ATS
Why your resume never gets a reply
By AI Resume Maker · 26 July 2026 · 8 min read
You applied to sixty roles on Naukri and heard back from two. The market is bad and there really were 400 other applicants, both true, and both hide the more useful answer: for most of those sixty, no human ever saw your resume. An applicant tracking system read it, failed to extract half of it, scored what was left against the posting, and put you below the cut line.
What an ATS actually does
An applicant tracking system is a database with a parser attached. When you upload a resume, the parser tries to turn a visual document back into structured fields: name, email, phone, each employer, each date range, each degree, a skills list.
It's good at this when your resume is a single column of plain headings and paragraphs. It gets worse as the design gets better, which is the uncomfortable part: the templates that look most impressive to you are usually the ones it handles worst.
Whatever the parser fails to extract doesn't exist. If "Kubernetes" sat inside a graphic or a text box it skipped, a recruiter searching the database won't find you, even though the word is right there on the page you sent.
The five formatting choices that lose you information
In rough order of how much damage they do:
- Two-column layouts. The parser reads left to right across the whole page width, so your sidebar skills interleave with your job history and both come out as nonsense.
- Tables. Same problem, worse. A skills matrix in a table often arrives as one run-on line, or gets dropped.
- Text inside images or icons. A parser reads text and can't see pictures, so contact details in a header graphic are invisible to it.
- Headers and footers. Many parsers skip them entirely. Never put your phone number or email there.
- Non-standard section names. "My Journey" instead of "Work Experience", "Toolbox" instead of "Skills". The parser is looking for the standard headings, so give it the standard headings.
Then it scores what survived
Once the text is extracted, the match is mostly vocabulary overlap between your resume and the job description, and it wants the actual terms rather than synonyms. Write "ReactJS" where the posting says "React.js" and some systems match you, some don't. Demonstrate stakeholder management without ever using the phrase "stakeholder management" and you'll score below someone less experienced who did use it. That one still annoys me, but it's how the matching works.
It's also why one resume sent to fifty postings does badly against all fifty: it's tuned to none of them.
Stuffing doesn't fix it. Repeating a keyword twenty times trips a different filter and reads as spam to the human who eventually opens the file. What works is quieter: every genuine skill you have, stated in the words the posting uses, once, in context.
The India-specific ones
A few habits are near-universal on Indian resumes and near-absent on the ones that get shortlisted at product companies:
- A photo. Standard on many Indian templates, unnecessary for almost every role, and a parsing hazard on top of that.
- Date of birth, marital status, father's name, nationality. Left over from a different hiring era. They take space and add nothing.
- The declaration block, "I hereby declare that the above information is true..." with a signature underneath. It has no function on a screened resume.
- "Objective: Seeking a challenging role in a reputed organisation where I can utilise my skills." This burns the most valuable line on the page saying nothing. One sentence naming your specialisation and what you build will do more.
- Percentages from Class 10 and 12, once you have a degree and any experience at all. Campus rubrics sometimes still ask for them, so check before you cut.
What to do this week
Take one resume and one job posting you actually want. Strip the resume to a single column with standard headings. Delete the photo, the declaration, and the objective. Then read the posting and check, requirement by requirement, whether your resume states the thing in the posting's own words.
Done by hand that's about forty minutes, most of it auditing rather than writing. We built the free ATS score checker to take the mechanical half off you: keyword match, missing sections, the formatting a parser will choke on. It runs in a few seconds, which leaves the forty minutes for the part only you can do.
One caveat worth saying plainly. Every ATS behaves a little differently, and vendors don't publish their parsing rules, so treat the specifics above as the pattern we see most often rather than a guarantee about any one employer's stack.
Put it into practice
Score your resume against a real posting with the free ATS score checker, or rehearse the HR round against an AI interviewer that asks the follow-ups.
Get Started FreeMore from the blog