Reports

Understanding candidate scores

How Hirevalu scores technical, communication and problem-solving performance, and what the overall score and hiring recommendation actually mean.

Updated

Every completed interview produces a report with four performance metrics, a set of criteria outcomes, a hiring recommendation, and the full transcript. This page explains what each one is derived from and how much weight to put on it.

The four performance metrics

Each is reported as a percentage.

Technical Experience

Role-specific skill demonstrated in the conversation, measured against the job description attached to the position. Because it is scored relative to the JD, the same candidate can score differently across two positions — this is intended. A backend engineer interviewing for a platform role and an application role should not score identically.

Communication

Clarity, articulation and engagement. Scored on how the candidate constructs and delivers answers, not on accent or fluency.

For interviews conducted in a candidate’s second language, read this metric alongside the transcript rather than on its own.

Problem Solving

How the candidate approaches problems they have not seen before: whether they clarify before answering, structure their reasoning, and adapt when given new information. This is the metric that most benefits from a longer interview — at 10 minutes there is often not enough room to observe it properly.

Overall Score

A combined assessment across the other three. Treat it as a sorting key, not a verdict — two candidates at 78% can be strong in completely different areas. The report’s ranking and score distribution show where a candidate sits relative to everyone else who interviewed for the position, which is usually more actionable than the absolute number.

Evaluation criteria outcomes

Separate from the percentage metrics, each criterion you defined on the position gets one of four outcomes:

OutcomeMeaning
MetClear supporting evidence in the interview
Partially metSome evidence, incomplete or qualified
Not metEvidence the candidate does not meet it
Not coveredThe interview produced no evidence either way

Not covered is not a negative signal about the candidate. It means the conversation never reached that ground — usually because the interview was too short for the number of criteria, or the criterion was written in a way that no answer could confirm. Persistent Not covered rows on the same criterion mean the criterion needs rewriting or the interview needs to be longer.

The report summarises this as a met count, which counts Met and Partially met together.

Hiring recommendation

A single label combining the metrics and criteria outcomes:

  • Very Strong
  • Strong
  • Moderate
  • Less

This is a summary for triage, not a decision. It has no knowledge of your budget, team composition, or how this candidate compares to others in your pipeline. Use it to order your review queue.

How to actually read a report

A workable order:

  1. Check completion status. Incomplete interviews are not comparable.
  2. Check the integrity result. A flag changes how you read everything else. See interview integrity.
  3. Read the high-importance criteria. These are your real filters.
  4. Look at the metrics for shape, not rank. Where is this person strong?
  5. Read the transcript for anyone you are seriously considering. The scores tell you where to look; the transcript is the actual evidence.

That last step matters more than it sounds. The scores exist to make 200 candidates triageable. For the eight people you actually advance, read what they said.

What the scores are not

  • Not a personality assessment. Hirevalu measures demonstrated performance in a conversation, not traits.
  • Not a hiring decision. Nothing is auto-rejected. A human advances or declines every candidate.
  • Not comparable across positions. Scores are relative to a specific job description and criteria set.