Bloom

How Resume Match Scores Actually Work

By Juan F. with Claude · Founder, Bloom · August 24, 2026

A speedometer-style gauge with its needle pointing to a high score, beside a resume where two lines carry a checkmark and one line carries an empty, unverified marker

A resume match score is a percentage guess at how much your resume’s wording overlaps with a job description’s — nothing more, nothing mystical. It rewards using the posting’s own language for skills, tools, and titles, and it punishes formatting that keeps a computer from reading your resume correctly. What it flatly cannot do is check whether any of it’s true. That’s the part nobody puts on the label, and it’s the part that actually matters.

If you’ve ever run your resume through a scanner and gotten back an “82% match,” here’s exactly what that number is — and isn’t — telling you.

What a match score actually counts

A match score isn’t measuring how qualified you are. It’s measuring text overlap between two documents: your resume and the job posting. In practice, that means it’s counting things that are countable, because that’s the only category software can check without shrugging.

Harvard Business School and Accenture’s Hidden Workers: Untapped Talent survey — over 2,250 executives across the US, UK, and Germany — found that more than 90% of employers use their applicant tracking systems to filter or rank candidates before a human ever opens the file. And the report doesn’t hide the ball on what these systems actually screen for: exact credential matches, employment-gap length, and other things that reduce to a string comparison or a date subtraction. Not judgment calls. Arithmetic.

Two more findings from that same report drive it home. Almost half of the companies surveyed said a résumé with an employment gap longer than six months gets automatically screened out — pure date subtraction, zero human judgment involved. And 88% of employers agreed that qualified high-skills candidates get vetted out of the process simply for not matching a job description’s exact criteria (94% for middle-skills roles). That’s not a vibe about how filtering feels from the outside. That’s employers describing their own funnel, out loud.

That’s the honest shape of a match score: a percentage built from things a machine can verify — does this word appear, does this phrase appear, does this credential string match exactly — not from any assessment of whether you’re actually good at the job.

Why formatting can wreck your score as badly as bad wording

A match score can only count what it can successfully read. If your layout confuses the parser, the right words in the wrong place might as well not exist.

This isn’t a hypothetical. In a 2026 field study of 204 real applications, a human-written resume in a plain single-column layout landed a positive response 31% of the time — the best result in the whole test — while the same kind of resume in a two-column layout with a sidebar dropped to 12%. (Reported by StudyFinds.) Kickresume’s CEO pinned part of the gap on single-column formats being easier for screening software to parse correctly. Sidebars, tables, and text boxes can get scrambled or skipped outright by systems built to read left to right, top to bottom, like a normal person.

Translation: your resume can contain every right word and still score badly if the layout keeps those words from being read the way you intended. When you’re chasing a match score, boring formatting is doing at least as much work as clever phrasing.

Here’s what a match score can’t tell you — and it’s the important part

A match score has zero mechanism for checking whether a claim is real.

Write “led a cross-functional team of eight” when you actually supported a team someone else led, and the score has no idea. It’s counting whether “cross-functional team” and a number show up near a leadership-flavored verb — not verifying your actual role. Same story for a tool you’ve listed but never touched, or a title you’ve quietly upgraded. All of it reads as a match. None of it survives an interview.

That gap is exactly where good candidates get rejected for the wrong reason, and weaker matches sail through for the wrong reason too. A high score gets you past the software. It says nothing about whether a human will believe what’s on the page once they start asking questions. Our piece on what an ATS actually scans for, using one real posting and three real resumes, walks through exactly how a keyword-level match and a genuine requirement match can point in opposite directions.

Treat the score as a floor, not the finish line

Here’s the useful reframe: a match score tells you whether you’ve cleared the minimum bar to be seen by a person at all. It does not tell you whether you should be hired — and chasing the number past that point, by adding words you can’t back up, actively works against you, because it just raises your odds of landing an interview you can’t get through.

The better target is a score that’s high because the underlying match is real: your true experience, described in the posting’s own words, formatted so it’s actually counted. That’s the same thing tailoring does — see our step-by-step guide to tailoring a resume to a job description — a match score is just one visible readout of whether you did the work.

FAQ

What does a resume match score actually measure?

Text overlap between your resume and a specific job description — whether your wording for skills, tools, titles, and requirements lines up closely enough for scanning software to count it as a hit. A measure of language overlap. Not qualification. Not ability.

Why did my resume get a low match score even though I’m qualified?

Two usual suspects: your resume describes real, matching experience using different words than the posting (so the overlap never registers), or your formatting — multi-column layouts, tables, text boxes — is keeping the software from reading the relevant section at all.

Does a high match score mean I’ll get an interview?

It means you’re more likely to clear an initial software or keyword-based screen. It says nothing about whether a human will find your claims credible once they actually read the resume, and it has no way to verify that any specific claim is true.

Should I add keywords to my resume just to raise the match score?

Only if they’re true. A term you can’t back up might raise your score and land you an interview you then can’t survive — which costs you more than the lower score ever would. The goal is a real match, described accurately, not the highest number you can manufacture.

Do different applicant tracking systems calculate match scores the same way?

No universal formula exists — these systems blend exact-word matching with configurable weighting that most vendors don’t publish. What is consistent: countable things (exact terms, credentials, dates, correctly parsed formatting) drive the score. Judgment calls don’t, because software doesn’t do those.

Sources

Bloom shows you how your resume actually maps to a job’s real requirements — not just a score, but which line of your history proves each one, and which claims still need real evidence behind them. Try it free →