Skip to main content

Understanding Scores

Ezra scores every candidate automatically after their interview. Here's what the scores mean and how to interpret them.

Score scale

All scores are on a 1 to 4 scale:

ScoreMeaning
4Excellent — response closely matches the ideal answer
3Good — solid response, misses some key elements
2Fair — acceptable but lacks depth
1Poor — response does not meet expectations

Overall score

The overall score is a weighted average of the question scores. The weighting is determined by the Importance level (1–5) you set for each question during training.

For example, if a question has Importance 5 and another has Importance 1, the first question has five times the impact on the overall score.

Question scores

Each question is scored based on how closely the candidate's response — and any follow-up answers — match the rubric you set during training (Great / Good / OK / Poor answer definitions) combined with the ideal candidate profile.

What Ezra does NOT score

The overall score is based only on the content of the candidate's answers. The following are surfaced as separate signals and do not affect the overall score:

  • Communication analysis (sentiment, communication style)
  • Cheat detection / fraud detection findings
  • Strengths and opportunities for growth
Ezra does not make hiring decisions

All decisions about whether to proceed with a candidate are made by the human recruiting team. Ezra's scores are one input among many.

Strengths and opportunities for growth

Below the scores, Ezra provides a written summary of the candidate's strengths and areas for development — evaluated relative to the job description, role requirements, ideal candidate profile, and question rubrics.

Editing a score

If you disagree with Ezra's score on a question, you can change it:

  1. Click Edit score next to the question
  2. Select the score you believe is more accurate
  3. Optionally explain why you disagree

Changing the score updates only that candidate's record. However, when you explain your reasoning, Ezra uses your feedback to update the evaluation rubric — improving how it scores future candidates for that interview. This is a powerful way to calibrate Ezra to your specific standards over time.