off or evaporates. This template records what each interviewer saw, where the evidence was thin, what risk you are knowingly accepting, and who decided — in a form you can revisit in six months.

Who fills it in

Hiring manager, completed with the full panel present.

How long it takes

Ten minutes per interviewer, twenty minutes in the debrief.

Reviewed by

Joseph Edney, 29 October 2026

When to use it

Use it when

  • You have a candidate at final stage and need a decision the panel can stand behind.
  • The panel disagrees and the discussion is circling.
  • You want a record of why a hire was made, to learn from later.

What good looks like

  • Each interviewer writes their recommendation before the debrief discussion starts, so nobody anchors on the loudest voice.
  • Every capability is marked as evidenced, partly evidenced or not tested — “not tested” is a legitimate and useful answer.
  • Accepted risks are named explicitly, with what you will do about each in the first months.
  • The record says who decided, not just what was decided.

Common mistakes

  • Scoring on a numeric scale and then averaging. Averages hide the one interviewer who saw a genuine problem.
  • Treating “no evidence” as “no concern”. Untested is not the same as satisfied.
  • Deciding in a corridor and writing the form afterwards to match.
  • Letting fatigue with the search decide. A process running long is a reason to fix the process, not to lower the bar.

The template

Anything you type stays in this browser tab, so reloading the page keeps it. It is not sent to us or added to the link, so it is safe to put real salary numbers and colleagues’ names in. Closing the tab clears it — copy or print before you go.

Candidate and process

Evidence against the scorecard

One row per must-have. Mark the evidence honestly: evidenced, partly evidenced, or not tested.

RowEvidenced / partly / not testedThe evidence itselfWho saw it
Must-have 1
Must-have 2
Must-have 3
Must-have 4
Must-have 5

Independent recommendations

Each interviewer completes their own row before the discussion.

RowRecommendationThe single strongest reasonThe main reservation
Interviewer 1
Interviewer 2
Interviewer 3
Interviewer 4 (optional)

Decision

This feeds directly into the onboarding handoff.

A worked example

Illustrative example. The company, role and answers are fictional, written to show a finished template. It is not a client, a quote or a result. The fictional clinic-scheduling startup from the earlier examples, the day after one candidate finished the final stage for the platform engineer role. Only two people assessed; the CEO's stage is recorded but not scored.

Candidate and process

Candidate reference
Candidate B, a reference only; no names, contact details or documents in this record.
Role
Platform engineer (first infrastructure hire)
Stages completed and who ran them
Stage 1 screening call and stage 3 operating interview with the CTO. Stage 2 pipeline conversation with a product engineer. Stage 4 with the CEO, not scored. References taken by the CTO with the candidate's permission.

Evidence against the scorecard

RowEvidenced / partly / not testedThe evidence itselfWho saw it
Must-have 1EvidencedRan the on-call rota at their last company and named three alerts they rewrote after a bad week, saying which changes were theirs. Both referees confirmed the on-call ownership.CTO (stage 3), references
Must-have 2EvidencedRebuilt a release process the whole engineering team used, explained what was wrong before, how people were moved onto it and what they would now do differently. On our backlog change, started from what it meant for the product engineers and changed the rollout plan when an objection was fair.Product engineer (stage 2)
Must-have 3Partly evidencedThe first outage was told in order, wrong turns included. The second went straight to the root cause, and the follow-up questions about dead ends got general answers.CTO (stage 3)
Must-have 4Partly evidencedSent a redacted runbook as the written follow-up. Clear and in order, but written for engineers who already know the system.Product engineer (stage 2 written follow-up)
Must-have 5Not usedThe scorecard has four must-haves.Not applicable

Independent recommendations

RowRecommendationThe single strongest reasonThe main reservation
Interviewer 1HireThe operating evidence was specific and first-hand, and the references matched it.Debugging under pressure is evidenced once, not twice.
Interviewer 2HireStarted from what the change meant for us, not from the tools, and listened when pushed.Argued one objection longer than it deserved before accepting it.
Interviewer 3Not scored. Stage 4 is a conversation, not an assessment.Recorded so the panel sees it: asked direct questions about runway and the next funding step.None recorded, by design.
Interviewer 4 (optional)Not used. Two people assessed this candidate.Not applicable.Not applicable.

Decision

Risks we are knowingly accepting, and the plan for each
Incident debugging is only partly evidenced. Plan: the CTO shares the first on-call rota, and the two of them write the first incident review together. The runbook style assumes expert readers. Plan: the first runbook is tested by a product engineer running a release from it.
Anything untested that we have decided not to test
Cloud cost ownership, a later scorecard outcome, was not tested. We judged it learnable in post and will review it at the end of the first quarter.
Decision, decision-maker and date
Offer. Decided by the CTO, with the CEO agreeing the package, the day after the final stage. Recorded the same day.
What the first thirty days must address as a result
A first release run from the runbook with a product engineer, the shared on-call rota, and a short note on cloud spend by the end of the first month.