Published on

How ATS keyword matching actually works (and how to read your match %)

Authors

There is a lot of folklore about applicant tracking systems (ATS): that they auto-reject 75% of résumés, that they choke on two-column layouts, that a hidden keyword in white text will sneak you through. Most of it is wrong. The useful mental model is simpler:

An ATS is a search engine, and the recruiter is the person typing the query.

What the machine actually does

When a job is posted, the ATS parses your résumé into fields (work history, skills, education) and indexes the text. A recruiter then searches that index — usually for the concrete skills in the job description: Kubernetes, Series A, HL7, Spanish, whatever the role needs. Résumés that contain those terms, in context, surface first.

That is why keyword coverage matters. Not stuffing — coverage. If the job description leans on "incident response" and "on-call rotations" and your résumé says "handled production issues," a keyword search for the former will not find you, even though you did exactly that work.

Reading your match percentage

JobHakken's ATS match % is an estimate of how well your résumé's language covers the terms a given posting emphasizes. It is not a grade on your career; it is a measure of overlap between two documents. A few ways to read it:

  • High (70%+). Your résumé already speaks this job's language. Apply, and spend your effort on the parts a keyword match can't see — the cover note, the referral.
  • Middling (50–69%). You are a real fit but phrased differently. This is the sweet spot for tailoring: adopt the posting's exact terms for skills you genuinely have.
  • Low (<50%). Either a stretch role, or your résumé and this job describe the same work in totally different vocabularies. Worth a look at which keywords are missing before you decide.

The number is a starting point for a decision, not the decision itself.

Turning the score into edits

The move is not "add every missing word." It is:

  1. Look at the keywords the posting weights most and that you are missing.
  2. Keep the ones you can back up with real experience.
  3. Rephrase existing bullets to use the posting's terms — same accomplishment, matched vocabulary.

When Jordan Rivera's résumé said "built data pipelines" and the job kept saying "ETL," the fix was one word, not a rewrite. The match jumped because the search could finally find the work that was already there.

That is the whole game: make it easy for the search engine to find the true things about you. JobHakken shows you the gap so you can close it honestly.