Resume ATS Score: What the Number Means and What Employers See

TL;DR: A resume ATS score is a number a checker tool calculates, not something the employer's applicant tracking system shows you. Each tool uses its own formula, so one resume can score 74 on one check and 96 on another. On the employer side, Greenhouse shows recruiters match categories and Workday HiredScore shows letter grades. Use a score as a list of fixes: clear parse errors first, then real keyword gaps for the specific job.

What is a resume ATS score?

Resume ATS score: a number from 0 to 100, or a percentage, that a resume-checking tool calculates from your file, sometimes compared with one job description, to estimate how well an applicant tracking system will read and match it.

The word "ATS" in the name is misleading. The applicant tracking systems employers buy (Workday, Greenhouse, iCIMS, Taleo and the rest) store your application, parse your resume into fields and let recruiters search, filter and rank. None of the vendor documentation cited in this article describes showing the applicant a score. The number you see on a checker's results page comes from the checker.

Checker scores fall into two families, and mixing them up explains most of the confusion in search results:

• Job-match scores compare your resume with one job description and report how many of its terms you cover. Jobscan's match rate is one example. Change the job ad and the score changes, even if your file stays the same.
• Resume-quality scores grade the document without a job ad: bullet strength, length, formatting, parse risks. Resume Worded's main score works this way.

Some tools blend the two. The cvmakeover.ai checker scores parsing, structure and content on every upload, and adds a keyword category only when you paste a job ad. Its results page says so in plain words: the score "is not an employer's ranking."

That leaves the useful question. When a tool says 62 or 81, what did it count, and does any of it match what the employer's system does with your application? The next two sections answer each half: what employers' systems show recruiters, then how three checker formulas work.

What employers' systems show recruiters instead of a score

Two large ATS vendors publish how their AI matching presents candidates, and neither uses a 0–100 resume score.

Greenhouse Talent Matching sorts applicants into categories. Its Talent Matching FAQ lists labels "such as Strong, Good, Partial, Limited, or Needs manual review," shown to recruiters and hiring managers in their dashboard. The same page states that Talent Matching "does not auto-reject or auto-advance any candidate"; the hiring team makes every disposition decision.

Workday HiredScore gives each candidate a letter. The candidate grades reference defines them by required ("basic") and preferred qualifications:

Two details matter for candidates. The grade is set against the requirements of one requisition, so the same resume can be an A for one job and a C for the next. And HiredScore's candidate profile documentation lists the file types it reads for its parsed "smart resume" view: DOC, DOCX, PDF, RTF and TXT.

How common is each grade? New York City's Local Law 144 requires employers that use an automated employment decision tool to commission a bias audit within one year of use and publish a summary. Pfizer's published audit of its HiredScore grades covers applications to requisitions with New York City hires open between August 2022 and February 2025. Among the 20,897 applications from women and 19,924 from men, 27–28% received an A, 43–44% an A or B, and 69–71% an A, B or C. Roughly three in ten were graded D.

The same document describes the grades as an aid that "does not supplant human judgment," with recruiters keeping "complete discretion" over whom they select. A grade sorts the pile; a person still decides.

How checker scores are calculated

Three checkers publish enough about their method to compare. They answer different questions, which is why their numbers disagree.

Jobscan states on its homepage FAQ that it recommends "a match rate score of 75%," that many users "see success even with just a 65% match rate," and that "it might not be possible to score above 75% without overstuffing your resume with keywords." Word count and measurable results are not part of the match rate.

Resume Worded explains on its score guide that its checks "are weighted by importance using insights from hiring managers, so the score is not a simple average." Its main score does not read a job description; a separate Relevancy Score does that. The bands run from "room to grow" below 80 to "exceptional" at 90–100.

The cvmakeover.ai checker, as of 13 October 2026, scores each check as pass (1), warning (0.5) or fail (0), weights some checks more than others, and averages four category scores by the percentages in the table. The keyword check passes at 75% coverage of the ad's terms, with required terms counting double. Two caps sit on top: a file with almost no extractable text scores at most 15, and any failed parsing check limits the total to 75, because a resume that parses into the wrong fields should not look "mostly fine."

So a Jobscan 70 is a coverage figure for one ad, a Resume Worded 70 is a writing-quality grade with no ad, and a cvmakeover.ai 70 says a parsing or keyword check failed. The same digits mean three different things.

One resume, nine scores: what moves the number

To show how much a score depends on its inputs, we ran one fictional resume through the cvmakeover.ai scorer nine times on 13 October 2026, changing one thing per run.

Method. The resume belongs to "Jordan Rivera," a marketing coordinator with two jobs, a bachelor's degree, a skills line and seven bullets, one of which starts "Responsible for." We gave the scorer the plain text a parser extracts from a clean one-column file, so each variant differs only by the change named. The product normally pulls keywords from a pasted ad with a language model; to keep the test reproducible we wrote two keyword lists by hand, each with five required and five preferred terms. List A fits a Marketing Coordinator ad (email marketing, HubSpot, event planning). List B fits a Digital Marketing Specialist ad (SEO, Google Ads, A/B testing).

Three results stand out.

The job ad moved the score by 22 points with an identical file. Rows 2 and 3 are the same text. Against an ad it fits, the resume scores 96; against a neighboring marketing role it scores 74, because it covers only 3 of the 10 terms. That gap measures fit with the role, and no formatting change will close it.

