Guide · Updated 2026-09-28
ATS-friendly CV formatting: what parsers actually read
An applicant tracking system converts your CV into a plain text record before anyone reads it. The formatting decision that matters most is not aesthetic — it is whether the parser can extract your experience at all. This guide covers the layouts that work, the ones that silently lose content, and the specific choices that cause data to disappear.
How parsing actually works
There is no single applicant tracking system. Employers use a range of products, and each extracts text in a slightly different way. What they share is a staged process: the document is converted to text, the text is split into sections using headings and layout cues, and the sections are stored as fields that a recruiter can filter on.
That middle step is where formatting does damage. When a parser cannot tell that a block of text is a job title rather than a company name, it guesses. When it cannot tell where one section ends and the next begins, it merges them. Nothing is reported back to you, so the failure is invisible from your side.
A layout that parses reliably
- A single column, top to bottom. This is the single most reliable choice.
- Standard section headings: Work Experience, Education, Skills, Certifications.
- Your name as a plain text heading at the top of the document.
- Contact details as text in the body, never in a header or footer.
- Skills as a plain list, comma or pipe separated.
- One consistent date format throughout, such as Jan 2021 - Mar 2023.
- PDF as the export format, with real selectable text rather than an image.
| Element | Effect |
|---|---|
| Single-column body | Reading order matches visual order. Safe. |
| Contact details in a header | Header text is commonly excluded. Data loss. |
| Sidebar with contact details and skills | Ambiguous order. Sections can be lost or swapped. |
| Icon-based skills | No text to extract. Silent loss. |
| Star ratings for skills | Graphic only. Not a data format. |
| Layout tables | Usually read column-wise, scrambling related text. |
| Text boxes and floating shapes | Ordering is unpredictable. Often skipped. |
| Standard headings | Give the parser reliable anchors. Safe and helpful. |
Avoid text boxes entirely
Designers reach for text boxes because they offer precise placement. For a CV they are actively harmful. A parser reads the main text flow, and text floating in a box often sits outside that flow entirely. If a layout requires a text box, the layout is wrong for a CV.
Make your headings do the work
Parsers use headings to find the boundaries between sections. If your experience section is titled 'Where I have worked' and your skills section is titled 'What I know', a parser looking for standard labels may not find either. Using conventional headings costs you nothing and gives the parser reliable anchors.
Keep dates consistent
Inconsistent date formats are a quiet problem. 'January 2021', '01/2021' and '2021' may all describe the same period but may be parsed as three different things. Pick one format and use it everywhere, including for education dates.
Where to go next
Formatting is only half of it. Once your CV parses correctly, it still has to be worth reading, which is a separate problem covered in the guide to quantified achievements.
Improve the CV you already have
Upload a PDF or DOCX and CVForge returns a sharper, ATS-safe, page-limited PDF. Your facts are preserved; gaps are reported rather than invented.
CVForge is an automated CV enhancement tool built by Januth Nimnal as part of OrinAI. It improves an existing CV and does not write one from scratch. About CVForge.