A resume ATS score is a number produced by a resume checker, not by the applicant tracking system an employer runs. No ATS shows candidates a score. You upload a file to a checker's website, it compares your resume with a job description and its own rules, and it reports a percentage on a scale it designed. Some employer systems do rank applicants on their side, against criteria the recruiter set, but you never see that number and no outside tool can reproduce it. Two of the vendors that sell scores say as much in their own copy. (ApplyEze, which publishes this blog, makes a job application tool and does not sell a score.) This post covers where ATS scores come from, what the real system does with your resume, how each popular checker scores, why the same resume swings from one tool to the next, and what to check instead.
Where ATS scores come from
The phrase "ATS score" suggests a number the system assigns and the candidate can look up. That number does not exist on the candidate's side. The checkers that sell scores say so:
- Rezi's score documentation says "real ATS systems don't give your resume an official score," and "Recruiters won't see the same percentage you see in Rezi."
- Enhancv's resume checker page says "there's no such thing as an ATS score," and goes on to say that no tool online that provides a score gives an actual one.
So each score is a model of what the vendor thinks a screening process cares about. Some of those models are thoughtful, and Jobscan is unusually open about its weights. But the number measures your resume against the vendor's rules, not against the employer's.
The candidate-facing side of the big systems confirms it. Greenhouse's candidate portal shows, at most, an active or inactive status or a stage name the employer chose to share, which our post on Greenhouse application status walks through. Workday shows whichever status label the employer configured. Neither shows you a number.
What the ATS really does
An applicant tracking system stores applications, parses resumes into fields, runs the employer's rules and gives recruiters ways to search and sort. Our guide to what an ATS is covers the whole sequence. Three parts matter for scoring.
Parsing. The system extracts text from your file to fill in profile fields. When parsing fails, Greenhouse's documented consequence is that the hiring team enters details by hand. It is a cost, not a verdict.
Ranking, on the employer's side. Some systems do score you, privately. In Greenhouse's Talent Matching, a recruiter sets up a calibration of skills for the job and rates each one's importance, and candidates land in one of five categories: Strong match, Good match, Partial match, Limited match, or Needs manual review. Greenhouse says it "does not automatically advance or reject candidates," and its candidate FAQ explains how to opt out of a match score, per application, where the employer enables that. Workday's HiredScore grades candidates A to D for recruiters, and its grade definitions tie the grades to the job's qualifications: B is "Candidate meets all basic qualifications," C is "Candidate meets most but not all basic qualifications."
Both are built from the employer's own criteria for one job, and both are shown to the hiring team. A checker on the open web has seen neither the calibration nor the recruiter's ratings, so its percentage cannot be the same number.
Rules on your answers. Automatic rejection mostly lives here, not in the resume. Greenhouse's application rules include an Auto-Reject rule that "uses custom application questions to screen and reject candidates based on their responses," with examples such as a candidate who requires a visa the company cannot provide. No resume score touches those answers.
If you have read that 75% of resumes are rejected by an ATS before a human sees them, that figure has no published method behind it. Our post on the ATS rejection myth traces where it came from.
How each checker scores you
Every checker scores something different, on a scale of its own. Here are four popular ones, from each vendor's own pages as of October 2026.
| Checker | What the score is | Scale and target | What it says about real ATS scores |
|---|---|---|---|
| Jobscan | Match rate between your resume and a pasted job description | Percentage; recommends 75%, and says many users see success at 65% | Says the match rate is based on five priorities, and that word count and measurable results are not factored in |
| Rezi | Rezi Score across Content, Format, Optimization, Best Practices and Application Ready | 1 to 100; 90+ is ready to apply, 50 to 89 has room to improve, below 50 needs revising | "real ATS systems don't give your resume an official score" |
| Enhancv | 27 checks across seven categories, scored in two tiers: how much of the resume it can parse, then the issues it finds | No range stated; above 80 is "mostly good" | "there's no such thing as an ATS score" |
| Careerflow | Two parts: a keyword (skill) score against a job description, and an analysis score for structure, formatting and completeness | No scale stated in its help article | A base resume without a job description gets only the analysis score |
Jobscan's five priorities, in order, are hard skills, education level (only when an advanced degree is in the posting), job title, soft skills and other keywords. That is the most transparent model of the four, and it shows what these scores are built from: mostly word overlap with the description you pasted.