One typographic choice cost 21 points. Row 4 contains every word of row 2. Spacing the letters of one heading turns "EXPERIENCE" into "E X P E R I E N C E" in the extracted text, which fails a parsing check and triggers the 75 cap. Greenhouse lists "a resume with spaces between the letters" among the causes of an unsuccessful parse, so this drop reflects a real reading problem.

Thresholds decide where effort pays. Adding two genuine skills raised keyword coverage from 80% to 93% and left the score at 96, because coverage had already cleared the 75% pass line. Rewriting one weak bullet added 5 points to the version without an ad (row 1 to row 7). A recruiter reading the resume might weigh those two edits the other way round. The number reacts to where the formula's thresholds sit, which is a property of our scorer and says nothing about Workday or Greenhouse.

Is 80% a good ATS score?

It depends on which number you are looking at, so answer two questions first: which tool produced it, and was a job ad involved?

For a job-match score, 80% is above Jobscan's own 75% recommendation. Pushing it toward 100% means copying more of the ad into your resume, and Jobscan itself warns that past 75% you may be stuffing. The more useful reading is which required terms are still missing. If the ad requires a certification you hold and your resume does not name it, add it. If it requires one you lack, the honest fix is not on the page.

For a resume-quality score, 80 sits in Resume Worded's "solid" band, just short of the 85 it calls good. Its checks cover impact, brevity and style, and the report lists which of them cost points.

For the cvmakeover.ai checker, look at the category breakdown before the total. A score at or just under 75 with a failed parsing check means a parse problem, and that is the first thing to fix. A score in the 70s with clean parsing and a weak keyword category means a gap against that specific ad.

What a score of 80 cannot tell you is your grade on the employer's side. HiredScore grades depend on required and preferred qualifications as the employer entered them, and Greenhouse's Talent Matching drafts its setup from the job post and scorecard attributes, after which the recruiter picks the skills, experience and job titles to match on. A checker cannot see either configuration.

Questions like "Is 70 a good ATS score?" or "Is 92 good?" have the same answer. The threshold that matters is the tool's own pass line for each check, and the most useful part of any report is the list of failed checks underneath the number.

How to use an ATS score without chasing it

Treat the score as a triage list. This order puts the fixes that change what an employer's parser reads ahead of the ones that only move a number.

1. Fix parse failures first. Unreadable text, letter-spaced headings, contact details in the page header, text boxes and two-column layouts change what lands in the employer's database. Each has a concrete fix in our ATS-friendly resume format guide.
2. Run it against the real ad. A score without a job description says nothing about fit. Paste the posting you are applying to, read the missing required terms, and decide for each one whether it is true for you.
3. Add only terms you can back up. A skill you list should appear in a bullet that shows it in use; our guide to the resume skills section walks through the pairing. A HiredScore grade is built on "basic qualifications," and the recruiter who opens your file reads the bullets.
4. Stop when the checks pass. Once parsing is clean and the required terms you have are present, further points come from the tool's thresholds, as rows 5 and 7 of the test showed.
5. Re-check the exported file. Scores describe one file. Export the final PDF or DOCX you will upload and test that file, because a font, a template or a save-as step can reintroduce a parse problem.

You can do steps 1 to 3 in one pass with the free cvmakeover.ai ATS checker: upload the file you plan to send, paste the job ad, and read the failed checks and missing keywords rather than the headline number. Other tools work too; the useful part of any of them is the same list.

For the wider picture of where AI ranking sits after parsing, and the notice and audit rules that apply to it, see our guide to AI resume screening.

ATS score questions

What is a good ATS score for a resume?+

It depends on the tool. Jobscan recommends a 75% match rate against a job description, Resume Worded calls 85 or above good on its quality score, and the cvmakeover.ai checker caps any resume with a failed parsing check at 75. Compare a score only with others from the same tool.

Is 70 a good ATS score?+

On a job-match tool, 70% is just under Jobscan's 75% recommendation and above the 65% it says often works. On Resume Worded's quality score, 70 falls in the band below 80 that it calls room to grow. Read which checks failed: a parse error matters more than a missing preferred keyword.

Do employers see my ATS score?+

Employers do not see the score from a checker you ran yourself. Their own systems may grade or categorize you: Workday HiredScore uses letters A to D, and Greenhouse Talent Matching uses categories such as Strong, Good, Partial and Limited, each measured against the job's requirements.

Can a low score get my resume rejected automatically?+

Neither Greenhouse nor Workday documents automatic rejection by match grade. Greenhouse says Talent Matching does not auto-reject candidates, and Pfizer's audit of HiredScore says recruiters keep complete discretion. Greenhouse does document auto-reject rules, but they run on answers to application questions, such as needing a visa the company cannot sponsor, not on a match score.

Can ChatGPT check my ATS score?+

It can compare your resume with a job ad if you paste both, but it has no fixed formula, so the number can change between runs. Ask it for the required terms your resume lacks and for any formatting it would expect a parser to misread, then verify the list yourself.

Key takeaways

• An ATS score is a checker tool's own calculation. Greenhouse's and Workday's documentation describes no score shown to applicants, and a score from one tool cannot be compared with a score from another.
• Job-match scores move with the job ad. In our test the same resume scored 96 against a fitting ad and 74 against a neighboring role.
• On the employer side, Workday HiredScore grades A to D on required and preferred qualifications, and in Pfizer's audited New York data about three in ten applications received a D.
• Parse failures deserve attention first because they change what the employer's parser stores; a single letter-spaced heading cost 21 points in our test and mirrors a failure Greenhouse documents.
• Once checks pass, more points come from the scorer's thresholds rather than from anything a recruiter reads, so stop optimizing and apply.