Resumes · 6 min read
Formatting a resume so the software reads it properly
Formatting advice for applicant tracking systems tends toward the extreme: strip everything out, use no design, submit something that looks like a document from 1998. That is unnecessary. A well-designed resume parses perfectly well as long as it avoids a specific and short list of things.
Here is that list, and the reasoning, so you can make your own judgements rather than following rules.
The file itself
Use a Word document or a text-based PDF. Both parse well in modern systems. The test for a PDF is simple: open it and try to select the text with your cursor. If you can, so can the parser. If you cannot, it is an image and nothing will be extracted.
Do not submit an image or a scan. A photographed or scanned resume extracts nothing.
Name the file sensibly. Your name and the word resume. Not "resume-final-v4-USE-THIS". Somebody is going to see the filename.
Follow the employer's instruction if they specify a format. Some systems genuinely handle one better than another, and the employer knows which.
Layout: the one that matters most
Use a single column for your main content.
This is the single most important formatting decision. Parsers read a document left to right across the full width. A two-column layout with a skills sidebar means the parser may read across both columns simultaneously, producing text that interleaves your sidebar with your work history into something unusable.
It is the most common cause of a badly parsed resume, and it is the most popular design choice in modern templates, which is why the two keep colliding.
If you want visual structure, get it from spacing, weight and clear headings rather than from columns.
Where text lives
Keep contact details in the body of the document. Some parsers ignore headers and footers entirely. If your phone number and email are in a header, they may be extracted as nothing, and you are the person who has to notice that.
Do not put content inside text boxes. Text boxes are separate objects in the document structure and are frequently skipped.
Do not use tables for layout. Tables used to position content produce unpredictable reading order. A table genuinely presenting tabular data is less of a problem, but avoid it for structure.
Do not put text inside an image. A skills section designed as a graphic with proficiency bars contains no text at all as far as the system is concerned. Those bars also communicate nothing meaningful to the human reading it, since nobody knows what seven out of ten in Excel means.
Headings and structure
Use conventional section headings. Experience or Work Experience. Education. Skills. Parsers look for recognisable headings to segment the document, and creative alternatives such as "Where I Have Made An Impact" defeat that.
You can be distinctive in how the document looks. Be conventional in what the sections are called.
Use a consistent structure for each role: job title, employer, location, dates, then what you did. Same order every time, so the pattern is obvious to both the parser and the reader.
Dates and titles
Use month and year. "March 2021 to June 2023" parses reliably. "Spring 2021" and "2021-23" do not.
Do not leave dates out to hide a gap. Missing dates are more noticeable than a gap, and a gap is a question rather than a problem. See explaining a career gap.
Use the recognised job title. If your internal title was unusual, put the standard equivalent and add yours beside it: "Customer Success Manager (internally: Client Happiness Lead)". Recruiters search for the standard term.
Fonts and visual choices
Use a standard, readable font. Anything common is fine. Avoid anything decorative or very light.
Ten to twelve point for body text. Smaller to fit more in makes it harder to read, and reading is the point.
Keep some white space. A dense wall of text is skipped rather than read.
Colour is fine in moderation, for headings or a rule. Keep body text dark on light.
Do not include a photograph unless it is genuinely conventional in your market and industry. In many countries it is discouraged specifically because it invites bias, and some employers remove them before review anyway.
The ten-minute check
Before you submit:
1. Save as PDF. Open it. Select all the text and copy it. 2. Paste it into a plain text editor. 3. Read what appears.
If your work history is in order, your contact details are present, and nothing is interleaved or missing, it will parse correctly. If the text is scrambled, you have found a column or a text box.
This takes a couple of minutes and it is the only check that actually tells you anything. It beats every list of rules, including this one.
Common questions
Should a resume be PDF or Word?
Either, as long as the PDF is text-based rather than an image. If the employer specifies, follow that.
Can I use a two-column resume?
It is the most common cause of parsing problems. If you want to, run the copy-and-paste test first and check the result is readable.
Do resume templates work with an ATS?
Most do. The ones that cause trouble use multi-column layouts, text boxes, tables for layout or graphics containing text.
Should I include a photo on my resume?
In most English-speaking markets, no. It is discouraged because it invites bias, and some employers strip them out before review.
How do I check my resume parses correctly?
Save as PDF, open it, select all, copy, and paste into a plain text editor. If the result reads correctly, the parser will read it correctly too.
The short version
Single column, contact details in the body, no text boxes or layout tables, no text inside images, conventional headings, month-and-year dates, and a text-based file. Then run the copy-and-paste test, which is worth more than any rule.