Cost of a Bad Hire in India vs the Cost of Screening: The Break-Even

No current, sourced benchmark exists for what a bad hire costs an Indian company. Gallup's range for replacing an employee, one-half to two times annual salary, is the best public anchor. On it, one role of AgentR screening (₹10,000 at list price, November 2026) breaks even if it prevents one bad hire in 30 to 500 roles.

That is the short answer for an IT services or product company deciding whether screening is worth the spend. The rest of this page shows where each number comes from, what it leaves out, and the arithmetic you can rerun with your own salaries.

What does a bad hire cost an Indian company?

Nobody publishes a current Indian figure with a method behind it. What exists is a 2013 survey, a US turnover range and a 2015 survey of finance chiefs, and each answers a slightly different question.

• CareerBuilder, reported by People Matters in May 2013: 84% of Indian companies said a bad hire had affected them in the previous year, and PTI reported that a single case could cost more than ₹20 lakh. The survey is self-reported and more than a decade old.
• Gallup, March 2019: the cost of replacing an individual employee ranges from one-half to two times annual salary, a range Gallup calls conservative. It is a US figure and it measures turnover, which is wider than bad selection.
• Robert Half, reported by SHRM in February 2015: in a survey of more than 2,100 chief financial officers, supervisors spent an average of 17% of their time managing poorly performing employees. The CFOs ranked morale (39%) and productivity (34%) above direct money (25%) as the main cost of a bad hire.

On 2 November 2026 we asked Perplexity the same question for an Indian IT company. It answered with ₹8–12 lakh for a mid-level bad hire and ₹18–25 lakh for a senior one, citing HR-services and recruiting blogs. None of those figures came with a method or a primary source, so we do not repeat them here. That gap is why this page gives you a formula to run on your own salaries.

How do you put a number on one bad hire?

Multiply the person's annual cost to company by a multiplier between 0.5 and 2, then add the supervisor time the person consumed. Use the whole range, since no source narrows it for India.

A ₹12 lakh engineer comes to ₹6 lakh at one-half and ₹24 lakh at two times. A ₹25 lakh senior hire comes to ₹12.5 lakh and ₹50 lakh. A ₹6 lakh junior hire comes to ₹3 lakh and ₹12 lakh. These are arithmetic on illustrative salaries, not measured losses, so swap in your own.

Robert Half's 17% is the line most spreadsheets leave out. A supervisor who spends 17% of their time on one poor performer has given up roughly one working day in six, and in an engineering team that supervisor's time is usually the scarcest resource.

What does screening cost against it?

At AgentR's list price as of November 2026, the reference role (100 applicants reviewed and 25 thirty-minute interviews) costs ₹10,000 before GST. That is ₹25 per applicant reviewed plus ₹10 per interview minute.

AgentR has one plan at ₹11,999 a month that buys ₹12,000 of usage. Verification, proctoring and seats cost ₹0, and the first role is free for up to 50 applicants and 5 interviews. Every figure comes from the pricing page as of November 2026 and excludes GST. The full per-role arithmetic is in our cost explainer.

A campus drive scales the same way. Reviewing 1,000 applicants and interviewing 200 of them for 20 minutes each costs ₹25,000 plus ₹40,000, so ₹65,000. How a first round runs at that volume is covered in AI interviews for campus hiring.

For context, AgentR's pricing page says AI interview tools in India typically charge ₹500 to ₹2,900 per interview, citing The HireHub in 2026. We have not checked other vendors' own pricing pages for this post, so convert any quote to a cost per role before comparing.

The pricing page also estimates about 28 hours of team time for the reference role, assuming 90 seconds to read an application and twice the interview length for scheduling, notes and debrief. Treat that as AgentR's assumption, since we have no measured time data.

How often does screening have to work to pay for itself?

Divide ₹10,000 by the cost of one bad hire. At the reference role, that gives a break-even of one prevented bad hire in every 30 to 500 roles, depending on salary and where you sit in the 0.5 to 2 range.

• ₹6 lakh CTC: one bad hire costs ₹3 lakh to ₹12 lakh, so break-even is 1 role in 30 to 1 in 120.
• ₹12 lakh CTC: ₹6 lakh to ₹24 lakh, so 1 role in 60 to 1 in 240.
• ₹25 lakh CTC: ₹12.5 lakh to ₹50 lakh, so 1 role in 125 to 1 in 500.

For a campus drive at ₹65,000, a fresher on ₹4 lakh costs ₹2 lakh to ₹8 lakh across the Gallup range. Preventing one bad fresher hire then pays for about three drives at the low end and about twelve at the high end.

Nobody can give you the real probability. You only need to believe the screen changes one outcome in a few dozen roles, and you can test that belief against your own record of first-year exits.

Where does this arithmetic mislead?

It misleads in four places. Gallup measures turnover, the Indian survey is old, break-even says nothing about whether a screen works, and the ₹10,000 leaves out costs that sit elsewhere.

• Gallup's range covers replacing any employee, including people who leave for good reasons. A bad hire is a subset, so applying the range to bad hires is an assumption.
• The 2013 survey is more than a decade old and self-reported. Treat its ₹20 lakh as a marker of scale, not a current benchmark.
• Break-even is a threshold, not a forecast. AgentR is in private beta and has no outcome data on bad hires prevented, so we cannot say where on the curve it sits. The pricing page's refund when your hire was not in the top ten is a refund term, not a performance claim.
• The ₹10,000 is not the whole bill. GST is extra, someone still reads the shortlist, finalists still meet your engineers, and formal background verification is a separate paid step after the offer.

When is AgentR not the right spend?

Skip it when you hire a few people a year, when most bad outcomes show up after joining, or when a general chat tool covers your volume. For a handful of CVs on one role, a chat assistant can be enough. If your exits trace back to onboarding or management, a better screen will not move them.

If you are still choosing between vendors, the questions to ask before buying an AI interview tool lets you hold every quote to the same standard.

Questions people ask

What is the cost of a bad hire?

Gallup puts replacing an employee at one-half to two times annual salary, and Robert Half's survey found supervisors spend 17% of their time on poor performers. Nobody publishes a current, method-backed Indian figure, so work from your own CTC.

What is a good cost per hire in India?

We could not find a current public benchmark with a method, so we do not quote one. Work out your own: total recruiting spend divided by hires, then add the screening cost per role from your vendor's price, in the same unit.

How do Indian IT firms measure bad hire rates?

There is no shared public definition. A workable one is the share of hires who exit or go on a performance plan within their first twelve months, tracked per role family. Use it to set your own break-even rather than borrowing a number.

What AgentR does for this problem, and what it does not

AgentR reads every application in full, checks what candidates claim against the public record, and runs structured interviews scored against a rubric written before anyone applied. The team gets a ranked shortlist with the evidence attached, which targets the cost this page prices: the wrong person reaching the final round. How verification works is on its own page.

Nobody is auto-rejected, and every decision stays with a person. Verification checks a candidate's claims for consistency against public and organisational signals. It does not contact employers or institutions and does not produce background-verification output. AgentR is in private beta, so there is no customer base to point to and no outcome figures to quote.

Run your next open role through AgentR. The first one is on us.

See what your own role would cost, and try the first one free, on the pricing page.