Test of 20 resume layouts finds contact details and section order lost in some formats
A tester rendered one synthetic resume in 20 layouts, spanning 14 PDFs and 6 DOCX files, and ran each through the text extraction libraries used by CVPosh's free ATS checker. The team counted how many of 37 known content items survived extraction. Contact details placed in DOCX headers or footers, drawn as images, or hidden behind link labels were lost in 4 of the 20 files. A resume saved as a single page image lost all content, while columns and positioned blocks kept the words but sometimes scrambled reading order. The authors stress that a high checker score does not guarantee clean parsing, and that the test covered only text extraction, not any employer's applicant tracking system.
This is an AI-generated summary. ShortSingh links to the original source for the complete article.
Discussion (0)
Log in to join the discussion and vote.
Log in