Letterin
Guides

What is an ATS, and how does it read your CV?

Applicant tracking software sits between you and the recruiter. It rarely rejects anyone on its own — but what it fails to read, the recruiter never searches, and that is the part you can fix.

In short

An ATS is a database, not a judge

An ATS (applicant tracking system) is the software a company uses to collect, store and search applications. It parses your file into fields — name, employer, title, dates, skills — and files the result.

The popular claim that these systems auto-reject three quarters of applicants is mostly a myth. Knock-out questions exist (work authorisation, a required licence), but the usual damage is quieter: a recruiter searches the database for a term, and a CV the parser read badly never appears in the results.

Where CVs actually break: the text layer

A PDF carries two things — the page you see and the text underneath it. If that text layer is missing or damaged, the file looks perfect on your screen and reads as nothing to a machine.

Two causes are common. A CV exported as an image, or scanned, has no text at all. And some design tools embed fonts in a way that loses the character map, so the extracted text comes back with digits and symbols inside words while the page still looks flawless. Copy a paragraph out of your own PDF and paste it into a plain text editor: what you see there is what the ATS sees.

Layout the parser reads in the wrong order

Parsers read a page in the order the text was written, not the order your eye follows. Two columns, tables, text boxes and content in headers or footers are where an entry loses its dates or a job title lands on the wrong employer.

Graphics that carry meaning fail for the same reason: a five-dot proficiency meter, a bar chart of skills or an icon standing in for a phone number is invisible in the text layer. If a value matters, it has to exist as real text beside the graphic.

Keywords are matched, not understood

Recruiter searches are literal. If the posting says "Power BI" and your CV says "data visualisation tools", you do not match the search, even though a person reading it would.

The fix is not keyword stuffing, which reads badly to the human on the other end. It is using the posting's own words for the work you genuinely did, and spelling out both forms of an acronym at least once: "Search Engine Optimisation (SEO)".

How to check your own CV

Three checks catch most of it. Copy-paste the whole PDF into a text editor and read what comes out. Confirm every date, employer and title survived. Then compare the terms in the posting with the terms in your CV, one by one.

Letterin does all three for free at /roast — it reads your file once to score it and hand you an honest roast alongside. You can also build the CV itself here, and every export carries a real, selectable text layer.

Questions

Does an ATS automatically reject my CV?

Usually not. Most systems score and sort rather than reject, and the filtering is done by a person searching the database. The exception is explicit knock-out questions an employer configures, such as work authorisation or a required certification.

Is PDF or Word safer?

A PDF with a real text layer is safe with every modern system, and it preserves your layout. A .docx is still asked for by some career portals, especially in Türkiye. Keep both; never send a scanned or image-only PDF.

Do two columns really break parsing?

Not always, but they are the most common cause of entries merging or losing their dates. If you want a two-column layout, check the extracted text before you send it — and keep a single-column version for portals that ask you to upload rather than apply directly.

Letterin builds your CV with a selectable text layer, scores its ATS readability before you send it, and rewrites it for the posting without inventing anything.

Try it free