Interview scorecard template, with a scale that holds up

How to use the scorecard

1. Agree the criteria before the first interview. The hiring manager and the recruiter pick three to six things the role needs and write what a 3 and a 1 look like for each. Writing the anchors is most of the work; the rest is filling in boxes.
2. Ask the knock-outs first. They are pass or fail, and a fail ends the screen. A candidate who cannot work the shift is done in three minutes, and so are you.
3. Score during the interview, not after. Note the evidence as you hear it: a phrase, a number, a step they described. Scores written an hour later drift toward a general impression.
4. Write one line of evidence per score. “Explained the refund policy back to a confused customer in two sentences” is evidence. “Good communicator” is not.
5. Work out the total and read it against the bands. Multiply each score by its weight, add them up and divide by 100. That gives a number from 0 to 3.
6. Submit before the debrief. Each interviewer sends their scorecard before seeing anyone else’s. Then the panel talks about the criteria where they differ.

A worked example

Two candidates for a customer support role, scored on the template above:

Candidate A: 3 × 30 + 2 × 30 + 2 × 20 + 1 × 20 = 210, divided by 100 is 2.1. Candidate B: 2 × 30 + 3 × 30 + 3 × 20 + 2 × 20 = 250, so 2.5.

A had more directly relevant experience, but B handled the actual support case better and explained it more clearly. Without the scorecard, the interviewer who liked A’s resume and the interviewer who liked B’s call would argue about impressions. With it, they are looking at the same four numbers and the evidence line behind each.

A third candidate scored 3, 3, 3, 3 but said they cannot work weekend shifts, which the role requires. That is a No, whatever the total. Knock-outs are not averaged.

Choosing a rating scale

Use a short scale and describe every point. The template uses 0 to 3:

• 0: no evidence, or could not do it
• 1: weak evidence
• 2: meets the bar
• 3: clearly strong, with specifics

Four points is enough to separate candidates and few enough that two interviewers will usually agree. The descriptions are what make it work. The US Office of Personnel Management recommends structured interviews in which every answer is rated against a written standard; a scale with no descriptions is just a number for how the interviewer felt.

Avoid 1 to 10 without anchors. Scores tend to bunch between 6 and 8, and the scorecard stops telling candidates apart. Avoid an odd number of points too: a middle option is where tired interviewers hide.

How many criteria, and how to weight them

Three to six criteria per interview. We learned the lower limit the hard way. On one role our scoring rubric had a single criterion scored 1 to 3, and the results only ever took three values: 33%, 67% and 100%. A whole applicant pool bunched into three groups, and no threshold could separate them. A scorecard with one row can only produce as many totals as it has points on the scale. Adding criteria is what spreads candidates out.

Weights say which criteria matter most for this role. The template gives experience and the core skill 30 each and the other two 20 each; a support role might weight communication higher, an engineering role the skill check. Keep weights in round numbers that add up to 100, so anyone can check the arithmetic.

Two rules keep weights honest:

• Must-haves are knock-outs, not heavy weights. If a candidate must hold a license, ask it as a pass or fail question. A heavily weighted criterion can still be outscored by everything else.
• Never score “culture fit”. It cannot be anchored to the job, and it is where bias enters. If you mean something specific, such as “works well across time zones”, write that as the criterion.

Screening scorecard vs interview scorecard

A screening scorecard is for the first round, usually a 10 to 20 minute call or AI interview. Its job is to decide who moves on, so it is short: the knock-outs, three or four criteria, and the logistics. Our guide to pre-screening interview questions has the questions that go with it, with role versions for customer service and inside sales.

An interview scorecard is for a later round with the hiring manager or a panel. It goes deeper on fewer areas, with harder questions and follow-ups. The interview questions to ask candidates list gives each question a follow-up that separates real experience from a rehearsed answer.

The same template works for both. Change the criteria, keep the scale, the evidence line and the decision rule.

Mistakes that break a scorecard

• Scoring after the debrief. The scores then record the group’s opinion, not each interviewer’s.
• Averaging away a failed must-have. A strong total does not cancel a shift the candidate cannot work.
• Comparing candidates instead of scoring against the anchors. “Better than the last one” is not a 3. Score each candidate against the description.
• Leaving the evidence column empty. Without it, a hiring manager cannot check the score, and the feedback you send has nothing to stand on.
• Never checking the scorecard against who you hired. If candidates who scored Hold keep turning into your best hires, the anchors are wrong. Revisit them after every few hires.

How HireQwik fills in the scorecard for the first round

HireQwik runs the first round as a 10 to 20 minute AI interview, and its scorecard follows the same structure as the template above. The criteria and anchors come from your job description, not a generic list, and your team can edit them before anyone is interviewed.

Knock-out questions are asked first, and a fail is recorded with the candidate’s own words. The core questions are then scored against written rubric levels: on our points rubric, the three core questions are worth 3, 3 and 2 points, communication up to 2, and a small bonus for domain depth, capped at 10. A candidate who answered fewer than two of the core questions is marked Incomplete rather than scored low, because an unfinished call is not evidence of a weak candidate.

The result is a verdict of Strong Go, Go, On Hold or No Go, with the evidence behind each score. A recruiter can override it, and can share the scorecard with a hiring manager as a link that needs no login. When the team records who was actually hired, HireQwik compares that with the verdicts for each role, which is the last mistake on the list above, fixed.