43% of resumes contain at least one formatting element that causes ATS parsing failures, according to a 2024 Jobscan analysis. Text boxes are the worst offender - parsers either append their content out of order or drop it entirely. Contact information in a Word header looks identical on screen but is invisible to most ATS systems. A quick test: paste your resume into plain Notepad. If anything is missing or out of order, it will fail ATS parsing too.
A resume parsing error does not announce itself. You do not get a rejection email that says “your resume data was corrupted.” You just do not hear back. The ATS received your file, extracted what it could, and either gave you a low score based on incomplete data or filed you as a contact with no discernible qualifications.
Understanding exactly which document structures cause parsing failures - and why - lets you eliminate these errors before they cost you.
What ATS Parsing Actually Is
When you upload a resume, the ATS does not read it the way you do. It runs a text extraction process that converts your document (PDF, DOCX, or whatever format) into raw text. Then a parsing engine analyzes that raw text, trying to identify which characters are your name, which are a job title, which are a date, which are a bullet point describing an achievement.
The extraction quality depends on the document structure. A cleanly structured DOCX extracts as perfectly ordered text. A poorly designed PDF can extract as scrambled characters, missing lines, or text blocks in the wrong sequence.
The parsing quality depends on the linguistic patterns the ATS recognizes. Standard section headers, standard date formats, and standard bullet point structures parse correctly. Non-standard structures get categorized incorrectly or skipped.
Most candidates never test what their resume looks like after extraction. They see the formatted version and assume that is what the ATS sees.
Text Boxes: Where Data Goes to Die
Text boxes are objects embedded in your Word document or PDF. You create them by going to Insert → Text Box, or by using a template that places content in floating boxes for visual layout.
The problem: when an ATS parser converts your document to text, it processes the main document body first, then attempts to process embedded objects separately. The processing order for embedded objects is implementation-dependent - some parsers get it right, some append text box content to the end of the document, some drop it entirely.
The most common failure pattern: your contact information is in a text box at the top of the page. It looks like it is at the top. The ATS extracts it and appends it after your last bullet point, or does not extract it at all. Your name and email are either in the wrong place or absent.
The same problem affects any content placed in a text box: a sidebar with your skills, a contact section in a graphical callout, a “featured accomplishments” box near the top of the page.
How to detect text box usage: Open your Word document and press Ctrl+A to select all. If any content does not get highlighted, it is in a text box or other floating object.
The fix: Delete the text box. Paste its content as regular paragraph text in the appropriate place in your document body. Reformat to look the same visually. The ATS will now see it.
Headers and Footers: The Invisible Zone
Your word processor’s header and footer are document zones outside the main body. Page numbers typically live in the footer. Some resume templates place your name and contact information in the header, repeating it on every page.
Most ATS parsers extract body text only. Headers and footers are not parsed. Your name and contact information are invisible to the system.
This is one of the most common parsing errors and one of the most costly, because without your contact information, there is no way to follow up even if the parser somehow gives you a high match score.
How to detect header/footer usage: In Microsoft Word, double-click the very top of the page. If a dashed border appears around the top area and the rest of the document grays out, you are editing the header. Your content is in the header zone.
The fix: Cut the content from the header. Close the header editing mode. Paste the content as regular paragraph text at the top of the first page. Set the document’s actual header to blank (no content). Your name and contact details will still appear at the top visually, but now as body text that parsers can see.
Tables: The Column Scrambling Problem
Tables are the most widely misunderstood formatting element in resume design. They are popular because they create clean visual alignment, particularly for dates and role titles. They cause parsing problems because ATS parsers read tables differently than humans do.
Reading order in a table:
A human reads a table as a grid with rows. The parser often reads it sequentially through the XML structure, which can go: entire left column first, then entire right column, or: row by row but treating each cell as a separate text block, breaking the continuity of your bullet points.
Common failure pattern with a two-column table:
Left column contains: “Senior Analyst | Accenture | Managed $3M consulting projects” Right column contains: “2022 – 2024 | Chicago, IL”
The parser extracts: “Senior Analyst Accenture Managed $3M consulting projects 2022 – 2024 Chicago, IL” - or worse, separates the columns entirely, processing all left-column entries first and all right-column entries second.
