What Counts as a Good Resume Score Depends on the Checker

An 85 can be good, average or just short of the next band, depending on the checker. Resume Worded calls 85 good; Rezi places it in a middle band; Bowling Green State University’s VMock guidance puts it one point below green.

That is why there is no universal “good resume score.” Scores are useful only when you know what the tool measured, where it set its bands and which findings led to the number. Resume Worded makes the same point: compare scores within one tool, not across different ones.

The table below places each tool’s own explanation side by side. One row is our builder, so the details of that row are drawn from our code and should be read as a maker’s account. The other rows come from the cited vendor and university pages, read on October 31, 2026.

The same number means different things

We did not run any resume through the other tools. Every cell about them repeats what a page in the sources says.

Tool	What the number measures, per its maker	What its maker calls good	Where an 85 lands
Jobscan	How well the resume matches one job description you enter	“a match rate of at least 75%”	Above the target, on a different measure
Resume Worded	“30+ checks”, on a scale from 0 to 100	“85 or above, and ideally 90 or above”	Good
Rezi	Five areas: Content, Format, Optimization, Best Practices and Application Ready	90+: “strong shape, and you’re ready to apply”	The 50 to 89 band
Enhancv	The share of the resume it could parse, and the issues its checker finds	“higher than 80” means “mostly good”	Mostly good
VMock, on one university’s page	A score from 0 to 100 “computed based on 3 core modules”	Green Zone, 86 to 100	Yellow Zone, 33 to 85
Our builder	Three parts: parsing, content and completeness	Ready, 85 and above	Ready, at its lowest point

The VMock row comes from a university’s description, not from VMock itself, and the St. Lawrence page shows that even schools using the same tool can phrase the goal a point apart. None of the checker pages we read puts a number on what each part of its score weighs. Rezi names five areas, Resume Worded sorts its checks into four groups and Enhancv names two factors, but how much each counts toward the total is not published on those pages.

A match rate answers a different question

Jobscan’s figure answers a different question from the others. Its tutorial contrasts the match rate with “other tools that grade how “good” your resume is” and says its own score “actually tells you how well you match the job”. The tutorial lists what the rate rests on, in order of priority: hard skills, education level (only when the posting asks for an advanced degree), job title, soft skills and other keywords. It also says “Resume word count and measurable results are not factored into the match rate”.

A match rate therefore moves with every job posting you paste in, while a resume score stays put until you edit the resume. Bowling Green’s page shows the two side by side for one tool: VMock’s resume score needs 86 to reach green, and its match score against a job description is green from 68%. The same 70 would be green on one and yellow on the other.

Our builder keeps the two apart as well. Its job match lists a posting’s keywords and shows which ones your resume prints, and that count is not part of the resume score. How it decides that a keyword is there is explained in our post on resume keywords from a job description.

The score is not an employer’s score

Two of the companies that sell scores say on their own pages that the number is not an employer’s. Rezi’s explainer says “real ATS systems don’t give your resume an official score” and “Recruiters won’t see your ATS score”. Enhancv’s checker page says “there’s no such thing as an ATS score – no tool online that provides a score gives an actual score.”

That does not mean nothing on the employer’s side rates applicants. Some applicant tracking systems can rate them for the hiring team, against criteria the employer sets. Greenhouse’s Talent Matching FAQ says the feature will “sort the candidate pool into match categories (such as Strong, Good, Partial, Limited, or Needs manual review)”, and that the results go to “recruiters and hiring managers in their dashboard”. That result is measured against the employer’s own setup, and no checker can reproduce criteria it has never seen. Our guides to Greenhouse and Ashby quote what each vendor documents, and a separate post traces the 75% ATS rejection claim to its oldest sources.

How our builder calculates its score

Our resume score is worked out by plain code with no AI in it, although the builder’s import and its writing help do use AI. The same text, settings and PDF give the same number every time, and every rule, weight and threshold is published on the builder page, along with the page a rule takes its idea from when it has one.

The total runs from 0 to 100 and is built from three parts. Parsing, which asks whether software can read back the contact details, entries, dates and headings, weighs 30%. Content, how the bullets and the summary are written, weighs 45%. Completeness, whether the parts a reader looks for are there, weighs 25%. Each part starts at 100 and loses points for each rule in it that fails. Tips are advice and weigh nothing. The bands are Ready from 85, Almost there from 60 to 84 and Needs work below 60.

Five rules are blockers, and any one of them caps the total at 49 however well the rest scores: no name, no email, placeholder text such as lorem ipsum, an empty resume, and a bracketed gap such as “[X]%” still on the page. The last one exists because a resume with an unfilled gap should not read as ready to send. The builder page publishes the rule for gaps in full, and our post on numbers in AI-written bullets shows where such gaps come from.

The weights are our own choices, and the builder page says so. A rule that links a career office guide or a parser vendor’s documentation shows where the idea came from; it does not mean that source endorses our numbers. The level you set for a resume, entry, mid or senior, changes four rules, among them the number of pages expected. Some rules rely on English word lists and are skipped for a resume in another language.

Why one resume can show two scores in our tools

Even within our own tools, one resume can show two scores, and the reason is the same one that separates the checkers: what each run has to work with. Two rules, the page count and a read-back of the text from the finished PDF, wait until the editor has drawn the PDF. The list of your resumes draws no PDF and leaves both out, so the number there can be higher than the one in the editor. The free resume checker runs the same rules on the resume as the import copied it. For a PDF it counts the pages of the file you uploaded, and it does not run the read-back at all. The builder page explains the arithmetic step by step.

The rules themselves also change from time to time. When we reworked the score in October 2026, a resume’s number could move in either direction without a word of it changing, and the builder page has a section on what changed in that version. A score from last month and a score from today may not be the same measure, in our builder or in anyone else’s.

The builder shows the score and its findings as you type, account or not; only the download asks you to sign up for a free one.

Use a score to choose the next fix

A higher score on one checker and a lower score on another tells you little by itself. Read the findings. If two tools flag a missing phone number or a job with no dates, fix that first. If only one objects, it may be enforcing a preference the other does not share.

Every score has a boundary. Ours cannot tell whether the resume fits a role, whether a claim is true or whether it will lead to an interview. It counts the visible structure of the document; a human still decides what the work means.