{"id":588,"date":"2026-09-21T14:39:33","date_gmt":"2026-09-21T11:39:33","guid":{"rendered":"https:\/\/soakjam.com\/knowledge\/?p=588"},"modified":"2026-09-21T14:39:33","modified_gmt":"2026-09-21T11:39:33","slug":"how-applicant-tracking-systems-read-cv","status":"publish","type":"post","link":"https:\/\/soakjam.com\/knowledge\/how-applicant-tracking-systems-read-cv\/","title":{"rendered":"How Do Applicant Tracking Systems Read a CV?"},"content":{"rendered":"\n

How do applicant tracking systems read a CV? Most begin by extracting text and turning familiar parts of the document\u2014such as contact details, employers, job titles, dates, education and skills\u2014into structured fields that recruiters can search and review.<\/p>\n\n\n\n

That process is often described as a robot deciding who gets a job. Reality is less uniform. An applicant tracking system, or ATS, can store applications, parse documents, collect screening answers and support recruiter workflows. Whether it ranks, filters or automatically rejects anyone depends on the product, the employer\u2019s configuration and the particular vacancy.<\/p>\n\n\n\n

\n

Quick answer:<\/strong> An ATS usually extracts machine-readable text, looks for section patterns and associates phrases with fields in a candidate profile. Recruiters may then search, filter or review that profile. A clear single-column CV, ordinary headings, selectable text, accurate dates and job-specific evidence reduce parsing problems. Keywords help only when they truthfully describe your experience.<\/p>\n<\/blockquote>\n\n\n\n

How do applicant tracking systems read a CV? The short answer<\/h2>\n\n\n\n
Stage<\/th>What may happen<\/th><\/tr><\/thead>
Upload<\/td>The system checks file type, size and whether the document can be processed<\/td><\/tr>
Text extraction<\/td>Text is read from a document layer or, in some systems, from optical character recognition<\/td><\/tr>
Parsing<\/td>Content is associated with fields such as name, email, employer, role, date, education and skills<\/td><\/tr>
Application questions<\/td>Answers may be stored or used for employer-configured eligibility rules<\/td><\/tr>
Search and filtering<\/td>Recruiters may search terms, locations, experience or other structured fields<\/td><\/tr>
Ranking or matching<\/td>Some tools offer scores or recommendations; others do not<\/td><\/tr>
Human review<\/td>A recruiter or hiring manager evaluates the application in context<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

There is no universal ATS test that all CVs must pass. A format accepted by one system can parse differently in another, and employer rules can matter more than the software brand.<\/p>\n\n\n\n

An ATS is a workflow system, not one secret score<\/h2>\n\n\n\n

Employers use applicant tracking systems to collect applications and move candidates through stages such as new application, recruiter review, interview, reference check and offer. The system can also keep correspondence, interview notes, consent records and reporting in one place.<\/p>\n\n\n\n

Resume or CV parsing is one feature inside that larger workflow. It saves a recruiter from typing every field manually and makes the database searchable. Greenhouse, for example, describes parsing as a way to locate resume details and build candidate information. Lever documents support for several common file types and explains that resumes are parsed into candidate profiles.<\/p>\n\n\n\n

Some platforms add matching, artificial intelligence or automated recommendations. Others leave selection almost entirely to people. An employer can also configure knockout questions\u2014such as legal work authorisation, a required licence or willingness to work at a stated location\u2014without analysing the CV at all.<\/p>\n\n\n\n

So \u201cthe ATS rejected me\u201d can describe several events: a required answer was not met, a recruiter declined the application, a parsing error hid information, a role was filled, or an automated rule applied. The rejection message rarely reveals which one.<\/p>\n\n\n\n

How do applicant tracking systems read a CV as text?<\/h2>\n\n\n\n

The first technical task is text extraction. A DOCX file stores paragraphs and formatting in a structured document. A normal PDF can also contain selectable text with coordinates showing where each word appears. The parser reads that underlying content rather than seeing the page exactly as a person does.<\/p>\n\n\n\n

A scanned PDF may contain only a photograph of a page. Some systems use optical character recognition, or OCR, but quality varies with resolution, contrast, language and layout. A visually sharp scan can still produce wrong characters or reading order.<\/p>\n\n\n\n

You can perform a simple check: open the final file, select all text, copy it and paste it into a plain-text editor. If the result is missing, scrambled or arranged in a confusing order, a parser may also struggle. This is not a complete ATS simulation, but it exposes many avoidable document problems.<\/p>\n\n\n\n

Do not upload an image merely because it looks identical on your screen. Unless the employer specifically requests a scan, use a text-based document created from a word processor or layout tool that preserves selectable text.<\/p>\n\n\n\n

Parsing turns a page into fields<\/h2>\n\n\n\n

After extraction, the system tries to identify what each piece of text means. A line containing an email pattern is likely contact information. A month-year range beside a company and role may be work experience. A recognised degree near an institution may be education.<\/p>\n\n\n\n

Parsers combine rules, statistical models and, increasingly, language models. They use headings, punctuation, location on the page and relationships between lines. Greenhouse explains that standard resume structure guides its parsing of details such as name, email, phone, address, current title and company.<\/p>\n\n\n\n

The result is not perfect understanding. A parser may mistake a client for an employer, split a two-line job title, attach a date to the wrong role or treat a sidebar as the start of the main text. Recruiters can often open the original document, but incorrect structured fields can affect search and sorting before they do.<\/p>\n\n\n\n

That is why clarity helps without requiring a dull document. A restrained visual hierarchy\u2014name, contact details, standard headings, role, employer, location, dates and evidence\u2014works for both machines and hurried people.<\/p>\n\n\n\n

\"ATS
Parsing turns document text into structured fields, but unusual reading order or unclear labels can produce mistakes.<\/figcaption><\/figure>\n\n\n\n

How do applicant tracking systems read a CV layout?<\/h2>\n\n\n\n

Reading order matters more than decoration. A person can glance across two columns and understand that the left sidebar contains skills while the right column contains experience. A text extractor may read across the page, place both columns one after another or interleave them line by line.<\/p>\n\n\n\n

Common sources of trouble include:<\/p>\n\n\n\n