How applicant tracking systems read a CV

Start here

If you have never thought about the machine reading your CV, read these two first.

• What an ATS is, and what it actually does

  The short version: it stores applications and pre-screens them. It is a filter, not a judge, and most of what it rejects it rejects for mechanical reasons.

• What makes a CV ATS-friendly

  The properties that survive parsing, ranked by how often they go wrong, with what each one is worth in Appliora's published scoring model.

How parsing goes wrong

Four failure modes account for most of the damage, and none of them are visible when you look at the PDF.

• Reading order and two-column CVs

  Why a sidebar can interleave itself with your job titles, how to tell whether it happened to your file, and when two columns are safe.

• The text layer, and why a scan scores near zero

  A PDF that is a picture of a CV contains no text at all. This is the one problem that caps a score outright rather than costing points.

• PDF or DOCX: which to send

  What each format gives a parser, which one the advert is really asking for, and the one case where the answer is neither.

• Tables, icons, text boxes and rating bars

  The design features that look professional and read as nothing. Includes what happens to contact details in a header.

Keywords and matching

The half of the problem that is about what you wrote rather than how it was laid out.

• ATS resume keywords, without the stuffing

  Where keywords come from, which ones count, and why a wall of terms scores worse than a dozen placed inside your experience.

• How to pull requirements out of a job advert

  Separating the must-haves from the wish list, and what to do with a requirement you only half meet.

• Tailoring a CV to one advert

  A repeatable 20-minute pass that changes what is said and what is emphasised, without inventing anything.

• What an ATS score actually means

  What a number out of 100 can and cannot tell you, and why no score is a prediction of an interview.

Evidence rather than folklore

Most ATS advice on the internet repeats a handful of statistics whose original source nobody can find. Appliora publishes what it can actually measure instead.

• The Appliora CV parsing benchmark

  One reference CV exported through every template in the collection and run through the same parser, with the method and the limits written out. It reports how Appliora's parser behaves, and says so - it is not a claim about any employer's system.

• The checker itself

  14 checks with every weight published, the extracted text shown back to you, and no signup. Run it as many times as you like.

What this cluster does not claim