What is an explained match score?
A match score measures the gap between a profile and a job, usually as a percentage. It is "explained" when it comes with its reasons: which required skills are present, which are missing, how experience compares with what the role requires.
It is the opposite of a "black box" score, which gives a number without saying where it comes from. An unexplained 82% asks the recruiter to take the tool's word for it; an 82% with "7 of 9 skills, 5 years of experience for 3 required, SQL missing" gives them something to decide with.
Why explain a match score?
Candidate matching is there to sort faster, not to decide for the recruiter. Without an explanation, the score becomes a truth nobody can check or challenge. A strong candidate can be pushed down because of an unrecognized synonym, and nobody notices.
The explanation also protects the candidate. If a criterion unrelated to the job influences the score, the breakdown makes it visible. That is the principle of explainable AI: a recommendation whose reasoning you can follow.
The legal framework points the same way. In Quebec, section 12.1 of Law 25 requires informing a person when a decision about them is based exclusively on automated processing, and letting them learn the main factors behind it. In Ontario, since 2026, public job postings from employers with 25 or more employees must disclose whether AI is used to screen, assess or select applicants.
Concrete example
A recruiter in Halifax posts a data analyst role. The tool ranks 40 candidates from her database. At the top, an 88% profile: Python, SQL and Power BI present, 4 years of experience for 3 required. Further down, a 61% profile: the explanation shows the resume mentions "statistical analysis in R" but not "Python".
Reading the breakdown, the recruiter decides R is transferable for this role and calls the candidate anyway. The score saved her time on the other 38; the explanation kept her from wrongly passing on someone.
Best practices
- Ask for the per-criterion breakdown before trusting a score: skills covered, skills missing, experience.
- Read the explanation for profiles just below your cut-off. That is where false negatives hide.
- Keep the decision human. The score orders a list; it does not reject anyone.
- Only share job-related factors with candidates: skills and experience, never a criterion unrelated to the work.
- Keep a record of recommendations and the decisions taken.
For the Quebec framework, read our guide to Law 25 in recruitment.
Common mistakes
- Setting an automatic rejection threshold. A threshold does not read explanations; a recruiter does.
- Showing the score to the hiring manager without the breakdown. The number alone then carries too much weight in the decision.
- Confusing precision with accuracy. A 78.4% is no more correct than a 78%: a heuristic score should be shown as a whole number.
- Never checking the tool's mistakes. Review a sample of low-ranked profiles now and then to see what the score misses.
With RecruitEasy
RecruitEasy never shows a score without its explanation: every match analysis shows the skills covered and missing and how experience compares with the role. The feedback sent to candidates only covers skills and experience, and the recruiter approves it before it goes out. See the matching feature.