ATS-Friendly Resumes: How to Pass the Automated Screen

1. How ATS scoring actually works

An applicant tracking system does three things: it parses your document into structured fields (name, job titles, dates, skills), it indexes the text so recruiters can search it, and in some systems it scores your application against the job description. The score is usually based on keyword frequency and the presence of specific phrases from the job posting.

Most ATS platforms are not sophisticated AI. They are database tools with pattern-matching. They look for headings they recognise ('Experience', 'Education', 'Skills'), pull dates out of consistent formats, and flag resumes that contain the exact words a recruiter typed into a search bar.

Understanding this makes the advice obvious: write real sentences, use standard headings, use the words the job description uses, and format the document so a text-extraction tool can read it linearly from top to bottom. That is the entirety of ATS optimisation done properly.

2. Reading a job description for the keywords that matter

Not all words in a job description are equal. Ignore the boilerplate ('fast-paced environment', 'self-starter') and focus on three types of keywords: role-specific skills and tools (React, IFRS, Salesforce, ISO 27001), action phrases that describe what the role does (stakeholder management, financial modelling, product roadmap), and qualifications or credentials that are listed as requirements rather than nice-to-haves.

If a skill appears multiple times or is listed under 'essential requirements', it is a high-priority keyword. If your background genuinely includes it, make sure the same phrase — not a synonym — appears in your CV. 'Revenue forecasting' and 'sales forecasting' are not the same string to a database.

A useful method: copy the job description into a text editor, delete the generic filler, and what remains is your keyword list. Compare it against your CV and close the gaps honestly. Do not add skills you do not have — you will be asked about them in an interview.

3. File format rules — PDF vs docx

The conventional advice used to be 'always send a docx because older ATS cannot read PDFs.' That advice is outdated. Modern ATS platforms — Workday, Greenhouse, Lever, iCIMS, Taleo — all parse PDFs reliably. Unless the application form explicitly asks for a Word document, a PDF is the better choice: it renders identically on every device and cannot be accidentally edited.

What does still cause real problems: PDFs that are actually image scans rather than text-based documents. If you scanned a printed CV and saved it as a PDF, the ATS cannot read a single word. Open the file, try to select and copy a sentence — if you cannot, neither can the parser. Export from a proper CV builder or word processor to produce a text-based PDF.

The filename matters too. 'CV.pdf' tells a recruiter nothing when they have 200 files in a folder. Use a clear naming convention like 'AmeliaClarke-ProductManager-CV.pdf'. It is a small thing that occasionally makes a real difference.

4. Use a clean, single-column structure

Heavily designed multi-column layouts, text inside images, and unusual fonts can confuse older parsers and even some modern ones. The issue is that ATS tools parse text in a single reading order — left to right, top to bottom. A two-column layout means the parser may read the left column of every section before it reads the right, which can produce nonsense field assignments.

The safest approach is a clear hierarchy of standard headings — Experience, Education, Skills — with real selectable text underneath. You can still have a visually polished document; you just need to achieve that polish through typography, spacing, and colour rather than through floating boxes and sidebar columns.

Templates designed with parsing in mind give you visual polish without breaking the reading order. Most well-built CV builders produce this by default.

5. What happens to multi-column layouts

Here is a concrete example of what goes wrong. Suppose your CV has a two-column layout where column one lists your job titles and dates, and column two lists your bullet points. When an ATS reads the raw text, it may extract: 'Senior Engineer 2019–2022 Product Manager 2022–2024' followed by all the bullet points from both roles without any connection between them. The result is a structured record with no clear role-to-achievement mapping.

Worse, skills listed in a sidebar are sometimes ignored entirely or appended to the last job entry, making it look as though you learned all your skills in a single role. If a recruiter searches for 'Python' and your Python experience is sitting in a sidebar that was not parsed correctly, your application will not appear in results.

The fix is not complicated: move to a single-column layout. You lose nothing of substance and gain reliable parsing across every platform the recruiter might be using.

