Recruiting teams are usually asked to improve both at once. Hire faster, and spend less doing it. Nobody says out loud that those two instructions pull in opposite directions, so the team quietly picks whichever one the last complaint was about.
They are also not the same kind of number. One is a duration and one is money, which means they cannot be traded off against each other directly. They can only be compared through a third figure, and most companies never calculate it.
What each one actually measures
Both get computed loosely, and both are easy to make look better without changing anything real.
| Metric | Definition | Reported as |
|---|---|---|
| Time to hire | Days from a candidate entering the pipeline to that candidate accepting an offer | Median days, by role family |
| Time to fill | Days from the requisition being approved to an offer being accepted | Median days, separately from time to hire |
| Cost per hire | Total internal plus external recruiting costs in a period, divided by hires in that period | Rupees per hire, segmented by role family |
| Cost of vacancy | Daily value of the output the role is not producing, times days open | Rupees per day, and rupees per vacancy |
Time to hire is the candidate-facing number. It is what determines whether you lose people to a company that moved faster, and it is the one worth optimising if you compete for scarce skills.
Cost per hire is the budget-facing number, and the standard definition includes internal cost. Recruiter salaries, and the hiring manager and interviewer hours spent on the process, belong in it alongside job boards, agencies, referral payouts, assessments and background checks. Most reported cost per hire figures include only the invoices, which is why they look reassuring and tell you nothing about where the effort goes.
A cost per hire built only from invoices does not measure cost. It measures which costs happened to arrive with a bill attached.
Why they pull against each other
Almost everything that makes hiring faster costs money. An agency retained to shorten a search, a premium sourcing licence, a higher referral bonus, paid job board placement, an assessment tool that removes a screening round, or a dedicated recruiter on one requisition rather than eight.
Almost everything that makes hiring cheaper takes longer. Sourcing in-house instead of through an agency, running one channel instead of four, letting a requisition wait for the right candidate rather than paying to accelerate it, or spreading a recruiter across more roles.
This is why measuring a team on both without stating a priority produces the behaviour you would expect. They optimise for whichever number is being looked at this quarter, and the other one drifts until somebody notices.
Where the two do not trade off
The trade-off is real at the margin, but there is a large category of delay that costs time and money simultaneously. These are the wins to take before anyone argues about budget.
- Profiles sitting unreviewed with a hiring manager, which extends the search and burns recruiter time chasing
- Interview scheduling taking a week, which loses candidates and consumes coordinator hours
- Feedback arriving four days after the interview, which forces re-engagement work and sometimes a repeat round
- Unclear role requirements, which produce rejected shortlists and a search that starts twice
- Interviewing more people than the decision needs, which spends senior time on candidates the bar had already excluded
- Slow offer approvals, which lose accepted candidates and restart the funnel at full cost
Every one of those adds days and rupees at the same time. If your time to hire is poor and your cost per hire is high, look here first, because nothing on this list requires a trade-off decision or a budget conversation.
The third number that settles the argument
You cannot compare 28 days against ₹1,31,000 until you know what a day of vacancy is worth. That is the cost of vacancy, and it converts the duration into money so the two become comparable.
The rough version is the value the role produces per working day, times the days it stays empty. For a revenue role, use the contribution the role is expected to generate. For a support role, use the fully loaded cost of the coverage arrangement instead: overtime, contractor cover, or the delay it creates for other people.
Apply a productivity discount, because a new hire does not produce full output on day one. If the ramp is three months, the first weeks of the vacancy are worth less than the calculation suggests, and pretending otherwise inflates the number in a way finance will spot immediately. Use a conservative figure that survives scrutiny rather than one that wins the argument on paper.
Once the daily figure exists, the decision is arithmetic rather than opinion. If the vacancy costs more per day than the premium you would pay to close it faster, paying for speed is the cheaper option, and you can show the working.
A worked example on one role
A senior engineering role at ₹24 lakh CTC, filled two ways. The numbers are illustrative, but the shape is what matters.
| Line | In-house search | Agency search |
|---|---|---|
| Recruiter time | ₹36,000 | ₹12,000 |
| Interviewer and hiring manager time | ₹60,000 | ₹40,000 |
| Job advertising and sourcing tools | ₹25,000 | Nil |
| Agency fee | Nil | ₹2,00,000 |
| Cost per hire | ₹1,21,000 | ₹2,52,000 |
| Time to hire | 62 days | 34 days |
The agency route costs ₹1,31,000 more and saves 28 days. Whether that is a good trade depends entirely on the vacancy cost.
