Why a 78% Match Score Doesn't Mean "Apply"

Key takeaways

• A match score counts how much of the posting's vocabulary appears in your resume. That is all it measures.
• It weights every requirement alike, so a missing license and a missing "nice-to-have" cost the same points — and it cannot see whether the posting is even live.
• Two resumes can both score 78%: one from a candidate who has done the job under a different title, one from a candidate who shares the words and not the work. The score cannot tell them apart. A verdict can.

Every scanner produces one: a percentage, sometimes with a color, usually with a recommended threshold. Jobscan suggests aiming for the high seventies or above. Other tools set their own bars. And every candidate who sees a number below the bar does the same thing — opens the resume and starts adding words until the number rises.

The number is not meaningless. It is a real measure of a real thing. The problem is that the thing it measures is not the thing you are deciding.

What the score actually measures

A match score compares two documents and reports overlap: how many of the posting's terms — skills, tools, phrases, sometimes titles — appear in your resume. Better tools normalize word forms and weight hard skills above soft ones. Some add formatting and section checks. But the core is lexical: words in one document, present or absent in the other.

That is a useful diagnostic for a specific job: once you have decided to apply, which of the posting's terms is your resume missing? Used that way, on a posting you have already judged worth the hour, it saves time.

Used as a decision — the score is high, so apply; the score is low, so don't — it fails in three ways.

What it cannot see

1. Gatekeepers

A match score weights every requirement alike. A missing professional license, a missing work authorization, a hard experience floor you do not clear — each costs the same points as a missing "familiarity with Power BI." So a resume can score 85% and be filtered before anyone reads it, because the one requirement it missed was the one the process cannot proceed without. The score has no concept of a requirement that outvotes the others. Should you apply if you meet only 60% of the requirements explains why the kind of requirement matters more than the count.

2. Posting health

A scanner reads the posting's text. It does not know the posting has been up for four months, that it was reposted three times unchanged, that the requirements list describes one specific person who already works there, or that the company announced a hiring freeze last week. A ghost posting and a live one produce identical scores. The hour you spend raising the number on a dead posting is the hour the score should have saved you. Ghost jobs and internal-candidate postings covers what to check instead.

3. Visibility versus fit

This is the failure that costs experienced candidates the most.

Take two resumes, both at 78% against the same posting.

Resume A belongs to someone who ran the exact function the posting describes for four years — under a different title, in another industry's vocabulary, with the relevant work three bullets down. The overlap is 78% because the words differ; the work matches almost entirely.

Resume B belongs to someone who has used most of the posting's tools and terms in adjacent roles and never held the accountability the posting describes. The overlap is 78% because the words match; the work does not.

Same score. Opposite situations. A should apply — after closing a presentation gap, which takes forty-five minutes. B should probably skip. The score cannot tell them apart because it measures vocabulary, and vocabulary is precisely what separates them.

The verdict that distinguishes these two is the reason the Apply / Adapt / Skip method has a middle option: ADAPT is "fit is real, resume hides it" — Resume A's situation — and no percentage can express it.

What raising the number does

When the score is treated as the target, the resume gets edited toward the posting's words. Done carefully, on a genuine ADAPT, that is tailoring. Done as score-chasing, it produces the resume recruiters recognize on sight: the posting's phrases repeated back, a skills block that reads like the job ad, claims the experience section does not support. The number goes up; the odds go down. Recruiters say this plainly and often, and modern screening — increasingly capable of reading context rather than counting exact terms — makes exact-match repetition less useful every year.

What to check instead

Before the score, three questions, in order:

1. Do I miss a gatekeeper? If yes, no score matters. Skip.
2. Is the posting live and written for an outsider? If not, apply in ten minutes as-is or skip; never the hour.
3. Is my fit real, and does my resume show it in twenty seconds? Real and shown: apply as-is. Real and hidden: adapt — and here, and only here, a keyword diagnosis is worth running, on the gaps the verdict named.

That order is what a decision engine does and a scanner does not. It is also the order that keeps the tailoring hour on the postings where it changes the outcome.

SmarterApply is a job application decision engine: paste a job posting and your resume, and it returns one verdict — APPLY, ADAPT, or SKIP — with the specific gaps that produced it. The first analyses are free, no card required. Run it on the posting you are deciding about.

FAQ

What is a good match score?

There is no threshold that means "apply." A high score on a posting with a missing gatekeeper is a filtered application; a low score for a candidate with a hidden match is a strong one. Use the score as a keyword diagnostic after the decision, not as the decision.

Do applicant tracking systems reject resumes below a percentage?

No. ATS platforms parse resumes and let recruiters filter on criteria they set — a required qualification, a location, a keyword. The percentage is the scanner's own diagnostic, not the employer's.

Should I stop using match scores?

No — use them in their lane. On a posting you have already decided is an ADAPT, a keyword diagnosis tells you which terms to translate. On every other posting, it is noise.

What does SmarterApply return instead of a score?

A verdict — APPLY, ADAPT, or SKIP — and the specific requirements behind it, separated into ones you meet and ones your resume shows. The gap between those two lists is what a score cannot express.