ResumeCrew
Menu

Guide · ATS and screening

ATS-friendly resume in India: what parsers read and what trips them

An ATS is a reader, a filing cabinet and, increasingly, a ranker. This guide explains what each of those does with your file, the format rules that follow from it, and the shortcuts that backfire.

Last reviewed
October 2026
Reading time
7 min

In short

  • An ATS reads your file into fields, stores them for recruiter search and, in newer systems, ranks candidates against the job. A file that cannot be read cannot be ranked.
  • Single column, standard headings, contact details in the body, text-based PDF under 1 MB. These rules come from vendor documentation, not taste.
  • Show each key skill in the Skills line and in at least one bullet that proves it. Never hide text.
  • No tool can promise a shortlist. A good score only removes avoidable reasons to be missed.

What an ATS does with your file

An applicant tracking system, or ATS, does three jobs. It reads your file into fields: name, email, employers, dates, degrees, skills. It stores those fields so recruiters can search them. And in newer products it ranks candidates against a job's requirements. According to vendor documentation and recruiter guidance, it rarely rejects anyone alone; it decides what a recruiter sees first.

Workday's HiredScore is a useful example because its documentation is public. It grades candidates from A to D, and its August 2026 release notes describe a language-model approach that checks each screenable qualification against the parsed resume and explains its verdict. The same documentation says no grade is assigned when a resume cannot be parsed. An unreadable file is not marked down. It is not scored at all.

So two things matter. The parser must read your file correctly, and what it reads must show evidence of the skills the posting asks for. Layout fixes the first. Honest, specific bullets fix the second.

Ten format rules, with the reason for each

Rule and reason
RuleWhy
Single column, no sidebarsParsers read across columns. A leading parser vendor reports only about 90% of column layouts come out correctly.
Standard headings: Summary, Skills, Experience, Education, Projects, CertificationsParsers sort text into fields by heading words. Creative headings land in the wrong field or none.
Contact details in the page bodyGreenhouse lists details placed in a header, footer or text box among the causes of a failed parse.
Dates as Mon YYYY – Mon YYYYYear-only ranges such as 2020-22 make tenure ambiguous.
Spell out company and job titleGreenhouse notes that bare company names and abbreviated titles such as 'Sr. Account Exec' cause trouble.
Plain round bullets, no icons or symbolsDecorative symbols can become garbage characters in extracted text.
No photo, logo, chart, skill bar or QR codeParsers ignore images, and images push a file towards size limits.
Text-based PDF with embedded fontsSAP's documentation says scanned or image-only PDFs fail to parse.
Under 1 MB, named Firstname_Lastname_Resume.pdfPortals set caps, and Greenhouse does not parse files above 2.5 MB. Clear names help recruiters who download in bulk.
Links as visible textSome parsers read only what is printed on the page.

Where Indian resumes go wrong

The common faults are consistent. Graphic two-column templates from design tools. A photo, date of birth, father's name and marital status at the top. A declaration paragraph at the bottom. A generic objective. Two to four pages for a fresher. One career site that reviewed 500 fresher resumes reported that about 340 used a graphic two-column layout; that is a single site's sample, not a national statistic, but it matches what recruiters describe.

None of these is a moral failing. They are habits from the biodata era. Most of them are also fixed by removing something, which is why the advice in what to remove from an Indian resume is short.

Keywords without stuffing

Use the posting's own words for skills and titles. Write 'Kubernetes', not only 'container orchestration'. Then give each key skill a home in the Skills line and in at least one bullet that shows it, because ranking tools increasingly look for evidence, not only presence. Jobscan describes its match rate as weighting hard skills most, then education level, job title, soft skills and other keywords, and it counts repetition: a skill that appears three times in a posting is expected more than once in your resume.

From a skill on the list to a skill with evidence

  1. Before

    Skills: SQL, Power BI, Excel

    After

    Built a Power BI dashboard on 60,000 rows of retail sales pulled with SQL; the store manager used it every Monday to plan staffing

    Why

    The same three keywords now appear inside a sentence that shows use, scale and an audience.

  2. Before

    Knowledge of Java and Spring Boot

    After

    Developed 6 REST endpoints in Java and Spring Boot for a student-fee tracker, covered by 25 JUnit tests

    Why

    Skill names plus a count of what you built. Easier for a person to trust and for a ranker to match.

Naukri, LinkedIn and company portals

Naukri recruiter search works on your resume headline, your Key Skills field and the resume text, so the profile fields matter as much as the file. Keep the headline to your role, your experience and three or four core skills, and fill Key Skills with skills your resume evidences. Exact weights are not public, and percentages quoted on blogs are unverified, so do not optimise for a number.