The fix: Eliminate tables from your experience section. Use a simple structure:
Senior Analyst, Accenture - Chicago, IL (January 2022 – March 2024)
• Managed $3M portfolio of consulting engagements across three business units
• Led team of 8 analysts during peak delivery periodsThis structure - title, comma, company, em dash, location, parenthetical date - parses correctly in every major ATS platform. It requires no special formatting. Use tabs for indentation in bullet points, not table cells.
Multi-Column Layouts: The Left-Right Sweep Problem
Multi-column page layouts, where you have a main content column on the right and a sidebar on the left (or vice versa), create a specific and well-documented parsing failure.
Most ATS parsers sweep the page from left to right, top to bottom, across the full page width. In a two-column layout, this means:
Page line 1: [Left column: “Skills”] [Right column: “Senior Product Manager”] Parser reads line 1 as: “Skills Senior Product Manager”
Page line 2: [Left column: “Python”] [Right column: “Google - San Francisco”] Parser reads line 2 as: “Python Google - San Francisco”
The result is interleaved content from both columns, completely scrambled. Your job title is next to a skill. Your company is adjacent to a programming language. The parser cannot make sense of any of it.
The fix: Use a single-column layout for any resume that will go through an ATS. The visual appeal of multi-column layouts is real - for human readers. For ATS screening, it is a liability.
If you want to use a visually impressive design for networking events, industry conferences, or direct human submission via email, maintain a separate “human-optimized” version. Keep your ATS version single-column.
Special Characters and Symbol Encoding
Certain characters cause extraction errors in some PDF rendering engines and ATS parsers. The risk is not universal - most modern systems handle Unicode correctly - but specific character types create problems in specific situations.
High-risk characters:
- Em dashes (-) in some older PDF renderers extract as question marks or nothing. Use en dashes (–) for date ranges, which render more reliably.
- Smart (curly) quotes ("") sometimes extract as garbled characters in Taleo and iCIMS. Use straight quotes (”) if you need quotation marks.
- Bullet point types: standard filled circles (•) are fine. Arrows (→), custom symbols (★), and icon characters from symbol fonts may extract as empty boxes or nothing.
- Accented characters in language skills sections sometimes cause encoding issues. “Bilingual in Spanish and French” is safer than including special characters like “é” or “ñ” if you are uncertain about the ATS’s character encoding.
The test: Save your resume as a .txt file or paste its content into Notepad. If all your characters appear correctly in that stripped format, they will parse correctly in the ATS. If you see question marks, boxes, or garbled symbols, fix those characters before submitting.
PDF Version and Creator Tool Issues
Not all PDFs are created equal. A PDF created by Adobe Acrobat’s “Save as PDF” function has a clean text layer with extractable characters. A PDF created by printing to a PDF driver may create a rasterized image with no text layer - it looks like text but is actually a picture of text.
An image-based PDF parses as completely empty. The ATS receives a blank document.
How to check: Open your PDF in any PDF viewer. Try to highlight text by clicking and dragging. If you can highlight individual words and letters, the text layer exists and is extractable. If clicking produces no selection or selects the entire page as an image, you have an image-based PDF.
Common causes:
- Scanning a printed resume
- Creating a resume in design software (Canva, Adobe Illustrator) and exporting as PDF without proper text layer export
- Some MacOS PDF conversion tools produce flattened PDFs
The fix: Create your PDF by saving directly from Microsoft Word (File → Save As → PDF) or from Google Docs (File → Download → PDF Document). These methods preserve the text layer. If you use Canva or similar design tools, export using their “PDF Print” option, which typically preserves text layers better than their standard PDF export.
What to Do Before Your Next Submission
Three quick checks before every application:
Check 1 - Plain text test: Copy all your resume content and paste into Notepad (Windows) or TextEdit in plain text mode (Mac). Read it from top to bottom. Does your structure survive? Are dates next to the right employers? Are your bullet points intact? If yes, your parsing will be clean. If the text is scrambled, find and fix the structural issue.
Check 2 - Header/footer check: Open your Word document and press Ctrl+Home to go to the very beginning. Are the first visible words your name and contact information? Click once on your name. Does a “Header” label appear? If so, move your content to the document body.
Check 3 - Object selection check: Press Ctrl+A in your Word document. Does all your content highlight? Any unhighlighted content is in a floating object (text box, shape) and may not parse.
These three checks take under five minutes and will catch the majority of parsing errors before they cost you an interview.