The Practical ATS Resume Guide

Applicant tracking systems are not intelligent. They are text parsers with rigid rules. Understanding what they can and cannot extract is the single biggest lever you have over whether your resume reaches a human at all.

This guide is not about beating the robot. It is about giving the robot exactly what it needs so it does not get in the way of the human who will actually decide.

What an ATS actually does

An ATS accepts an upload, converts the file into plain text, tries to identify common resume fields, and stores those fields in a database against your application. The parsing step is the fragile part. Everything before it — the visual design of your document — is invisible to the ATS. Only the plain-text extraction matters.

The fields an ATS typically tries to identify:

• Contact details (name, email, phone, location)
• Employer names and job titles
• Employment dates
• Education (degree, institution, dates)
• Skills and keywords
• Certifications

If any of these fields are ambiguous in the extracted text, the ATS records them as missing or wrong. The recruiter sees a profile with blank fields. In most cases your application is filtered out before a human looks at the PDF.

What breaks ATS parsing

Most failures come from five structural choices:

Multi-column layouts. Two-column resumes are common in design-forward templates. The parser reads the sidebar first or last, jumbling the reading order. A job title from the sidebar can be attached to a company from the main column.

Tables and text boxes. Text inside a table cell is sometimes extracted correctly and sometimes not. Text inside a text box often is skipped entirely. Either way, the field is lost.

Icons that replace words. A phone icon instead of the word "Phone", a location pin instead of the word "Location". The ATS sees a glyph. The field is empty.

Non-standard section headings. "Where I've Worked" is human-readable and ATS-invisible. The parser is looking for "Experience", "Education", "Skills". These standard headings are cues.

Fonts that contain no text. If you convert your resume to a vector graphic — for example by exporting it as a JPG then embedding that into a PDF — there is no text to parse. The ATS sees an image.

The ATS-safe resume structure

Every template in Advoke CV's ATS & Classic category follows this structure, and every resume that has to pass through an ATS should follow it too:

• Single column. No sidebars, no parallel content areas.
• Plain text order in the DOM. The visual reading order is the DOM order is the parse order.
• Standard headings. Summary, Experience, Education, Skills, Projects, Certifications, Languages, Awards, Volunteer, Publications.
• Left-aligned dates or right-aligned dates. Both work. Avoid dates in the middle of a line.
• Consistent date format. "January 2020 – March 2023" throughout. Not "Jan 2020 – Mar 2023" sometimes and "2020-2023" elsewhere.
• Full words, not icons. "Email", "Phone", "Location" as text labels, or as clearly-marked links.
• Standard fonts. System fonts parse cleanly. Highly custom fonts sometimes do not.

Keywords — honestly

The ATS does not care about a keyword count. It cares about keywords appearing in the right field. A "Python" mention in a Skills section is extracted into the skills database. A "Python" mention inside a project bullet is not, unless the parser is sophisticated enough to extract from free text.

The right approach: put the tools and technologies you actually use in the Skills section, and use them naturally in your experience bullets. Do not invent keywords to match a job posting. Do not bury the tools you know in prose.

Exporting correctly

The output format matters as much as the layout. PDFs parse reliably if they are text-based PDFs. Word documents parse reliably if the document uses styles rather than raw text boxes. HTML parses reliably if it is well-formed.

Avoid: image-only PDFs, scanned PDFs, PDFs where the text is embedded as an outline, files with passwords, files over 10MB.

Advoke CV exports text-based PDFs from the browser's native print engine. The output is searchable, selectable, and parses cleanly in every major ATS.

A final pre-submission checklist

• Copy-paste test from the PDF produces readable text in a sensible order.
• Section headings use the standard names.
• Every employer, title, and date appears on a single line or in a clear visual pair.
• Contact details are plain text, not an image or an icon.
• Skills are in a Skills section, not buried in prose.
• No image-only content anywhere in the PDF.
• The file is under 10 MB and has no password.

If everything on that list is true, you have done what you can for the ATS. The rest of the outcome depends on the human.