I build an interview scorecard before sourcing starts, not after the first impressive candidate arrives. The scorecard turns a broad job description into a small set of capabilities the panel can assess consistently. It does not predict performance by itself. It gives the team a shared way to collect and discuss evidence.
Start with the work
I ask the hiring manager to name the outcomes expected in the first months of the role. Then I group those outcomes into four to six competencies. A useful competency describes something a person must do, such as prioritizes customer issues under changing information. It should not describe an identity, personality, or preferred communication style.
| Competency | Definition | Evidence to seek |
|---|---|---|
| Prioritization | Chooses what to do first when demands compete | Explains tradeoffs, urgency, and decision criteria |
| Ownership | Moves work forward and follows through | Names actions, decisions, and measurable deliverables without overstating results |
| Technical judgment | Selects an appropriate solution for the context | Compares options, constraints, risks, and validation steps |
| Collaboration | Makes progress with people who have different goals | Describes alignment, disagreement, and communication choices |
I keep required skills separate from useful extras so an optional, trainable skill does not lower the score.
Use anchored ratings
Each competency needs a definition and rating anchors. I use a four-point scale so the interviewer has to choose whether the evidence is below, at, or above the requirement.
| Rating | Anchor |
|---|---|
| 4 | Evidence is clearly above the level required for this role. |
| 3 | Evidence meets the level required for this role. |
| 2 | Evidence is mixed, incomplete, or below the current requirement. |
| 1 | Evidence is substantially below the requirement or was not demonstrated. |
| N/O | The interviewer did not assess this competency. |
I add a specific anchor for specialized competencies. For example, a 3 for prioritization might mean the candidate explains how they balanced customer impact, deadlines, and capacity. Without that detail, interviewers may score confidence or similarity instead of capability.
Assign ownership by interview stage
The panel should not ask every interviewer to assess everything. I map each competency to a stage and interviewer, then give each person two or three core questions. This reduces duplicate conversations and makes an N/O rating meaningful.
| Stage | Primary owner | Competencies | Output |
|---|---|---|---|
| Recruiter screen | Recruiter | Motivation, logistics, basic role alignment | Confirmed requirements and candidate questions |
| Hiring manager interview | Hiring manager | Ownership, prioritization | Evidence notes and ratings |
| Skills interview | Subject matter interviewer | Role-specific skill, judgment | Work sample or scenario rating |
| Panel interview | Cross-functional panel | Collaboration, communication | Independent ratings before debrief |
The interviewer guide includes the question, a follow-up prompt, examples of evidence, and what the interviewer must not infer. I provide structure without forcing identical conversations.
Separate score from recommendation
A scorecard can show a pattern, but I do not let an average automatically make the hiring decision. A candidate may meet every core requirement while having one skill that can be learned, while another may sound strong but lack evidence on a critical competency.
I ask for both:
- A rating and evidence for each assigned competency
- An overall recommendation with the most important reason
The hiring manager owns the final decision. If the team overrides a score, I record the job-related reason. That creates a decision trail and helps reveal whether the rubric needs improvement.
Scorecard template
INTERVIEW SCORECARD
Candidate:
Role:
Interviewer:
Interview stage:
Date:
Competency:
Definition:
Question or exercise:
Evidence observed:
Rating: 1 / 2 / 3 / 4 / N/O
Evidence-based rationale:
Competency:
Definition:
Question or exercise:
Evidence observed:
Rating: 1 / 2 / 3 / 4 / N/O
Evidence-based rationale:
Most convincing evidence:
Most important gap or risk:
What should be tested next, if anything?
Overall recommendation: Strong advance / Advance / Hold / Do not advance
Final decision owner:
Decision rationale:
Calibrate before launch
Before using the scorecard, I run a short panel calibration with sample answers. We discuss what earns each rating and whether any criterion duplicates another or is impossible to observe. After several interviews, I compare notes and adjust unclear anchors.
The best scorecard is not the longest one. It is the one that keeps interviewers focused on the work, gives candidates a consistent assessment, and leaves the hiring team with evidence it can explain.