Your designed CV looks great and scans terribly
The layout is the problem, not your experience. See what a parser actually pulls out of your file, then rebuild it to read cleanly.
The rejection you never hear about
Before a human reads a word of your CV, software converts your file into plain text. That conversion is dumb on purpose: it walks the document, grabs whatever text it can find, and hands over a stripped string. Anything that isn't text in the main body of the document is at risk of vanishing, and nobody tells you it vanished.
Designed templates are built for the eye, not for that conversion. A side panel puts your skills in a separate column, so the parser may read straight across the page and glue a job title to a software name. A header puts your name and phone in a zone many parsers skip entirely. Icons replace the words "Email" and "Phone". Skill levels become dots, and dots contain no text at all.
The fix isn't more keywords. It's a rebuild of the file itself. Keep every achievement, every date, every number you worked hard to earn, then move it all into a single column of ordinary body text with real headings. The content stays. The layout tricks go. Your CV stops being a picture of your career and becomes a document a machine can read out loud.
Six template habits that quietly break parsing
Each one looks like good design. Each one costs you text on the other side of the upload.
The two-column layout
Parsers read left to right. Your side panel gets interleaved with your job history, and both come out scrambled.
Text boxes and shapes
Text inside a box floats outside the document flow. It can land anywhere in the extract, or nowhere at all.
Headers and footers
Your name, phone and email sit in a zone plenty of parsers skip. A nameless CV is a dead CV.
Icons instead of labels
A tiny envelope isn't the word "Email". The parser sees a picture and moves on without your contact details.
Skills rated in dots
Four filled circles say nothing in plain text. The skill either appears as a word or it doesn't appear.
A photo where the name goes
Israeli and most European CVs don't need one, and a large image at the top pushes your name out of first position.
The plain-text rebuild, step by step
Same career, same achievements, a file that survives the upload.
Collapse to one column
Delete the side panel and paste its contents back into the flow: skills under a Skills heading, languages under Languages.
Move contact details into the body
Name on the first line, then phone, email and city as ordinary text. Nothing in the header or footer.
Use real headings
Experience, Education, Skills. Plain words, styled with heading formatting, not shapes or coloured bars with text pasted on top.
Write dates as text
One consistent format per role, such as 03/2021 to 08/2024. Avoid timelines, bars and graphics that imply dates visually.
Turn ratings into words
Replace dots and bars with plain levels: advanced, intermediate, native. Better still, show the skill inside a bullet about real work.
Export a PDF with a live text layer
Open the file and try selecting a sentence. If you can't highlight it, the parser can't read it either.
Side by side
What you designed vs what the scanner extracts
| On the page | What the parser extracts | What the recruiter ends up seeing |
|---|---|---|
| Name and phone in the header | Often skipped along with the whole header zone | A CV with no candidate name attached |
| Skills in a left side panel | Read across the page, mixed into your job history | Job titles glued to software names |
| Skill shown as four filled dots | Nothing. Dots carry no characters | The skill is simply absent from your profile |
| Job title inside a text box | A loose fragment, detached from its dates | A role with no company and no timeline |
| Phone number beside a phone icon | Digits with no label to identify them | Contact details the system can't file |
| Design exported as a flat image PDF | An empty text layer | A blank record and an automatic rejection |
See your CV the way the scanner sees it
Upload your file and get the extracted text back, plus the exact layout elements that are costing you content.
Questions people ask before they rebuild
Do applicant tracking systems really reject two-column CVs?
They rarely reject them outright. They mangle them. Some modern parsers detect columns correctly, plenty still read straight across the line and merge your side panel into your job history. You can't tell which system a given employer runs, so the safe move is one column. You lose nothing but the design flourish.
Is PDF or Word safer for CV parsing?
A PDF with a real text layer parses reliably and looks identical everywhere, which is why most Israeli and international employers prefer it. A Word file parses fine too, but formatting can shift on someone else's machine. The genuinely dangerous file is an image-only PDF exported from a design tool, since it contains no readable text at all.
Should I put a photo on my CV?
In Israel and across most of Europe, no. It adds no information, it raises bias questions, and a large image at the top of the page often displaces your name from the first position the parser looks at. If you're applying somewhere a photo is genuinely standard, keep it small and put your name above it.
Won't a plain one-column CV look boring to a recruiter?
A clean CV doesn't look boring, it looks senior. Recruiters skim for job titles, dates and numbers, and a single column delivers those faster than any side panel. You still get typography, white space and bold headings. What you drop are the elements that were never adding meaning anyway: icons, dots and decorative bars.
How do I check whether my CV parses correctly?
Quick test at home: open the PDF, select all, paste into a blank text file and read the result. If sections are scrambled or your contact details are missing, so is the parsed version. For a fuller picture, HRLens analyses the file for free and lists the specific layout elements breaking the extraction.