Applicant tracking systems, decoded

Understand how software reads your resume.

Tools, tests, and evidence-based guides for building a resume that extracts cleanly, without chasing a fake universal “ATS score.”

Local text processingPrimary-source claimsNo invented benchmarks
DOCUMENT TRACEREADY
WORK EXPERIENCE
SKILLS
typescriptcloudflaresql
  1. 01Extracttext + structure
  2. 02Parsesections + fields
  3. 03Routesearch + workflow
HEURISTIC ≠ HIRING DECISIONv1.0

Local utility

Find the terms your resume may be missing.

Paste resume text and a job description. The matcher compares meaningful terms and phrases on your device. It is a revision aid, not an ATS simulation.

Open keyword matcher

Failure map

Where resume parsing breaks

A file can look perfect and still produce bad candidate data. These are document-level failures you can inspect before applying.

E-01

No text layer

A scanned or image-exported PDF may contain pixels without searchable words.

Inspect file formats
E-07

Reading order drift

Columns and positioned text can extract in a different order than they appear.

Reduce layout risk
E-12

Field ambiguity

A parser can separate a job title, date, or employer from the entry it belongs to.

See parser stages
E-18

Invisible contact data

An icon cannot replace a real email address, phone number, or link label.

Run the checklist

Format decision

PDF vs DOCX is the wrong first question.

Start with the employer’s requested type. Then ask whether the exported file has usable text, a sensible order, and the structure you intended.

Read the full file analysis →
PROPERTYPDFDOCX
Visual stabilityHighVariable
Native structureVariableExplicit XML
Image-only riskPossibleLow
Complex layout riskReading orderTables / boxes
RESULT: VERIFY THE EXPORTED FILE

Core guides

Start with the mechanism

Method

Documented. Observed. Unverified.

Documented A primary source states the behavior.

Observed A public product or file exposes the attribute.

Unverified A claim lacks a reproducible public test.

Read the methodology

Research queue

Protocols before results

We publish what a test will measure before attaching a result to a product.

Common questions

Direct answers

What is an ATS?

An applicant tracking system stores applications and supports recruiting workflows. Resume parsing, keyword search, matching, application questions, and rejection rules are separate capabilities that can be built in, configured, or supplied by another product.

Can an ATS read a PDF?

Many systems accept PDF files. A native-text PDF is very different from an image-only scan. Search the final PDF and paste its text into a plain-text editor to catch basic extraction and reading-order problems.

Are two-column resumes ATS-friendly?

They can be, but columns add reading-order risk. The real test is whether extracted text stays in the intended sequence. A single column is the conservative default when you cannot test the target parser.

What is a good ATS score?

There is no universal passing score. Commercial checkers use their own rules. Some employer products offer configured match labels or grades, but those do not create one shared score across hiring systems.

Does an ATS reject resumes automatically?

Some employer setups include automatic rules, such as rejection tied to a required application-question response. That is different from every resume being rejected by a hidden universal keyword score.

Does the keyword matcher upload my resume?

No. The matcher runs in your browser and does not send entered text to a server or database.