An ATS-friendly resume is one that applicant tracking software can read correctly: a text-based, single-column file with standard headings, consistent dates and skills written as plain text. Follow the portal's file instructions; if there are none, a clean DOCX or a tested text-based PDF works.
- An ATS-friendly resume is a text-based file whose contact details, roles, dates, skills and education extract in the right order.
- Use one column, conventional headings, consistent dates and normal body text; avoid text boxes, images and complex tables.
- Follow the portal's file instructions; otherwise a clean DOCX is the conservative choice, and a text-based PDF works if you test it.
- Use the job's real terms inside evidence bullets instead of pasting keyword lists.
- Test your own final file; a third-party "ATS score" is not the employer's configuration.
What "ATS-friendly" can and cannot promise
"The ATS rejected my resume" can describe several different events. Separate them before trying to fix the document.
| Stage | What may happen | What the candidate can control |
|---|---|---|
| File acceptance | The portal accepts or rejects a file type or size | Follow the stated format and size rules |
| Parsing | Software extracts name, roles, employers, dates and education | Use readable text, conventional structure and consistent information |
| Search or matching | A recruiter or configured system searches, filters or compares application data | Include relevant, truthful terms in context and answer application questions accurately |
| Eligibility rules | Required answers, location, work authorisation or qualifications may be checked | Read mandatory criteria and do not hide a real mismatch |
| Human review | A recruiter or hiring manager judges relevance and evidence | Make the resume easy to scan and show outcomes, scope and judgement |
Workday describes resume parsing as populating candidate fields, notes that results vary with format and word order, and lets people review the parsed data (Workday resume-parsing documentation). That is different from an automatic rejection or a universal score.
Some employers use knockout questions or automated matching; others use the ATS mainly to collect and search applications, and candidates rarely know which. Where a job post discloses its selection process, as job posts on hiARed do, read it before applying. Otherwise, build a document that preserves your information and helps a person evaluate you.
Use a structure software and people can follow
A dependable order is: name and contact details; a short role-specific summary if it adds context; relevant skills; work experience in reverse chronological order; projects (especially for freshers and career changers); education and certifications; and optional extras such as languages or publications.
Use familiar headings such as Work Experience, Projects, Education and Skills. Greenhouse lists images, text boxes, headers, footers, columns and complex tables among common causes of failed or partial parsing (Greenhouse parsing guidance). Roche's candidate guidance makes similar recommendations and explains that parsed data may populate applications or support recruiter searches (Roche resume-parsing FAQ).
| What goes wrong | Why it matters | Safer fix |
|---|---|---|
| Contact details sit only in a header or text box | Some parsers omit them | Put name, phone, email and location as normal text at the top |
| Two columns interleave | Roles, dates and skills lose their sequence | Use one column or confirm the extracted order |
| Resume is a scanned image | There may be no text layer | Export from the source document |
| Skill levels shown only as charts or icons | Meaning disappears in extraction | Write the skill and evidence in text |
| Dates use several formats | Chronology becomes ambiguous | Use one pattern, such as Jan 2023 – Mar 2025 |
| Important words only inside a hyperlink | The label may be lost | Keep the evidence in text and show the URL separately |
| Abbreviations unexplained | Search and readers can miss context | Use the full term once, then the abbreviation |
Designers and creative candidates can still use a visual portfolio: submit the simple resume where a portal expects one and link the portfolio separately.
Choose DOCX or PDF based on the application
- Follow the employer's instruction. If the portal requests PDF or DOCX, use it.
- Use DOCX as a conservative fallback. Roche recommends DOCX over PDF for parsing accuracy in its process.
- Use a text-based PDF when layout matters. Export it from a word processor; never scan it or flatten it into an image.
- Test the exported file before sending it (see below).
A PDF can look correct while containing broken text order or an image instead of selectable text, and a DOCX can fail if it relies on floating text boxes or nested tables. Give the file a clear name such as Asha-Mehta-Data-Analyst-Resume.pdf, and never include passwords, government ID numbers, bank details or other information an employer does not need at this stage.
Match job language with evidence
Keywords help when they identify real requirements and hurt when they replace evidence. Build a small requirement map from the job description:
| Requirement | Mandatory or preferred? | Your evidence | Where it belongs |
|---|---|---|---|
| SQL analysis | Mandatory | Built weekly revenue and retention reporting | Skills and experience bullet |
| Dashboard tool | Preferred | Created Power BI operations dashboard | Skills and project bullet |
| Stakeholder communication | Mandatory | Presented findings to sales and finance leads | Experience bullet |
| Three years' experience | Mandatory | Two years and four months | Do not disguise the gap; decide whether to apply |
Use the employer's terminology where it is accurate. Do not paste the description, repeat a skill in every section, add hidden white text or claim a tool you have only watched in a tutorial. Never rewrite an official title into a more senior one.
Then put the terms inside proof, using action + object + context or scale + result or purpose. Illustrative examples:
Weak: Responsible for social media and reports.
Stronger: Planned the weekly content calendar for three product accounts and built a campaign report that helped the marketing lead compare reach, enquiries and cost per enquiry.
The stronger version invents no percentage; add a measured result only if you can support it. Freshers can use projects the same way, for example: "Analysed 18 months of public bus-arrival data in Python, documented missing-data assumptions and compared two forecasting baselines in a reproducible notebook."
Test the resume yourself
You do not need a commercial score to find the serious problems.
- Text layer. Select all text in the final file and paste it into a plain-text editor. Check that name, contact details, headings, employers, titles, dates and bullets appear in a sensible order.
- Portal fields. When an application pre-fills your profile, compare every field with the resume and correct missing employers, dates, degrees and skills. Recruiters may search the saved data rather than the attachment.
- Relevance. Mark where the resume proves each mandatory requirement. If the evidence is buried, move it up; if it does not exist, keywords will not fix that.
- Human scan. Can a recruiter see your direction, most relevant work and strongest evidence in a short scan? Remove dense summaries and repeated skills.
Commercial scanners can help compare versions, but do not keep adding phrases merely to raise a third-party percentage.
Tailor the emphasis and run a final check
Keep a factual master resume and create versions for a few role families. Tailoring should change emphasis and language, never employers, dates, titles, results or skills. Generative AI can spot unclear bullets, but it can also invent metrics and inflate responsibility, so verify every suggestion.
Before submitting, confirm:
- the portal accepts the file type and size, and the text is selectable in the right order;
- contact details sit in the body, and headings and dates are consistent;
- every mandatory requirement has truthful supporting evidence;
- bullets explain work, context and outcome, with keywords used naturally;
- links work, the filename is professional and application fields match the document;
- a person has checked spelling, numbers and any AI-assisted edits;
- you have saved the submitted version with the job description.
The broader job-search workflow explains which applications deserve tailoring and how to diagnose weak response rates. HR teams evaluating automation can read the guide to AI in recruitment.
