Skip to content

ATS resume checker: keyword match with a job description

Match your resume against a job description, on your device.

Paste the whole posting, requirements included
Paste the text, or drop the file

The job description and resume are read and compared in this browser; nothing is uploaded or saved. The site counts page views, not what you paste. PDF reading loads the pdf.js library only when you drop a PDF.

What does an ATS do with your resume?

An applicant tracking system (ATS) is the software an employer uses to collect applications, and this ATS resume checker shows the two parts of it you can influence: whether your file turns into clean text, and whether that text contains the words the posting uses. Most of what people fear about "ATS rejection" comes from those two steps, not from a robot reading your career and deciding against it.

The usual sequence runs like this. You upload a file. A parser converts it to plain text and tries to fill fields: name, email, employers, titles, dates, schools, skills. The recruiter then sees your application in a list and can filter it with screening questions ("Do you have a nursing license in Ohio?") or search the whole pool by keyword. Some employers also let the system rank applicants against the posting.

From file to search result: where a resume can drop out
Your file PDF, DOCX Plain text parser Fields jobs, dates, skills Search recruiter's words scanned image: no text at all text box, table: section lost or mixed "Excel" vs "MS Excel": not found

That last step matters more than most guides admit. Greenhouse's help pages describe its candidate search as Boolean: AND, OR, NOT, quoted phrases and a trailing asterisk. It does not describe automatic synonyms; its own example for a broader search spells the alternatives out (trainer OR instructor OR teacher). So if the posting says "financial modeling" and your resume says only "built models", a recruiter typing the posting's words can miss you. This checker treats known synonyms as matches but labels them, so you can see where your wording differs from the posting's.

Filtering is real and widespread. In a 2021 Harvard Business School and Accenture survey of 2,275 executives in the US, UK and Germany, 63 percent of employers used a recruitment management system, and more than 90 percent of those used it to make the first cut or rank candidates. The authors argued that rigid filters screen out many capable people.

How this checker reads a posting

The tool splits the posting into sections by their headings, then pulls out terms three ways: a list of about 230 common skills, tools and certifications with their synonyms and acronyms; two- and three-word phrases the posting repeats or puts in its requirements; and single words that repeat or are written as names (Tableau, Lambda, GAAP). Words that every posting uses, such as "experience", "strong" and "team", are ignored.

How much a term counts, by where it appears in the posting
Part of the postingWeight per mention
Job title (first line) 2.0
Requirements, qualifications, "what you bring" 2.0
Responsibilities, "what you will do" 1.4
Nice to have, preferred, bonus 1.1
Text with no clear heading 1.0
About us, benefits, equal opportunity text 0.15

Source: This tool's own weighting. Known skills and tools count 1.6 times; repeated phrases 1.2 times.

Each term is then looked up in your resume, ignoring plural and tense ("reconciled" finds "reconciliations") and accepting listed synonyms ("JS" for JavaScript, "k8s" for Kubernetes, "budget vs actual" for variance analysis). Short acronyms only count in capitals, so "ai" inside an ordinary sentence does not pass for AI. A phrase where you have the words but not together, such as "cohorts" and "analysis" for "cohort analysis", is marked as close wording and counts half.

Worked example (the one behind "Try an example"): the data analyst posting yields 18 terms with a total weight of 53.2. The resume has eight of them, including SQL, Python, Tableau, Excel and "split tests" for A/B testing, and close wording for three more, such as "Segmented customers" for customer segmentation. The score is (matched weight + half the close weight) ÷ 53.2, which comes to 61 percent. Adding Power BI would only be honest if you have used it, which is why the tool lists what is missing instead of rewriting anything for you.

Where to put a missing keyword

A missing term you genuinely have belongs where a recruiter expects to see proof of it. The posting's exact words help, but they should sit inside real evidence: what you did, with what, and the result.

Where each kind of term usually goes
Kind of termBest placeExample
Tool or software Skills list and the bullet where you used it Built weekly Power BI reports for 12 stores
Certification or license Certifications section, exact name and year ACLS, American Heart Association, 2025
Method or practice The experience bullet where you did it Ran variance analysis for the $85M budget
Acronym Both forms once, then either Search engine optimization (SEO)
Soft skill A result that shows it, not a word in a list Trained 30 new hires; error rate fell 40%

Source: Advice drawn from the University at Buffalo career resource center and Greenhouse search documentation.

Never hide keywords

White text, tiny fonts and pasted copies of the posting are old tricks. Parsers read hidden text, so it shows up in the recruiter's preview as a block of nonsense, and a person reads every resume that gets an interview.

Formatting that trips parsers

Textkernel, which sells a resume parser to recruiting software makers, names converting documents to text and handling column layouts as work its parser does for its customers; both are steps where a resume can go wrong. Newer parsers cope with far more than older ones, but you cannot know which one an employer runs. When you drop a file here, the text box shows the plain text the file produced; if it reads out of order, a parser may be seeing the same thing. To see exactly what changed between two versions of your resume, paste both into our text compare tool.

Common format problems and the fix
ProblemWhat can go wrongFix
Scanned or image PDF No text layer, so the record arrives empty Export from Word or Docs as PDF
Contact details in the header Some systems skip headers and footers Put them in the first lines of the body
Tables and text boxes Cells merge; boxed text can be dropped One column, tab stops for dates
Creative headings "Where I've been" is not read as experience Experience, Education, Skills
Seasons, two-digit years "Fall 2023" or "'21" cannot be dated "Sep 2023", "2021"
Icon fonts for phone and email Symbols come out as junk characters Plain words: Email, Phone

Source: University at Buffalo School of Management career resource center; Greenhouse Support, November 2026.

File type is less of a gamble than it used to be. Greenhouse, for example, accepts .doc, .docx, .pdf, .rtf and .txt, and says a resume can fail to import when it is a non-standard file type or contains characters its parser does not recognize. If a posting names a format, use it.

Common mistakes

  • Chasing 100 percent. A resume that matches every word of a posting usually reads like a copy of it. Most strong applications land well short of full overlap; cover the requirements you actually meet and let the rest go.
  • One resume for every job. Postings for the same title use different words. Keep a master version and adjust the summary, skills list and two or three bullets for each application.
  • Listing soft skills. "Excellent communicator" matches the word and proves nothing. A bullet with a number does both.
  • Pages of text. Long bullets bury the keyword. Our word counter helps keep a one-page resume near 400 to 700 words.
  • Forgetting the cover letter and the exit. If you are leaving a job, a clear resignation letter or a two weeks notice letter keeps the reference you may need for the next application.

Common questions

Is this the same as an ATS score?

No. There is no single ATS score. Each employer configures its own system, many never rank resumes at all, and the ones that do use their own methods. This tool reports a keyword match: how much of the posting's weighted wording appears in your resume. Treat it as a checklist, not a prediction.

Do applicant tracking systems reject resumes automatically?

Sometimes, mostly through knockout questions such as work authorization, a required license or minimum years. The keyword side usually works as search and ranking: a resume that lacks the recruiter's words is not deleted, it just does not come up.

Should I send a PDF or a Word file?

Either, if it is a text PDF or a plain DOCX and the posting does not ask for one. Drop your file into the checker: if the text that appears is complete and in order, the format is not your problem.

Can I use a two-column resume template?

It is a risk. Many current parsers handle columns, but older ones read straight across the page and mix your skills sidebar into your job history. The layout check above flags PDFs where text sits side by side.

Is my resume uploaded anywhere?

No. The comparison runs in your browser and nothing you paste or drop is sent to a server or stored. Closing the tab clears it. The site counts page views, not what you paste.

Sources

Last reviewed November 13, 2026.