One sheet per interviewer, per interview.
An interview scorecard records what a single interviewer learned about a candidate against criteria the team agreed before the interview. Each interviewer fills in their own copy, so their evidence is captured before the group discussion.
The criteria and questions in the file are examples for a general role. Replace them with requirements from your own job description before using the scorecard with a candidate.
Score what the role needs, not a general impression.
Start from the job description. List the few things a person must be able to do in the role, then turn each into a criterion an interviewer can observe in a conversation.
- Keep the list short. Four to six criteria per interview is usually enough. A long list spreads attention thin and the ratings become guesses.
- Pair each criterion with a question. Ask every candidate the same opening question for a criterion, then follow up on what they say.
- Write down what strong evidence looks like. An evidence anchor such as “a specific example with the candidate’s own contribution and the result” helps two interviewers rate the same answer the same way.
- Split criteria across the panel. When several people interview, give each person a defined area rather than asking everyone to cover everything.
Give every number a meaning.
A rating only compares across interviewers when everyone reads the scale the same way. The template uses a 1–5 scale with a written meaning for each point:
| Rating | Meaning |
|---|---|
| 1 | No evidence or a clear concern |
| 2 | Limited evidence; important gaps remain |
| 3 | Meets the requirement for this role |
| 4 | Strong evidence with specific examples |
| 5 | Exceptional evidence across the whole question |
Leave a rating blank when the interview did not cover that criterion. A blank is not a low score: it stays out of the weighted score entirely.
A weighted score is a summary, not a decision.
Weights (1–5) say how much each criterion matters for this role. Weighted score = the sum of weight × rating, divided by the sum of the weights of the criteria that were rated. This fictional example uses the six criteria in the template:
| Criterion | Weight | Rating | Weight × rating |
|---|---|---|---|
| Core skill for the role | 5 | 4 | 20 |
| Problem solving | 4 | 3 | 12 |
| Collaboration | 3 | 4 | 12 |
| Ownership | 3 | 2 | 6 |
| Communication | 3 | 4 | 12 |
| Interest in the role | 2 | 3 | 6 |
The result is 68 / 20 = 3.4. That is close to “meets the requirement” overall, but the single 2 on ownership is the more useful signal: the panel should look at the evidence notes behind it before deciding whether it matters for this role.
Collect the sheets before anyone talks.
The value of a scorecard is independent evidence. Ask every interviewer to submit their ratings and notes before the debrief, then discuss the criteria where ratings differ most. Record the final recommendation as Proceed, KIV (keep in view) or Do not proceed, together with the reason.
A structured scorecard makes the evidence easier to compare. It does not make weak criteria fair or guarantee an unbiased result, and the hiring team remains responsible for the decision.
The same structure, attached to the candidate.
In TiLab ATS, scorecard templates are organised into categories and questions. A question can be a 1–5 score, a single or multiple choice, or a written description, with an optional comment box. Every job requires a scorecard, and the interviews for that job inherit it.
Each interviewer submits their own scorecard. In weighted mode the verdict follows from the weighted average: by default Proceed from 3.5 and KIV from 2.5, with thresholds you can change. In manual mode the interviewer chooses the verdict. The candidate’s record then shows how many panelists have submitted and whether their decisions are split.
See how this fits the wider interview workflow on the structured interviews page.