Methodology
How Syntheve calculates your ATS match score
Six deterministic pillars, summed to a score out of 100. No black box: here is exactly what is measured and why.
The short version
Every ATS match score is the sum of six independently scored pillars, each worth a fixed number of points that add up to 100. The scoring itself is rule-based and deterministic: the same resume text and job description always produce the same pillar points, whether you are on the Free or Pro plan.
A separate step uses an AI model to write the plain-language summary and suggestion text you see in your report. That step only describes and explains a score that has already been calculated — it never sets or adjusts your points.
The six pillars
Points earned in each pillar sum to your overall score out of 100.
Section completeness — 10 points
Resumes built in the Syntheve builder start from a fixed 10/10 baseline, since the content is already structured data rather than a PDF layout to reverse-engineer. Uploaded PDFs are scored by how completely a parsing pass can recover structure, refined further when multiple parsers are run and their results compared.
Hard requirements — 25 points
Starts at 25 and deducts for specific, checkable gaps the job description states explicitly: years of experience, minimum education level, required field of study, required certifications, required languages, location, and work authorization. Anything the posting does not mention is not scored.
Skills match — 25 points
Must-have skills are worth up to 15 points and nice-to-have skills up to 10, split by how many of the posting's listed skills your resume evidences. An exact mention earns full credit; a closely related term (Pro plans only) earns partial credit for narrative purposes — the underlying point total is identical for every plan.
Experience match — 20 points
Blends three signals: how much of the posting's stated responsibilities your experience text actually covers (up to 12 points), how your tenure compares to any stated minimum (up to 5 points), and whether your most recent role is current or recent (up to 3 points).
Title & seniority — 10 points
Up to 5 points for meeting or exceeding the seniority tier the posting implies (junior, mid, senior, and so on), and up to 5 for how closely your own job titles align with the posting's title.
Education & certs — 10 points
Points for meeting a stated minimum education level, holding required certifications, holding preferred certifications, and matching any required field of study — each scored independently and only when the posting specifies it.
The years-of-experience knockout
One rule can override the pillar math directly: if a job description states a minimum years-of-experience requirement and your resume's inferred timeline falls clearly short of it, your overall score is capped at 38 — no matter how strong the rest of your resume is. This mirrors how many real screening steps treat a hard experience floor, and it is the most common reason a resume that otherwise reads well still scores low.
The fix is usually mechanical: make sure every role on your resume has clear start and end dates, so the timeline used for this calculation is accurate.
What Pro’s semantic matching does — and does not do
Pro plans run an additional embeddings-based step that recognizes when a skill missing from your resume is closely related to something you do list — for example, crediting familiarity with a broader analytics toolset when a specific tool is missing. This step only enriches the wording of your AI-written summary and suggestions.
It does not change your skills pillar points. The point total for every plan is always calculated with semantic matching switched off, so free and Pro users see the same score for the same resume and job description.
What this score can and cannot predict
Treat the score as a structured, repeatable second opinion on how closely your resume's text aligns with a specific job description — not a guarantee about any employer's internal hiring software or a prediction of interview odds.
- It cannot see a specific employer's internal ATS configuration or search behavior.
- It cannot account for applicant volume, competition, or internal-candidate preference.
- It cannot capture subjective recruiter or hiring-manager judgment.
- It only evaluates what is present in the job description text you provide — not unstated expectations.
Because the underlying rules are deterministic, the most reliable way to raise your score is to add truthful, specific evidence for gaps the report identifies — not to reformat around the scoring rules themselves.
FAQ
Methodology questions
No. The 0-100 score itself comes from six deterministic rule-based pillars comparing your resume text to the job description you provide — the same input always produces the same score. An AI model is used separately to write the plain-language summary and suggestion text, but it only describes results that are already computed; it cannot change your points.
More help in the help center.