--- title: "How to Write an ATS-Friendly Resume (Beat the Bots in 2026)" description: "ATS has evolved beyond simple keyword matching. Learn how modern Applicant Tracking Systems use semantic matching, what formatting still gets you auto-rejected, and how to optimise your resume for both smart and basic ATS." canonical: "https://mortit.com/blog/ats-friendly-resume-guide" --- Resume Writing # How to Write an ATS-Friendly Resume ATS has changed more than most people realise. Here's what actually matters in 2026 - and what advice is stuck in 2019. 12 min read Updated February 2026 TL;DR Most ATS advice online is outdated. Modern systems like Workable and Greenhouse now use **semantic matching** that understands related concepts - not just exact keywords. But plenty of systems still use basic keyword matching, and **you don't know which one you're applying through**. The smart play: mirror the job description's exact terms (for basic systems) AND write **context-rich bullet points** that show depth (for semantic systems). Formatting basics still matter - **single-column, standard fonts, no tables or graphics, .docx format**. ## What ATS Actually Does (And It's Changed More Than People Realise) Most advice you'll read about **Applicant Tracking Systems** online is stuck in 2019. People still talk about ATS like it's a glorified Ctrl+F. And some of the cheaper systems? That's basically what they are. Literal keyword matching. Very basic. But the bigger platforms - **Workable, Greenhouse, iCIMS**, and others - have moved on. Many of them now use something called **semantic matching**. What this means is the system doesn't just look for the exact word you typed. It understands, to some degree, that "project management" and "programme management" are basically the same thing. Or that someone who lists "data visualisation" probably also knows their way around dashboards, even if they didn't spell it out. These systems turn your resume and the job description into what's called **embeddings** - mathematical representations of meaning. Then they compare how similar those two chunks of meaning are. It's not perfect. It's not a human reading your resume with empathy and nuance. But it's a lot smarter than pure keyword matching was. #### Important Not every company is using the fancy stuff. Plenty of smaller businesses, government agencies, and older enterprise setups are still running basic ATS that does literal text matching and nothing more. **You don't know which one you're applying through.** So the smart play is to optimise for both. Here's what most ATS - smart or basic - is doing with your resume: 1 #### Parsing Ripping your resume apart into structured data. Name, contact info, work history, skills, education. If it can't parse a section, that section basically doesn't exist to the system. 2 #### Matching Comparing your resume against the job requirements. Basic systems do literal text matching. Smarter ones use semantic similarity - meaning they can recognise related concepts and synonyms, not just identical words. 3 #### Ranking Scoring you based on how well you match. Recruiters typically only look at the top 10-20 candidates. Everyone else might as well have not applied. ## The Formatting Mistakes That Still Get You Auto-Rejected This stuff hasn't changed much. The newer AI-powered systems are better at handling messy formatting than the old ones, but why risk it? These are all easy fixes. Even the semantic matching can't save you if the **parser mangled your text** before the matching engine ever sees it. ### Tables, Columns, and Text Boxes ATS reads left-to-right, top-to-bottom. Multi-column layouts get scrambled. Your carefully designed two-column resume? The parser might read "Senior Marketing" from column 1 and then jump to "Bachelor of Science" from column 2. Total gibberish. ### Headers and Footers A lot of systems straight-up ignore them. If your name and contact info are in the header, you've just submitted an anonymous resume. ### Graphics, Icons, and Images That little skill bar showing you're "85% proficient in Python"? The ATS sees nothing. Completely invisible. The AI-powered systems don't magically read images either - they're still working with the parsed text. ### Creative Section Titles **Use This:** Work Experience **Not This:** Where I've Made an Impact Creative titles mean the parser can't categorise your work history. Even the smarter systems rely on the parser getting the structure right first. Stick to the standard headings: **Work Experience, Education, Skills, Certifications**. It's not sexy. Neither is unemployment. ### File Format Submit as **.docx** unless they specifically ask for PDF. Some older systems still struggle with PDFs. If they give you a choice, .docx is the safe bet. The newer systems handle PDFs fine, but you don't know what you're dealing with on the other end. ## The Keyword Strategy (Updated for Systems That Actually Think) This is where things have gotten more interesting. The old advice was "mirror the job description exactly." That's still good advice - but it's not the whole picture anymore. ### For Basic Keyword-Matching Systems You still want to: - Copy the job description into a doc. Highlight every skill, tool, qualification, and requirement they mention. - **Mirror their exact language.