What you pay for the number varies. As of October 2026, Jobscan's plan page lists a free tier with 5 resume scans a month, $49.95 a month, or $89.95 every three months. Rezi's pricing page lists a free plan, Pro at $29 a month and Lifetime at $149 once. Our Jobscan review looks at what its match rate does and does not tell you, and our Resume Worded review does the same for another checker's score, its plans and its refund rules.
Why scores swing
People notice the swings first. One Reddit user wrote that they scored 98% on Jobscan and it did nothing for them. Another, after a score of 30, asked whether the tool was really supposed to work that way. A third thread asked how legit Jobscan is because the score seemed impossible to beat. None of this means the tools are broken. It follows from how they are built.
- Every tool measures something different. A keyword match rate, a 100 point quality score and a 27 check audit will not agree, because they are not measuring the same thing. Moving between tools changes the ruler, not the resume.
- The job description is half the input. A keyword score is overlap with one posting. Paste a different posting and the same resume gets a different number. That is correct behavior, and it is also why a single "ATS score" for a resume means little.
- Tools disagree about what counts as a match. Jobscan sells an "AI Keyword Synonym Detector" on paid plans, so the same wording can score differently by plan. Greenhouse's own Talent Matching FAQ says it "compares related terms on each resume to the calibrated skills you set," and that "a longer list of matched terms doesn't always mean a higher match score." A checker that rewards every extra matched word is rewarding something the employer's tool may not.
- Each checker runs its own parser. If a checker misreads a two column layout, it scores text that may not match what the employer's system extracted, in either direction.
- Overlap is not proof. Jobscan says word count and measurable results are not part of its match rate. A resume can list every term in the posting, prove none of them, and score in the 90s. That is the 98% with nothing to show.
- High scores have a ceiling built in. Jobscan warns that scoring above 75% may not be possible "without overstuffing your resume with keywords." Chasing the last points pushes a resume toward stuffing, which a human reader sees at once.
What to check instead
The useful parts of a score are things you can check yourself in minutes, against the posting you are actually applying to.
| The score claims to measure | Check this instead |
|---|---|
| ATS readability | Open your PDF, select all, copy, and paste into a plain text editor. If the text reads in order with your name first and jobs in sequence, a parser can read it. Our ATS friendly resume format post covers the failures to look for. |
| Keyword match | List the posting's required hard skills. Each one you have should appear in the posting's spelling and be proven in a bullet. Our post on hard skills for a resume works through this. |
| Job title fit | Does your most recent title, or your summary, tell a reader in one line that you do this job? Jobscan ranks job title third in its own weights. |
| Qualifications | Count the basic qualifications in the posting and mark which you meet. Workday's HiredScore grades turn on exactly this split between basic and preferred. |
| Auto-reject risk | Answer the screening questions slowly. Work authorization, sponsorship, location, licenses and minimum years are where rules act. |
| Recruiter appeal | Read the top third of the page as if you had 30 applicants to get through. Is the match obvious without hunting? |
There is a better reason to match a posting's words than any percentage. In the 2021 Harvard Business School and Accenture study Hidden Workers: Untapped Talent, 88% of employers believed qualified high-skills candidates were screened out because they did not match the exact criteria in the job description, and 94% believed it for middle-skills workers. That is employer belief from a survey, not a rejection rate, but it is the honest case for using the posting's language for work you have done.
Checkers are not useless. A keyword list from a pasted job description can show you a required skill you forgot to mention. Use the list and ignore the number.
Where ApplyEze stands
ApplyEze, which publishes this blog and makes a job application tool, does not sell an ATS score, because no ATS exposes one. It writes a resume for each posting and a cover letter when the form asks for one, as PDF and DOCX, and reads each file back with a parser check before upload, a check on the same failures the copy and paste test above looks for. It brings the posting's named skills and tools onto your skills line, even when your own record does not list them. There is no step to remove one before a live run, so use a dry run to see what it writes, and be ready to discuss anything on the page. A certification the posting asks for is not added unless you hold it. Your employers, job titles, dates, schools, degrees and certifications come from your own record, and code rejects a rewrite that adds a figure, employer, school or credential your record does not hold. If you tailor by hand, list only skills you can discuss in an interview and read the tailored file before it goes out.