Candidate Scorecard

What a scorecard contains

A well-designed scorecard for one interview stage contains:

1. The competencies being assessed. Each interview stage should assess 2–4 competencies, not the full list for the role. Competencies are defined at the role level (e.g., “Strategic thinking,” “Technical depth in Python,” “Cross-functional communication”).

2. Definitions of what “strong” looks like. For each competency: a written description of what a strong, adequate, and weak response looks like. Without this, two interviewers have no shared frame of reference.

3. Rating scale. Most tools use: Strong, Inclined, Not Inclined, Strong No. Some use numeric scales (1–5). The key is consistency — the same scale, used by all interviewers for the same role.

4. Space for specific evidence. Scorecards should capture what the candidate actually said, not a summary judgment. “Candidate solved the system design problem by starting with requirements rather than implementation” is evidence. “Seemed smart” is not.

5. Overall recommendation. Hire, No Hire, Strong Hire, or a numeric equivalent.

Scorecard quality in major ATS tools

Greenhouse: Best-in-class scorecard builder. Competency-level definitions visible to interviewers at scoring time. Required fields configurable. Scores are hidden from other interviewers until all are submitted (eliminating anchoring bias). EEOC-compliant note anonymisation is native.

Ashby: Good scorecard functionality with strong analytics on scorecard data (interviewer quality scoring, correlation with post-hire performance). Less mature on competency definition depth than Greenhouse.

Workable: Basic scorecard — star ratings and free-text notes. No competency definition capability. Functional for small teams but doesn’t enable structured interviewing at Greenhouse’s depth.

Lever: Adequate scorecard — similar to Workable. Lever’s strength is CRM/sourcing, not scorecard depth.

Breezy HR, JazzHR: Basic rating systems. Not designed for structured interviewing.

Common scorecard mistakes

Assessing too many competencies per stage. Interviewers who assess 8 competencies in a 45-minute interview assess none of them well. Limit each stage to 2–3 competencies maximum.

Generic competencies without definitions. “Culture fit” as a scorecard criterion is not a competency — it’s a container for bias. Replace with specific observable behaviours: “Demonstrates ability to collaborate across disciplines,” with a defined strong/adequate/weak description.

Submitting scores after the debrief. If interviewers discuss before scoring, the second person’s score is measuring the first person’s opinion, not the candidate. Enforce independent scoring before debrief.

Inconsistent use across candidates. If scorecards are used for some candidates but not others in the same pipeline, the data is useless for comparison and creates legal exposure.

Legal implications

Scorecard data is discoverable in employment discrimination cases. A well-completed scorecard (specific evidence, consistent criteria, structured assessment) is a defence. A poorly completed scorecard (“doesn’t seem like a culture fit,” inconsistent criteria use) is a liability.

Best practice: treat every scorecard as if it will be reviewed by an employment attorney — because it might be.