Job seekers often picture an applicant tracking system (ATS) as a gatekeeper robot that reads CVs and silently rejects most of them. Recruiters picture it as the place where they spend their day moving cards between columns. Both are partly right. An ATS is first a database and a workflow engine, and only secondarily a filter. Understanding its parts explains what it can and cannot do for a hiring team.
The core: a record for every candidate and every requisition
Before an ATS, a mid-sized company typically ran hiring on email threads and spreadsheets. A recruiter downloaded CVs from a job portal, renamed the files, tracked status in Excel, forwarded profiles to hiring managers, and chased feedback over WhatsApp. When that recruiter left, much of the history left with them.
An ATS replaces this with two linked objects. A requisition (or job) holds the approved position: title, department, location, budget band, hiring manager, number of openings and the interview plan. A candidate record holds the person: CV, contact details, source, every application they have made, every interview, all feedback and every email exchanged. The link between them, often called an application, carries the candidate’s stage for that particular job.
This sounds mundane, but it is the most valuable thing an ATS does. If a strong candidate was a silver medallist for a role eight months ago, the next recruiter can find them, see why they were not selected, and contact them again. Without that memory, every requisition starts from zero.
The pipeline: stages, owners and timestamps
Each requisition has a pipeline of stages, for example: applied, screened, assessment, technical round, manager round, HR round, offer, accepted, joined. Moving a candidate from one stage to the next records who moved them and when. That timestamp trail produces almost every recruitment metric automatically: time to shortlist, time in each stage, interview-to-offer ratio, source effectiveness and offer acceptance rate.
Stages also carry actions. Moving someone to “technical round” can trigger a scheduling link to the candidate and a calendar invite with a scorecard for the panellist. Moving someone to “rejected” can send a polite, templated email after a delay so the rejection does not arrive minutes after an interview. The pipeline view lets a recruiter see at a glance that, say, nine candidates have been waiting more than five days for hiring manager feedback, which is where most delays hide.
Parsing, search and the “ATS-friendly CV” question
When a CV arrives, the ATS parses it: it extracts name, phone, email, employers, dates, education and skills into structured fields. Parsing quality varies. Plain layouts parse well; CVs built as images, multi-column designs, text inside graphics or unusual section headings can lose information. When parsing fails, the candidate is not usually rejected; their data is simply harder to find, because searches run against the parsed fields and the extracted text.
That is the honest version of the “ATS-friendly CV” advice. The system rarely rejects a CV because of formatting. But a CV whose skills section could not be read will not appear when a recruiter searches the database for “Kafka” six months later. Using standard headings, real text and the exact names of tools you have used makes you findable.
Where automatic filtering really happens
Automated rejection does exist, but it is usually configured by people, not invented by the software. The common mechanisms are:
- Knockout questions in the application form: “Are you willing to work from Hyderabad?”, “Do you have at least three years of Selenium experience?”, “What is your notice period?” A candidate who answers outside the set range can be moved to a rejected or hold stage automatically.
- Ranking or match scores that compare a CV to the job description. Many recent systems use machine learning or large language models for this. Recruiters often sort by score and read only the top of the list, which is filtering in effect, even if nothing is formally rejected.
- Assessment cut-offs from integrated coding or aptitude tests, where a score below a threshold moves the candidate out of the pipeline.
Each of these deserves scrutiny. A knockout on “minimum 5 years” will reject a strong candidate with 4.5 years. A match score trained on past hires can reproduce past biases, for example against career breaks, which disproportionately affect women. A responsible team reviews knockout rules per role, samples auto-rejected applications periodically, and treats AI scores as a reading aid rather than a decision.
Integrations that make or break it
An ATS on its own is an island. Its usefulness grows with its connections: job portals and LinkedIn for posting and importing applicants, the careers page on the company website, calendar systems for scheduling, assessment platforms, background verification vendors, e-signature for offers, and the HRMS so that a hired candidate’s data flows into onboarding and payroll without being retyped. Each manual hand-off reintroduces the spreadsheet problem the ATS was meant to solve.
In India, the integration with job portals such as Naukri, and with referral workflows, usually matters more than fancy features, because those two channels supply a large share of applicants for many companies.
Data, consent and retention
An ATS quickly becomes one of the largest stores of personal data in a company: lakhs of CVs with phone numbers, addresses, salaries and sometimes identity documents. Under the Digital Personal Data Protection Act, 2023, the company needs a lawful basis for holding that data, should collect only what it needs, and should not keep it indefinitely without reason. Practical consequences include clear consent text on application forms, retention rules that purge or anonymise old candidate records, role-based access so a hiring manager in one team cannot browse every applicant’s salary, and care when sharing profiles with recruitment agencies.
What an ATS will not fix
An ATS records a process; it does not create a good one. If interviewers do not submit feedback, the pipeline stalls just as it did by email, only now with a visible timestamp. If stages are configured inconsistently across teams, the metrics become meaningless. If recruiters keep side spreadsheets, the database decays. The companies that get value from an ATS are those that agree on the pipeline, insist that decisions are recorded where everyone can see them, and look at the data regularly to find where candidates are waiting.
Written by the RP Strativa Team.
Leave a Reply