How ATS actually works.

ATS is not a mystery

Applicant Tracking Systems (ATS) sound scary because most articles about them are vague. “Use keywords.” “Avoid tables.” “Don't use headers.” Great — but why?

The truth: ATS is boring software. It parses text. It matches keywords. It ranks. That's it. What makes it feel opaque is that most tools score you with an AI guess instead of the actual rule set.

Deterministic vs guessed scoring

A deterministic score is reproducible. Same resume + same job description = same score, every time. Every lost point traces to a specific rule.

A guessed score is not reproducible. Two runs give different numbers. You can't improve what you can't trace.

We use deterministic scoring because your career deserves better than a slot machine.

The six rule families

1. Parseability — can the ATS read your file cleanly?

2. Keyword coverage — do your terms match the JD, in the right sections?

3. Section structure — does the ATS find the sections it expects?

4. Skills fidelity — do your skills appear with context, not just as a list?

5. Metric density — do accomplishments carry numbers?

6. Recency & relevance — is the most relevant work weighted highest?

How to optimise, in order

Fix parseability first — no images-of-text, no exotic fonts, no columns that break.

Then close keyword gaps — pull them from the JD, land them in the right section.

Then tighten metrics — numbers beat adjectives every time.

Then re-score. If your rewrite scored lower, you know exactly which rule broke.