The essentials in 30 seconds
- 01One column, standard type, literal headings.
- 02A PDF with selectable text, never a scanned image.
- 03Advert keywords placed inside your experience.
01
A robot reads it before a human does
An Applicant Tracking System pulls the text out of your PDF, splits it into fields and scores it against the advert. If it can't read a section, that section doesn't exist.
02
Single column, always
Two columns, text boxes and tables scramble the reading order during extraction. Use one column, standard fonts and literal section headings: Experience, Education, Skills.
03
No graphics or icons carrying information
Language and skill rating bars can't be read — write “English C1” instead. Never put information inside images, headers or footers.
04
Keywords come from the advert
Copy the exact terms from the job description (tools, certifications, job title) and work them into your experience naturally. No stray keyword lists at the bottom.
05
Check your score
Upload your CV to the checker: it scores you out of 100 and shows the criteria you fail, so you can fix them before applying.
06
What an ATS actually does with your PDF
An Applicant Tracking System does four things in order: it extracts the text from the file, splits it into fields (name, employers, dates, titles, skills), compares that content with the advert's requirements and ranks applications by match. The recruiter sees an already-scored list — and on high-volume roles, opens the first screen of results.
The critical step is the first one. If extraction fails — because the text sits inside an image, a table, or two columns the converter interleaves — the system doesn't score you badly: it scores you empty. A beautiful CV the machine can't read competes exactly like a blank page.
- Text extraction: only a PDF with selectable text works.
- Field parsing: it needs literal section headings.
- Comparison with the advert: it measures terms, not intentions.
- Ranking by match: it decides which CV gets opened.
07
The five formatting decisions that decide whether you're read
Single column whenever you can choose. Two columns can pass without tables or text boxes, but the reading order is more fragile and breaks on older boards. Section headings must be the literals the system expects: Experience, Education, Skills, Languages. “My journey” or “What I can do” force the parser to guess, and it guesses wrong.
No information inside graphics. Language bars, skill wheels and icons carrying your phone number are invisible to extraction: the fact has to be written. And nothing in headers or footers, because many converters discard them before they start.
- One column and a standard typeface embedded in the PDF.
- Literal section headings on their own line.
- Month/year dates, in the same format throughout.
- Written levels (“English C1”), never bars or stars.
- Contact details in the body, not in the file header.
08
Placing keywords without writing for a robot
Keywords come from the advert, not from a trending list. Copy the job description, mark the recurring nouns (tools, certifications, processes, job title) and keep ten to fifteen. Then place them where they weigh: headline, statement and the achievements of your most recent role.
Position beats quantity. “SAP” inside an achievement with a date and a result is worth far more than “SAP” in a loose list at the bottom, because the system values the term in context. Hidden text, white-on-white and footer stuffing don't work: they're detected or ignored, and some systems drop the whole application.
09
How to check it before you send anything
There's a thirty-second home test: open your PDF, select all the text, copy it and paste it into a blank document. What you see is roughly what the filter reads. If the order holds, no section is missing and the dates still sit beside their employer, the format is safe.
Then upload the file to the checker: it returns a score out of 100, the criteria you fail and the specific fix for each, including the advert keywords missing from your text.
Check this before you send it
- Text copied out of the PDF keeps its order and loses no sections.
- Section headings use the standard literals.
- No fact lives in an image, table, header or footer.
- Ten to fifteen advert terms appear in context.
- The file is under 2 MB and is a PDF, not a scan.
- The checker score is above 80 out of 100.
Mistakes that cost you the interview
Sending a scanned or image-exported PDF
The filter extracts nothing, so the application enters with a zero score and nobody ever sees it.
Using tables to lay out experience
On extraction, cells are read in an order that mixes employers with dates and leaves achievements with no role to attach to.
Creative section headings
“My professional journey” stops the system identifying the experience block, so that part of the CV goes unscored.
Keyword stuffing with no context
A block of twenty terms at the bottom hurts human readability and adds almost nothing to the automated match.
Frequently asked
Questions we get asked most
- How do I know if my CV passes the ATS?
- Copy all the text out of the PDF and paste it into a blank document: if the order holds and nothing is missing, the filter reads it too. The checker also scores you out of 100 and names the criteria you fail.
- Do two-column CVs pass an ATS?
- They can, provided there are no tables, text boxes or information inside images — but the reading order is more fragile. For a heavily contested role or an older job board, single column is the safe bet.
- Which file format does an ATS read best?
- A PDF with selectable text. Word shifts layout on another machine, and a scanned PDF is an image: the filter extracts nothing and the application scores zero.