Parsing failures waste your time and cost placements. Understanding how ATS CV parsing actually works helps you fix the root causes and avoid sending unreadable CVs to clients.
TL;DR: ATS parsers try to extract text and structure from CV files using one of four main technologies. They fail mostly because of six common layout traps, like tables or multiple columns. Recruiters can reduce errors by standardising CV intake and stripping problematic formatting. Distill fits best by cleaning and reformatting CVs automatically before submission.
What an ATS parser is (and isn't)
An ATS CV parser is software that scans a CV file and pulls out key details: name, contact info, skills, employment history, education, and dates. It organises this data into fields for searching and matching.
But parsers aren't perfect document readers. They don't see the CV like you do. They don't judge style or tone. They're not extracting every word for reading pleasure.
Their job is mechanical: turn varied CV formats into reliable, structured data. This means they expect certain layouts and text flow. Anything outside those expectations risks misreading.
They're not magic. A CV that looks great to a human can confuse a parser if it uses tables, columns, or unusual fonts. If the parser fails, the recruiter spends extra time fixing entries or resubmitting.
The 4 parsing technologies in production today
Most ATS parsers use one or a combination of these technologies to extract data:
Rule-based parsing Uses predefined patterns and keywords to identify sections. For example, looks for "Employment History" heading, then grabs lines below. Works well for consistent, simple CVs but brittle with unusual layouts.
Regular expressions (regex) Pattern matching on text strings. For instance, regex finds phone numbers by matching digits in expected formats. Powerful but fragile if formatting changes slightly.
Machine learning parsers Trained on thousands of CV samples to recognise sections and extract fields. More flexible with varied layouts but can still be confused by complex formatting or graphics.
Optical Character Recognition (OCR) Used for scanned PDFs or images. Converts pixels to text before parsing. Accuracy depends heavily on scan quality and clean text layout.
Most ATS vendors combine rule-based and machine learning techniques. OCR is a fallback for non-text files but less reliable.
Why parsers fail: the 6 layout traps
Parsing errors come down to six common layout issues that confuse software:
Tables and grids Parsers often read tables left to right, top to bottom, but some ATS can't parse tables at all. Tables may jumble employment dates and roles, mixing columns and rows unpredictably.
Multiple columns Two or three-column CVs look neat to humans but read as jumbled text streams by parsers. They can merge columns or swap content order, losing meaning.
Headers and footers Contact details in headers or footers are frequently ignored or dropped because many parsers only scan the main body text.
Text boxes and shapes Text inside boxes or graphic elements often gets skipped or extracted as gibberish, as parsers read linear text only.
Fonts and special characters Unusual fonts, bullet styles, or special characters can break regex patterns or machine learning predictions, leading to missing or malformed data.
Embedded images or logos Parsers don't read text inside images or logos. If contact info or key data is embedded, it's lost.
Each trap increases the chance that key details go missing or mix up, forcing manual fixes.
What recruiters can do at intake to avoid failures
Fixing parsing errors after CV submission wastes time. Instead, control the format at intake:
Request standard Word or PDF CVs Avoid scanned images or unusual formats.
Ask candidates for single-column layouts without tables Clear, linear text flows parse far more reliably.
Get contact details in the main body, not headers or footers Essential info should be visible to parsers.
Avoid text boxes, graphics, and logos in CVs Keep it simple and text-based.
Use common fonts and standard bullet points Simple characters reduce parsing errors.
Check CVs quickly on submission Spot obvious formatting traps early and ask for a rewrite if needed.
This upfront effort saves hours downstream.
Where Distill fits in the workflow
Distill automates CV cleaning before submission. It strips names, emails, phone numbers, photos, and graduation years to anonymise data for compliance. But it also reformats CVs into ATS-friendly Word documents.
If you send 20+ CVs a week to clients with parsing issues, Distill removes tables, converts multi-column layouts to single column, and moves headers into the main text. This reduces parsing errors without manual rework.
Distill doesn't replace ATS parsing but prepares CVs to fit their expectations. It's a practical step between candidate submission and client upload.
FAQ
Why can’t ATS parsers read tables or columns reliably?
Most parsers read text linearly. Tables and columns break that flow. Parsers may jumble cells or miss data if they can't detect table boundaries correctly.
Can I improve parsing by converting PDFs to Word?
Sometimes. Native Word documents tend to parse better than PDFs. But poorly formatted Word files with tables or columns cause the same problems.
Are some ATS parsers better than others?
Yes. Some use advanced machine learning and handle varied layouts better. But no parser is perfect. Layout best practices still matter.
Does Distill integrate directly with ATS platforms?
Distill works independently to clean CVs before upload. It doesn't currently integrate natively with ATS platforms but fits easily into existing workflows.
What about scanned CVs or images?
OCR can help but is error-prone. Best to avoid scanned CVs where possible.
Parsing is a mechanical process with clear limits. Knowing the technologies and layout traps helps you avoid wasted time and messy data.
If you regularly send CVs to Bullhorn, SmartRecruiters, or other ATS platforms and face parsing issues, try Distill. It prepares your CVs by reformatting and anonymising them to your client's ATS specs — saving hours on manual fixes.
Try Distill free with your ATS export today.