6. Mirror the language of the job description

An ATS ranks candidates partly on keyword relevance. Read the job posting carefully and use the same terms it uses — if it says 'stakeholder management', use that phrase rather than a synonym, provided it is true of you. Spell out acronyms at least once, for example 'Search Engine Optimisation (SEO)', so both the long and short forms are searchable.

This is not about gaming the system — it is about speaking the same language as the team that wrote the description. They used 'data visualisation' rather than 'data reporting' for a reason. Matching their vocabulary signals that you understand the role and the field.

One area where many candidates lose points: job title alignment. If you held the title 'Growth Analyst' but are applying for a role called 'Marketing Analyst', and your CV never uses the phrase 'marketing analysis', a keyword search for that phrase will not surface you. Where your past work genuinely overlapped with the target role's vocabulary, use both.

7. Keep headings and dates conventional

Use ordinary section titles. A creative heading like 'Where I've Made Magic' may read well to a human but can prevent the system from recognising your work history section at all. Stick to standard titles: Work Experience (or Experience), Education, Skills, Certifications, Languages.

Write dates in a consistent format throughout — 'Jan 2021 – Mar 2023' or '01/2021 – 03/2023', but not a mixture of both. Month-Year is preferable to year-only because it gives the parser enough information to calculate duration. Avoid burying dates in paragraphs rather than displaying them prominently next to job titles.

• Save and submit as PDF unless the posting specifically asks for .docx.
• Avoid headers and footers for critical information — some parsers ignore them.
• Do not put key content inside graphics, icons, text boxes, or tables.
• Use standard section headings that any recruiter would recognise.
• Write dates consistently in Month-Year format next to each role.
• Spell out acronyms at least once so both forms are searchable.

8. Testing your own CV before you apply

You can approximate what an ATS sees by running a simple test. Open your CV PDF, press Ctrl+A to select all, then copy and paste the full text into a plain text editor. Read through what you get. If it is readable prose in a logical order — contact details, summary, then each role with its bullets — your parsing will likely be clean. If you see columns merged together, bullets from different jobs jumbled, or words from a text box appearing somewhere random, fix the layout.

Several free tools let you paste a job description alongside your CV and show you the keyword gap. These are useful for a quick check before each application. Do not optimise for the tool — optimise for the real job description, and use the tool to sanity-check your work.

Ask someone who has never seen your CV to read the plain-text version and tell you what role and seniority they would expect you to be applying for. If the answer does not match the job you are targeting, the problem is almost certainly in the summary or the top bullet points.

9. What not to do — keyword stuffing and gimmicks

Some guides suggest pasting hidden white-text keywords or repeating terms unnaturally. Modern systems and recruiters catch this, and it reads as dishonest. Keyword stuffing also produces bizarre results when a human reads the document — and a human always reads it eventually.

Similarly, 'optimising' your CV by making it a wall of keywords without context helps no one. A recruiter searching for 'SQL' who finds it listed seventeen times in a skills section but never evidenced in an actual role will assume you are padding, not skilled.

Earn relevance the right way: describe genuine, relevant achievements using the vocabulary of your field. A single bullet point that demonstrates you used SQL to build a reporting pipeline that saved 10 hours a week is worth more than SQL listed twenty times in isolation.

10. ATS-readiness checklist

Run through this before every application. It takes under five minutes and closes the most common reasons CVs fail automated screening.

• Document is exported as a text-based PDF (text is selectable, not a scanned image).
• Layout is single-column with no text boxes, tables, or sidebar columns in the main content.
• Section headings use standard names: Experience, Education, Skills, Certifications.
• Key contact details are in the body of the document, not only in the header or footer.
• Dates are in a consistent Month-Year format next to each job title.
• All keywords from the 'required' section of the job description appear naturally in the document.
• Acronyms are spelled out at least once.
• Plain-text copy-paste test passes — content reads in logical order.