How the ATS readiness score is calculated
Eight categories, 100 points, every weight published. A score you cannot interrogate is a number designed to make you buy something — so here is ours, including the parts that make us look worse.
The weights
| Category | Points | What it measures |
|---|---|---|
| Parseability | 20 | Whether the text a parser pulls out of the finished PDF matches the document you wrote — same content, same order. |
| Bullet quality | 16 | How the achievement lines read: length, whether they start with an action, and whether they say what happened rather than what you were responsible for. |
| Contact completeness | 12 | Whether a recruiter and a parser can both find who you are and how to reach you. |
| Section structure | 12 | Whether your sections use the names parsers are calibrated for, in the order they expect. |
| Quantification | 12 | How much of your experience is stated in numbers rather than adjectives. |
| Dates | 10 | Whether every entry is dated in a form a machine can read. |
| Length & density | 10 | Page count against how much experience you have, and how densely the page is filled. |
| Skills coverage | 8 | Whether the skills section is specific enough to be searched. |
What earns and loses each category
Parseability — 20 points
Whether the text a parser pulls out of the finished PDF matches the document you wrote — same content, same order.
- Earns: Every section, employer, date and bullet extracting in the order a reader sees them.
- Loses: Text missing from the extraction, sections arriving out of order, or a file with no text layer at all. This is the largest single category because it is the only failure that can make everything else irrelevant.
Bullet quality — 16 points
How the achievement lines read: length, whether they start with an action, and whether they say what happened rather than what you were responsible for.
- Earns: Bullets that open with a verb, run one to two lines, and describe a result.
- Loses: First-person phrasing, bullets over about 30 words, single-word fragments, and lines that describe duties instead of outcomes.
Contact completeness — 12 points
Whether a recruiter and a parser can both find who you are and how to reach you.
- Earns: Name, email, phone, location, and at least one professional link.
- Loses: A missing email or phone, contact details buried in a header or footer, or a link that is not a URL.
Section structure — 12 points
Whether your sections use the names parsers are calibrated for, in the order they expect.
- Earns: Summary, Experience, Education and Skills present, conventionally named, canonically ordered.
- Loses: Invented section names, a missing Experience or Education section, or an order that puts skills ahead of experience.
Quantification — 12 points
How much of your experience is stated in numbers rather than adjectives.
- Earns: Bullets carrying a figure: a percentage, a count, a duration, an amount.
- Loses: A resume where nothing is measurable. Partial credit is proportional — this is a share of your bullets, not a threshold.
Dates — 10 points
Whether every entry is dated in a form a machine can read.
- Earns: Month and year on every role and qualification, consistently.
- Loses: Missing end dates, year-only ranges, mixed formats, and ranges that run backwards.
Length & density — 10 points
Page count against how much experience you have, and how densely the page is filled.
- Earns: One page for most people; two once there is enough relevant work to fill them.
- Loses: A resume that spills a few lines onto a second page, or a page so sparse it reads as unfinished.
Skills coverage — 8 points
Whether the skills section is specific enough to be searched.
- Earns: Named, grouped, concrete technologies and practices.
- Loses: Generic traits, a handful of entries, or a list so long that nothing in it is a signal.
Why parseability is worth the most
Because it is the only category whose failure makes the others meaningless. A resume with excellent bullets that extracts in the wrong order, or extracts as nothing at all, is a resume the system never read. Everything else on this page is an improvement to a document that is already being read.
It is also the half most tools cannot measure. Scoring text you pasted into a box tells you nothing about your file — and your file is the thing that might be broken. What an ATS reads from your PDF shows the extraction from one of our own templates, including an artifact we cannot fix.
Two honest limitations
- This is not a prediction of whether you get an interview. It measures whether your document is readable, complete and specific. A 94 on a resume applying for the wrong job is still the wrong application.
- Templates with a side column give up one guarantee, and the score says so. A PDF’s text comes out page by page, so on a two-page resume a side column’s text lands after whatever fitted on page one. Those templates are scored against same-content rather than same-text, and the report prints a zero-point note naming it. See two-column resumes and ATS.
The version that runs on an upload
The free checker cannot compare your PDF against a document, because a stranger’s upload has no structured version of itself — so it measures the same eight categories from the extracted text alone, using the same rules for action verbs, first-person phrasing and quantification. Same rubric, one less input.