How ResumeScorePro scoring works
ResumeScorePro is a rule-based estimator of how well a resume is likely to survive applicant-tracking-system (ATS) screening. Every check is a published rule you can inspect — not a black-box model, and not a hiring guarantee.
The 21 checks, in five dimensions
When you paste a resume and a job posting, the scanner runs 21 checks. Each check scores from 0 to 1; a dimension's score is the average of its checks, and the overall 0–100 score is the weighted sum of the five dimensions. Scores of 85 and above are our "likely to pass" zone, 70–84 is borderline, and below 70 is high rejection risk. Those bands are our own calibration, not any employer's cutoff.
Keyword match — 35% (4 checks)
JD keyword match rate. The share of the posting's key terms that appear in your resume. Keywords are extracted from the posting by frequency — common filler words are removed, multi-word skill phrases such as "project management" count extra — up to 24 terms.
Top-5 JD terms present. Whether the posting's five most prominent terms appear in your resume.
Recognized hard skills. How many skills from our hard-skill dictionary (tools, platforms, and methods) appear in your resume.
Keyword depth. How many of the matched terms you use at least twice, rather than mentioning once.
Impact — 20% (4 checks)
Quantified achievements. Counts measurable results — money, percentages, multipliers, and team or project sizes. Years, phone numbers, and dates are stripped out first so they cannot inflate the count.
Strong action-verb openers. The share of your bullet points that open with a strong verb (led, built, delivered, grew) instead of weak phrasing such as "responsible for".
Bullet-point experience. Whether your experience is broken into scannable bullet points rather than dense paragraphs.
$ / % metrics present. Whether any money or percentage figure appears at all.
Format — 20% (8 checks)
Email address present and phone number present: the contact details parsers look for. Contact info at top: your email appears near the start of the document. Standard section headers: Experience, Education, and Skills headings a parser can recognize. Summary section: a Summary, Objective, or Profile heading. Employment dates present: enough dates for a parser to build your timeline. No excessive ALL-CAPS lines, and no overlong lines (over 45 words).
Brevity — 15% (3 checks)
Resume length in range. Full credit in the 300–600 word range; very thin and very long resumes score less. Concise bullets: an average bullet length of 28 words or fewer scores best. No filler buzzwords: words such as "very", "various", and "leverage" reduce the score.
Skills — 10% (2 checks)
Hard skills detected. How many recognized hard skills the scan finds; the check scales up to eight. Dedicated skills list: a comma-separated skills list (a Skills section) is present.
Free scan vs the Complete Report
The free scanner and the keyword checker run entirely in your browser — your text is never transmitted. Your most recent scan is kept only in your own browser's local storage, on your device, so the report page can pick it up where you left off. The Complete Report's item-by-item breakdown is also computed locally. Only its AI consultant section involves a server: if you buy the report, your resume and the job posting are sent once, over an encrypted connection, to our server and to OpenAI to generate that section. They are processed for your report only, are not stored by us, and are never used to train models.
What the score is not
ResumeScorePro is not affiliated with any ATS vendor (such as Workday, Greenhouse, Lever, iCIMS, or Taleo) and does not replicate any vendor's proprietary algorithm. A high score means your resume is parseable and well aligned with the posting's language; it cannot predict interviews or offers.
Data and sources
The checks are distilled from published ATS guidance and recruiter practice, and our blog guides cite the sources behind individual recommendations. We do not scrape candidate data, and resumes processed by the tools are never collected as training data.
How we update the checks
We revise the checks when published ATS guidance changes. When we change a check, we note the change and its date on this page.
Limitations
Pasted text only. The scanner reads the text you paste; it cannot judge how a particular PDF or Word file will be parsed by a specific employer's system.
English-language postings. Keyword extraction and the verb and filler lists are built for English.
Lexical matching. The scanner matches whole terms as written; the keyword checker additionally applies light stemming. Neither understands synonyms the way a human reader would.
Comparable only within ResumeScorePro. An 85 here is not an 85 in another tool; every checker weights different things.
Report an error
If you think a check scored you wrongly, email support@resumescorepro.com with the check's name and what you expected. Please don't send your full resume unless you choose to share it. We review every report and fix confirmed errors in the open.
Last updated: October 15, 2026 · Reviewed by the ResumeScorePro editorial team