Skip to main content
ATSEyeATSEye
Methodology

How ATSEye scores your resume — no black box.

Most resume tools hand you a number and keep the reasoning hidden. ATSEye does the opposite. Your score comes from a deterministic, six-layer engine — the same code every time, with a traceable reason for every point. This page explains exactly what each layer measures, how personalized and general scoring differ, and — just as important — what the score does not claim to predict.

Score my resume free

First, the honest framing

What the score is — and what it isn't.

It is

A deterministic, 0–100 estimate of how well your resume matches a job on the signals an ATS and a recruiter weigh — ATSEye's own, reproducible assessment, built to give you a clear, explainable target to improve against.

It isn't

A prediction that any specific employer's ATS will accept or reject you — different systems are configured differently and no tool can know each one's rules — and it is not a guarantee of an interview. The hiring decision stays with people.

The engine

Six weighted layers, and what each one measures.

Resume parsing and job-description analysis run first as prerequisites. These six weighted layers then determine the number — the weights below are for personalized (with-JD) scoring and sum to 100%.

Keyword coverage

40%

Whether your resume actually contains the required skills and terms from the job — the signal a recruiter's search filters on. It carries the most weight because it's the most literal match, but stuffing is self-defeating: the integrity layer detects and penalizes padded, repeated, or context-free keywords.

Experience alignment

20%

Do your years, seniority, and trajectory match what the role asks for? Parsed from your actual dates and titles — not guessed.

Impact

15%

Quantified, outcome-led bullets versus vague duty statements. Resumes that lead with results score higher because that's what recruiters shortlist on.

Semantic match

10%

How closely the meaning of your experience maps to the job — so a relevant accomplishment counts even when it doesn't use the exact keyword.

Soft-skills signal

10%

Evidence of collaboration, ownership, and communication expressed through your work — not a checklist of adjectives.

Format & parseability

5%

Clean structure and section naming that an applicant tracking system can extract without dropping content to graphics or columns.

Two more checks run alongside the score but carry no weight of their own: a readability diagnostic (surfaced as guidance) and an integrity guardrail that deducts points for keyword stuffing or fabricated details. AI-generated insights are advisory and never set the score.

Why it's built this way

Three principles behind the number.

Deterministic, not a guess

The weighted layers are computed by code, so the same inputs always return the same score. The AI never sets the number.

Reproducible, with honest caveats

Stable within an engine version. Scores are versioned, and OCR of scanned PDFs can vary slightly — we say so rather than pretend otherwise.

Truthful by construction

Optimization re-scores through the same engine and can't invent employers, dates, or metrics. A score you can't defend is worse than no score.

Your resume data

Uploaded files are deleted right after the text is extracted, resume text is not logged in plaintext, and personal details are redacted before any AI model processes your content.

For hiring teams

When ranking candidates, the score is a recommendation to support a human decision — never an automatic accept or reject. People review the shortlist and make the call.

Try it

Put the methodology to work.

Free ATS resume checker

Get your deterministic six-layer score against any job description.

Check

AI resume optimizer

Fix the flagged gaps with truthful rewrites, re-scored by the same engine.

Optimize

What an ATS score means

The number in plain English — and why most online scores are unreliable.

Read

What is a good ATS score?

Realistic targets and why the threshold moves with the employer.

Read

Questions

How the score works — frequently asked

What does the ATSEye score actually represent?+
It's a deterministic estimate — on a 0–100 scale — of how well your resume matches a job across the signals an ATS and a recruiter weigh. It is ATSEye's own assessment, not a prediction that any specific employer's applicant tracking system will accept or reject you. Different employers use different software configured in different ways; no tool can know each one's rules. What the score gives you is a consistent, explainable target to improve against.
Why does the same resume always get the same score here?+
Because the six weighted layers are computed by deterministic code, not by asking a language model to guess a number. The same resume and the same job description always produce the same score. When the number moves, your edits moved it — not randomness. (The AI insights and rewrites are generated separately and never set the score.)
Is the score truly reproducible?+
Within a fixed engine version, yes. The weighted layers are reproducible for the same inputs. Two honest caveats: scores are keyed to a scoring-algorithm version, so historical scores stop being directly comparable after an engine update; and a scanned (image) PDF goes through OCR first, which can introduce small variation in the extracted text. For normal text-based PDFs and DOCX files at a fixed version, the score is stable.
What's the difference between scoring with and without a job description?+
With a job description you get a personalized match score (Mode A) that weighs keyword and semantic alignment against that specific role. Without one you get a general ATS analysis (Mode B) that focuses on formatting, impact, experience, and readability — the JD-relative layers are set aside and their weight is redistributed. The biggest, most specific gains almost always come from scoring against a real JD.
Can AI optimization inflate my score by inventing things?+
No. Every AI rewrite is re-scored by the same deterministic engine and passes an anti-fabrication gate that blocks invented employers, dates, degrees, and metrics. The score is also monotonic — an approved optimization never regresses below your starting point. The goal is a higher score you can still defend in the interview.
Does a high score guarantee an interview?+
No — and any tool that promises that is overselling. A strong score means your resume is well-matched and clearly parseable, which removes a common, avoidable reason for being filtered out. The hiring decision still rests with people, and relevant context may live outside the resume. ATSEye improves your odds by fixing what's fixable; it doesn't guarantee an outcome.

Methodology last reviewed 25 July 2026 · scoring engine version 1.5.2.

See your six-layer score — free.

Upload your resume, paste a job description, and read the reason behind every point. No signup.

Try ATSEye free