Methodology
How the resume score is calculated
Nine weighted categories, some measured by code and some judged by an AI pass. Here is the whole method, the actual weights, and the one thing we deliberately refuse to put a number on.
Two engines, deliberately
Some of what makes a resume good can be measured, and some of it has to be read. Length, contact details, skills coverage, and formatting are checked by ordinary code, so they return the same answer every time. Whether a bullet shows a real result, whether your summary earns its space, whether the wording is doing any work: those need judgment, and an AI pass handles them. Both feed one 0-100 score.
What the code measures, and what it is worth
ATS-friendly format is 10 points, skills 7, length 6, and contact details 5. These are the boring categories, and boring is the point: they never drift between runs, so if your score moves you know the content moved rather than the scorer having a different day. They are also deliberately light, because nearly every finished resume scores full marks on all four, and a score built mostly out of things everyone gets right cannot tell you anything useful.
What the AI judges, and what it is worth
Measurable results is the heaviest category in the whole score at 22 points, because it is the thing that most separates a resume that gets read from one that gets skimmed. Word choice is 16, strong summary 12, spelling and grammar 12. If you have attached a target job, match to the job adds a further 10.
One category only appears if it applies
If a resume lists no jobs and no projects at all, a work history category worth 15 points is added and scored low, with a note suggesting a project, coursework, volunteering, or an internship. It is the only conditional category: on any resume that has either one, it is not part of the total. That is on purpose, so a category about a gap most people do not have cannot quietly move everyone else's score.
Where the two disagree, the code wins
A few categories could be reached from either direction. When they collide, the deterministic result takes precedence and the AI's opinion is discarded. Language models are not reliably consistent on the same input, and a score that wobbles when nothing changed is worse than a score that is occasionally strict.
Re-checking is anchored, not re-rolled
When you re-check a resume, the previous report goes back in with it, and each category is expected to hold unless that section genuinely changed. Without that anchor an unchanged resume would score slightly differently every time, which would teach you to re-roll instead of to edit.
The ATS part is a checklist, never a number
This is the deliberate one. Plenty of tools print an ATS score out of 100. We show pass, attention, or a note across the things that actually decide whether a parser can read your resume: single-column reading order, recognizable section headings, parseable dates, clean text, and a keyword match once you have added a target job.
Why we publish this
Most tools that score your resume will not tell you how. That makes the number hard to argue with and hard to trust, which is a bad combination when you are about to change your resume because of it. So here is the whole method, including the weights.
The short version: a 0-100 score built from nine categories, plus a tenth that applies only if your resume lists no jobs and no projects. Some are measured by code and some judged by an AI pass, merged into one number with a per-category breakdown and a list of specific fixable issues. Nothing about it is proprietary enough to be worth hiding, and you can argue with any part of it once you can see it.
Questions about scoring
- How is the resume score calculated?
- Nine weighted categories out of 100. Measured by code: ATS-friendly format (10), skills (7), length (6), contact info (5). Judged by an AI pass: measurable results (22), word choice (16), strong summary (12), spelling and grammar (12), and match to the job (10) when you have attached a job description. A tenth, work history (15), is measured by code but added only when a resume lists no jobs and no projects at all. They merge into one score, and where a category could come from both, the code result wins.
- Why is there no ATS score out of 100?
- Because it would be invented. Every applicant tracking system parses differently, so no tool can score you against all of them, and a made-up number invites you to optimize for it. You get a checklist instead. In testing, a layout-driven ATS score also scared people away from two-column designs that parse perfectly well, which is real harm in exchange for a false precision.
- Is the score the same every time I run it?
- The code-measured categories are, by construction. The AI-judged ones are anchored: a re-check receives the previous report and holds each category steady unless that part of the resume actually changed. So an unchanged resume should not drift.
- Which category should I fix first?
- Whichever heavy one you are weakest on. Measurable results is worth 22 points and is the most common weak spot, usually because bullets describe duties rather than outcomes. Word choice at 16 is next.
- What goes into the ATS-friendly format category?
- Three signals, weighted, and all three are about your content rather than your design. Recognizable section headings carry half of it, because a heading a parser cannot place is what actually loses your work history. Parseable experience dates are 30 percent and parse-safe characters the remaining 20 percent. The template you pick is deliberately not scored: which design reads best for an applicant tracking system is shown in the ATS check as a note, never as a deduction, because choosing a two-column layout is a preference and not a mistake.
- Does a high score mean I will get interviews?
- No, and any tool implying otherwise is selling you something. The score measures whether your resume is well built: specific, readable, parseable, and matched to the job. Whether you get called depends on the role, the competition, and timing. It is a useful second opinion on the document, not a prediction about the outcome.
- Does it cost anything to get a score?
- No. Importing a resume returns a score at no cost, and re-checking is free too. The keyword match against a job description is also free. Exporting the finished PDF is the only part with a one-off cost, and there is no subscription.
More on pricing and exports is in the general FAQ.