Suppose the team's output per engineer works out to about ₹16,000 per working day, and you apply a 50 per cent discount for ramp and for the fact that the team partially absorbs the gap. That is ₹8,000 a day, so 28 days of vacancy costs roughly ₹2,24,000. On those numbers, spending ₹1,31,000 to save 28 days is worth doing, and the case takes one slide.
Change the role and the answer reverses. A support role with a queue that other people can absorb, where the daily cost of the gap is ₹1,500, makes the same 28 days worth ₹42,000. Paying ₹1,31,000 to save it is a poor decision, and the in-house search is right even though it is slower.
Which one to prioritise, by role
The answer is not a company-wide policy. It is a per-role-family default that the team can apply without asking.
- Revenue-generating roles where the seat directly produces output: prioritise speed, and expect to pay for it
- Scarce or specialist skills with a small candidate market: prioritise speed, because the constraint is availability and slow processes lose the few candidates who exist
- Leadership roles: prioritise quality above both, accept a long search, and measure time to fill rather than time to hire so the sourcing effort is visible
- High volume roles with abundant supply: prioritise cost, and build a repeatable process rather than paying per hire
- Roles where the work is genuinely being absorbed by the team: prioritise cost, but track how long you are relying on that absorption, because it has a limit
Writing that down once removes most of the recurring argument, and it gives recruiters a defensible answer when a hiring manager asks why the agency was or was not used.
How each metric gets gamed
Both respond well to being managed rather than improved, which is worth knowing before you set a target on either.
Time to hire shortens if you start the clock later, count only candidates who were already in a pipeline, exclude searches that were cancelled or restarted, or lower the bar so the first adequate candidate gets the offer. The last one is invisible in the metric and shows up nine months later in first-year attrition.
Cost per hire falls if you exclude internal costs, push screening and sourcing work onto hiring managers, or count a period with an unusual number of easy hires. Moving work from a recruiter to five engineers does not reduce cost. It moves it somewhere nobody is counting, and it usually increases the total.
Neither of these is dishonest most of the time. They are the natural consequence of putting a target on a single number, which is the argument for reporting them together.
Never report either one alone
The pair constrains itself. Time to hire falling while cost per hire climbs is a purchasing decision, and it may be the right one. Both improving at once usually means process delay was removed, which is the best outcome available. Cost falling while time climbs means somebody decided to be patient, which needs to be a decision rather than a drift.
Add quality of hire as the third leg, even in a rough form such as first-year retention, because it is the only thing that catches the failure mode both metrics share. A fast, cheap hire who leaves in seven months improved two numbers and cost more than either of them saved.
Frequently asked questions
What is the difference between time to hire and cost per hire?
Time to hire measures days from a candidate entering the pipeline to accepting an offer. Cost per hire measures money, as total internal and external recruiting cost divided by hires in the period. One is a speed measure and the other a budget measure, and they usually move in opposite directions.
Which is more important, time to hire or cost per hire?
It depends on what the vacancy costs per day. Where the empty seat costs more per day than the premium needed to fill it faster, speed is the cheaper choice. For high volume roles with abundant supply, cost discipline usually wins.
How do you calculate cost per hire?
Add total internal recruiting costs, including recruiter salaries and hiring manager and interviewer time, to total external costs such as agency fees, job boards, referral payouts and assessments. Divide by the number of hires in the same period.
How do you calculate cost of vacancy?
Estimate the value the role produces per working day, apply a discount for ramp time and for work the team absorbs, and multiply by the days the role stays open. Use a conservative figure that will survive scrutiny from finance.
Does reducing time to hire increase cost per hire?
Usually, where the reduction comes from buying speed through agencies, paid sourcing or higher referral bonuses. Not where it comes from removing delay inside your own process, which improves both at once.
What is a good time to hire?
It varies too much by role, seniority and market for a single number to be useful. Track your own median by role family and watch the trend rather than comparing against a published average.
Should internal costs be included in cost per hire?
Yes. The standard definition includes internal recruiting cost alongside external spend. Excluding it is the most common reason a reported cost per hire looks good while telling you nothing.
How can both metrics be improved at the same time?
By removing delay that costs time and money simultaneously: profiles waiting for review, slow interview scheduling, late feedback, unclear requirements and slow offer approvals. None of those require a budget trade-off.
If the two numbers are being argued about in your organisation, the argument is usually a proxy for a missing figure. Work out what a day of vacancy costs for two or three role families, write down which default applies to each, and report speed, cost and first-year retention on the same line. The debate tends to end there, not because anyone won it, but because it turns into arithmetic.