Workday's documentation says its parse does not fill the Languages and Skills fields, so expect to type those, and check every pre-filled field. On LinkedIn Easy Apply the profile does most of the work, so keep its titles and dates identical to your PDF. Employers compare the two, and verification firms compare both with your documents.

Test your own file in two minutes

  1. 01

    Copy and paste

    Open the PDF, press Ctrl+A, copy, and paste into Notepad. The text should come out in the same order as the page.

  2. 02

    Read the top

    Your name, phone and email should appear first as plain text, and headings should read as words, not as letter-spaced 'S K I L L S'.

  3. 03

    Check the lines

    Each company, title, date range and degree should sit on its own clear line with the same separators.

  4. 04

    Run the checker

    Upload the file to the free ATS checker. It lists each rule it tested and offers a one-tap fix where one exists.

Questions people ask

Does a high ATS score mean I will be shortlisted?

No. Scores from any tool, ours included, estimate parse-safety and keyword fit. They are not a prediction from one employer's system, and a recruiter still decides. A good score removes avoidable reasons to be missed. The rubric is published at how the score works.

Is a Word file safer than a PDF?

Both parse when they hold real text. Some users report that DOCX handles dates and skills more reliably on Naukri and older systems, but that evidence is anecdotal. Use a text-based PDF by default and keep a DOCX for recruiters who ask for one.

Will an ATS reject an AI-written resume?

Jobscan's review found that no major ATS natively detects AI authorship. Human readers do notice sameness: identical bullet rhythm, generic verbs, numbers with no context. Write from your own facts and use AI to tidy wording, never to supply claims.

Can I use a two-column template if it looks better?

Some vendors argue modern parsers handle two columns well, and that may hold for the best engines. You rarely know which system a company runs, and smaller tools read straight across the page. One column is the safe default.

What ATS score is good?

In our scoring, below 60 needs work, 60 to 79 is good and 80 and above is strong, with 85 a sensible target. A match score against a job description is separate, and 75 or more is a common target.

Sources

Numbers on this page come from these sources, and the text says where a source is a vendor blog or a survey. Where a claim has no source here, it is guidance, not fact.

  1. Workday HiredScore release notes, 5 August 2026https://doc.workday.com/hiredscore/en-us/workday-hiredscore/hiredscore-release-notes/2026/august/wednesday--5th-august.htmlLanguage-model qualification matching; no grade when a resume cannot be parsed.
  2. Workday admin guide: resume parsinghttps://doc.workday.com/admin-guide/en-us/human-capital-management/recruiting/candidates/set-up-prospects-and-candidates/hdc1552497830785.htmlParse does not fill Languages and Skills; avoid images and image-based styles.
  3. SAP KBA 2081576: SuccessFactors resume parsinghttps://userapps.support.sap.com/sap/support/knowledge/en/2081576Text-based PDFs work best; scanned or image PDFs fail.
  4. Greenhouse: Unsuccessful resume parsehttps://support.greenhouse.io/hc/en-us/articles/200989175-Unsuccessful-resume-parseHeader and footer contact details, tables, text boxes, graphics, bare company names, abbreviated titles.
  5. Greenhouse: Supported file formats and limitshttps://support.greenhouse.io/hc/en-us/articles/360052218132Files above 2.5 MB are not parsed.
  6. Textkernel: Improving extraction from column resumeshttps://textkernel.com/learn-support/blog/improving-extraction-from-column-resumesAbout 90% of column CVs render correctly after its fix.
  7. Jobscan tutorial: how the match rate is builthttps://www.jobscan.co/jobscan-tutorialOrder of weighting and the role of repetition. A vendor's description of its own tool.
  8. Duke Pratt: Thwarting hidden resume hackshttps://pratt.duke.edu/news/thwarting-prompt-injection/Summary of the study of about 200,000 resumes.
  9. Measuring real-world prompt injection attacks in LLM-based resume screening (arXiv 2605.28999)https://arxiv.org/pdf/2605.28999Hidden prompts in about 1% of resumes; sevenfold rise between July 2024 and November 2025.
  10. Jobscan: Can an ATS detect an AI-written resume?https://www.jobscan.co/blog/can-ats-detect-ai-resume/No major ATS natively detects AI authorship. A vendor review.
  11. Jobrise: 15 resume mistakes freshers make in Indiahttps://jobrise.io/en/blog/resume-mistakes-freshers-india/A 500-resume review by one career site.

Next step

Build yours in the chat

Answer a few questions, one at a time. The reviewer drafts each line, asks for the numbers it cannot guess and never adds a fact you did not give. Already have a resume? Check it first and see what a parser reads.