ATS-friendly résumés: what actually matters

What an applicant tracking system is

An applicant tracking system (ATS) is the database and workflow tool a company uses to run hiring. Workday, Greenhouse, Lever, iCIMS, SAP SuccessFactors and Taleo are common examples. When you apply, the system stores your application, runs your résumé through a parser that tries to fill in fields (name, email, phone, each job with employer, title and dates, education, skills), and gives recruiters a place to search, filter, tag and move candidates between stages.

The parser is the part that touches your document. It extracts the text, then uses section headings, date patterns and position on the page to decide which text belongs to which field. Everything else in the system works on those fields and on the recruiter's own actions.

The rejection robot mostly does not exist

The common picture of an ATS is a filter that scores your résumé against the posting and discards it below a threshold. That is rarely how these systems are configured. Most rejections that feel automatic are one of two things: a knockout question in the application form (work authorisation, a required licence, willingness to relocate) answered the wrong way, or a recruiter who searched or filtered the candidate pool and never opened your file.

That changes what you should optimise for. Rather than beating an algorithm, you want two things: to be findable when a recruiter searches for the terms in the posting, and to be readable in the few seconds a person gives the page once it is open. Both depend on the same properties: correct parsing and clear, specific wording.

What breaks parsing: reading order

A parser reads text in the order the file stores it, which is not always the order you see. In a two-column layout or a table, the left column's text may arrive interleaved with the right column's, so that a date lands next to the wrong job or a skills list breaks a sentence in two. Text inside an image (a name set as a graphic, a skills chart, a logo bar) is not text at all to the parser and disappears. Headers and footers are often skipped or stored separately, so contact details placed there may never reach the contact fields.

The fix is dull and reliable: one column, contact details in the body under your name, no tables for layout, no graphics that carry information.

What breaks parsing: headings, dates and job blocks

Parsers map sections by their headings, so use the standard ones: Summary or Profile, Experience or Work Experience, Education, Skills, Certifications. “Where I've been” or “My toolbox” may be read as part of the previous section. Within experience, keep each job block in the same order (title, employer, location, dates) and give every job both a start and an end.

Dates are the most common field error. Use one format throughout, with months: “Mar 2021 – Present” or “03/2021 – 06/2023”. Avoid seasons (“Spring 2021”), year-only ranges when months are known, and words like “now” or “ongoing”. Write “Present” for your current job.

File formats

A PDF that contains real text is fine with every current system; so is a DOCX. The PDFs that fail are scans (an image of a page), exports from design tools that convert text to outlines, and documents where a font was embedded so badly that letters come out as symbols. Turkish and other accented characters are a known trouble spot: if the font lacks the glyph, the parser gets a question mark.

Test it yourself. Open the PDF, select all, copy, and paste into a plain text editor. If the words appear in the right order with the right characters, a parser will read them the same way. If a posting asks for a specific format, send that format.

Keywords: findable, not stuffed

Recruiters search the database with terms from the posting: job titles, tools, certifications, sometimes with Boolean operators. The way to surface is to use those terms, spelled as the posting spells them, once or twice in natural places: the skills list and a bullet that shows you using the skill. Where an abbreviation and a full form are both common, write both the first time: “search engine optimisation (SEO)”, “Certified Public Accountant (CPA)”.

Do not hide keywords in white text or tiny fonts. The recruiter sees the parsed text, where hidden words are visible and read as an attempt to game the system. Do not use skill bars or star ratings: they carry no searchable text and no one knows what four of five stars means.

Online ATS scores

Websites that return an “ATS score” run their own parser and their own keyword heuristics and print a percentage. Employers' systems do not produce that number for recruiters. Some systems offer a match or ranking feature, but it is not the checker you used, its logic is different, and the recruiter still opens and reads the files that surface.

The useful part of such a tool is the parse preview: if it shows your employer in the title field or splits your dates, your layout has a problem that an employer's parser may share. Fix the layout and ignore the percentage.

Checklist

Before you upload, confirm the following.

• One column; no tables, text boxes or graphics carrying information.
• Name, phone, email, city and any profile link in the body, at the top.
• Standard section headings.
• Each job: title, employer, location, dates with months, in the same order every time.
• The posting's exact terms appear in the skills list or in bullets, with abbreviations spelled out once.
• A PDF with selectable text (or DOCX if asked); copy-paste test passed.
• File named with your name and the word résumé or CV, such as Jane-Okafor-Resume.pdf.