Resources / Evidence-based analysis

Your “ATS compatibility score” can't see what you think it sees.

A resume checker can give you a number in seconds. The number may reveal useful risks, but it did not reach into a specific employer's Workday or Greenhouse account and watch your application being processed.

Upload a resume to almost any checker and you receive a result such as “78/100,” “ATS compatibility: good,” or “12 issues found.” The language makes the result feel like a direct diagnosis from the employer's system.

That is not how the measurement works. In our live test of Lever, Oracle Taleo, Workday, and Greenhouse, one Greenhouse application exposed zero candidate-visible parsing output. No autofill, preview, or structured resume fields appeared. The system may still have parsed the file for a recruiter, but the candidate could not observe the result.

If the real candidate flow does not expose its parsing output, a third-party checker outside that employer's account cannot be measuring that output. It is measuring a proxy.

What a resume checker actually measures

A typical checker combines three legitimate signals:

1. Its own parser. The tool extracts text and sections from your file. If a mainstream parser produces broken reading order or drops contact details, that is a useful warning—even though it is not the employer's parser.
2. Format heuristics. It looks for columns, tables, text boxes, image-only content, nonstandard headings, and other patterns that have caused problems in some systems.
3. Keyword overlap. It compares resume language with a job description. This estimates vocabulary alignment and searchability, not whether the file parsed correctly.

Those signals can answer a useful question: Does this file look generally parseable and aligned with this job description? They cannot answer the more confident question: Will this employer's ATS pass me?

The three layers hidden inside one score

Compressing all three layers into one number creates false precision. The file layer is controllable and testable. The employer's parser and configuration are mostly outside a candidate-side tool's visibility.

Why no universal score can measure “the real ATS”

Every application is a configured environment

The ATS vendor is only part of the system. Employers decide which fields appear, which are required, which questions eliminate candidates, which account values persist, and how the application is localized. In our test, Lever exposed 7 resume-derived fields, Taleo about 18 across 11 steps, Workday 3 before region rules interfered, and Greenhouse none.

The same resume met four different environments. A score calculated before opening the application cannot anticipate a configuration it has never seen.

The candidate view has a hard evidence ceiling

Even inside a live application, candidates can verify only what the form displays. We used a strict classification rule: “not shown” is not “failed to parse.” Any external checker sees less about that employer instance than the candidate in the application, not more.

This is why a precise score should be interpreted as the checker's internal diagnostic—not an acceptance probability, a vendor-specific compatibility rating, or a prediction of recruiter behavior.

When an ATS score is still useful

1. As a smoke test

If a reputable parser cannot extract your text, contact details, headings, or reading order, fix the file. Another parser may fail differently, but a known failure is not worth carrying into an application.

2. As a keyword mirror

Comparing your resume with a real target job can uncover missing tools, credentials, and role language. Add only terms you genuinely know and support them inside experience bullets. This improves both recruiter searchability and the human read.

3. As a layout tie-breaker

When two versions communicate equally well, prefer the one that extracts into clean plain text. A diagnostic is useful for eliminating avoidable risk, not for selecting the version with the most impressive number.

Audit the application, not only the file

The highest-leverage verification happens inside the real application, where the actual environment may expose its output. Use this five-minute audit:

• Verify every pre-filled value. Look for duplicated phone numbers, wrong current-job flags, invented dates or locations, and missing certifications.
• Keep checking through multi-step flows. Structured records and errors may appear several screens after upload.
• Review stored account data. On Workday-style systems, region and historical profile values may override current resume data.
• Use the final review as your source of truth. Correct what the application shows before submitting.
• On black-box forms, optimize for the recruiter. Clear structure, truthful keywords, and quickly scannable evidence survive more environments than score chasing.

The bottom line

An ATS score is a parseability-and-keyword estimate. Use it to find specific, controllable problems. Do not read it as a pass probability or proof of compatibility with a specific employer's system.

Then inspect the real application. What it pre-fills, rejects, overwrites, or hides is the only system-specific evidence available to you—and evidence is more valuable than a proxy.

FAQ