** If they say "project management", don't write "managed projects". If they say "Salesforce", don't write "CRM software". - Include both the acronym and the full term. Write "Search Engine Optimisation (SEO)", not just "SEO". Some systems look for one but not the other. ### For Semantic Matching Systems Here's what matters more now: **context**. These newer systems don't just count whether a keyword appears. They look at the meaning around it. **Weak (keyword only):** Skills: Agile **Strong (context-rich):** Led a cross-functional team of 12 through an Agile transformation that reduced delivery time by 30% The embedding models pick up on the richness of that sentence. They understand you didn't just know the word Agile - you actually did something with it. What this means practically: include the exact keywords (because you might be dealing with a basic system), but **also make sure your bullet points tell actual stories**. The semantic models reward depth. A skills dump at the bottom of your resume still matters for the literal matchers, but the meat of your experience section is what the smarter systems really focus on. #### Don't Keyword Stuff Keyword stuffing is even more pointless than it used to be. The old trick of pasting the job description in white text at the bottom? Most systems catch that now. The semantic models don't care about frequency - they care about **relevance and meaning**. Repeating "machine learning" fourteen times doesn't make you more of a match than using it once in a meaningful context. A recruiter does eventually read the resumes that get through, and the AI screening is now smart enough to flag stuffing too. If you want to dig deeper into which keywords matter most for your industry, check out our [resume keywords by industry](https://mortit.com/blog/resume-keywords-by-industry) guide or learn [how to match your resume to specific job descriptions](https://mortit.com/blog/how-to-match-your-resume-to-jobs). ## The Format That Consistently Works After a lot of trial and error, this is the safest setup. It looks boring. It is boring. It works. #### ATS-Safe Resume Format - Single column layout - Standard fonts (Arial, Calibri, Garamond) - Clear section headings (Work Experience, Education, Skills) - Reverse chronological order - Bullet points using normal bullet characters - Your name and contact info in the body of the document (not the header) - No images, no graphics, no tables - File name: FirstName\_LastName\_Resume.docx Not sure which overall format works best for your situation? Our [guide to resume formats in 2026](https://mortit.com/blog/best-resume-format) breaks down when to use chronological, functional, or combination layouts. ## So What Does This Actually Mean for You? The game hasn't completely changed. Good formatting still matters. Keywords still matter. [Tailoring your resume](https://mortit.com/blog/resume-writing-guide) to each role still matters - probably more than anything else. What *has* changed is that the smarter systems now give you a bit more room to breathe. You don't need to be quite as paranoid about using the exact precise wording from the job ad, because some systems will understand synonyms and related concepts. But "some systems" is doing a lot of heavy lifting in that sentence. You have no idea whether the company you're applying to is using cutting-edge semantic AI or something that was built in 2015 and hasn't been updated since. **So the move is: write for both.** Mirror the key terms from the job description (covers you for the basic systems). But also write substantive, context-rich bullet points about your actual experience (which the smart systems reward, and which humans like to read too). You basically need a different version of your resume for every job, or at least for every type of role. A generic resume that tries to cover everything will usually lose to a tailored one that mirrors the exact job description. Yes, it's time-consuming. Yes, it's annoying. But that's the reality of how the system works right now. **Quick ATS Test:** Take your current resume, copy-paste it into a plain text editor (Notepad, not Word). If it's still readable and everything's there in a logical order, it's probably ATS-friendly. If it's a mess, that's roughly what the ATS is seeing. ## The Tailoring Problem (And How to Solve It) Here's the catch: every job description is different. The keywords that work for one application may not work for another. And now you need to optimise for *both* basic keyword matching and semantic depth. Manually tailoring a resume takes 30-60 minutes per application. Do that for 20 jobs and you've spent an entire work week just on resumes. This is exactly what [MORT's AI Resume Builder](https://mortit.com/features/resume-builder) solves. It reads the job description, identifies the important keywords, and adjusts your resume's summary, skill order, and wording to match - in about 2 minutes. It handles both sides of the equation: mirroring exact keywords for basic ATS *and* strengthening context around your experience for semantic systems. Same optimisation, without the manual work. ## Stop Guessing What the ATS Wants MORT's AI Resume Builder analyses job descriptions and creates tailored, ATS-optimised resumes that work with both basic keyword matchers and modern semantic systems. Upload your resume once - get a tailored version for every job you apply to. [Learn About Resume Builder](https://mortit.com/features/resume-builder) [Try Free Resume Builder](https://app.mortit.com/signup) ## Keep Reading ### [Complete Resume Writing Guide](https://mortit.com/blog/resume-writing-guide) Everything you need to write a resume that works ### [Resume Keywords by Industry](https://mortit.com/blog/resume-keywords-by-industry) The exact keywords to boost your ATS score ### [How to Match Your Resume to Jobs](https://mortit.com/blog/how-to-match-your-resume-to-jobs) Tailor your resume for each application ### [Best Free AI Resume Builders](https://mortit.com/blog/best-free-ai-resume-builders) ATS-friendly free options with real AI tailoring