Time to hire: how you measure it and where the delays really come from

Echipa HR 365 · reviewed 2026-09-02 · 5 min read

Time to hire is measured from the moment a candidate enters the process to the moment the offer is accepted. The total figure is only useful for knowing whether you have a problem; to solve it, you need it split by stage — and there you see that most of the lost days are waiting days, not working days.

Two indicators that get permanently confused

Time to hire starts when the candidate enters the process. Time to fill starts when the post opens. The first measures how quickly the recruitment team works with a person; the second, how long the post stays empty — including the weeks beforehand spent approving a budget and writing a job description.

The confusion has consequences: an excellent time to hire can coexist with a disastrous time to fill, if the post spent three weeks in approval. Recruitment reports the first figure, operations feel the second, and both are right.

Almost half the duration is days when nobody does anything with the candidate, they are simply waiting for a window in somebody’s calendar. No improvement to screening touches that figure. Two fixed windows a week, reserved in advance in the interviewing managers’ calendars, typically cut 5–8 days from the total without changing anything else in the process.

How you measure it without inventing data

What has to exist for the indicator to be real

1. The system — records the date of every stage transition, automatically, not by later manual entry
2. Recruitment — defines what a “qualified candidate” means before measuring, or the attraction stage is negotiable
3. Recruitment — reports the median, not the mean — one position filled in 4 months moves the mean and hides the pattern
4. Recruitment — separates positions by role family; an operational post and a leadership one have nothing to compare

The median rather than the mean is the recommendation that most changes the report. The distribution of recruitment durations is almost always skewed — a few very long positions pull the mean up and make everything look slow, when in fact most are fine and two are stuck.

What the indicator does not do

It says nothing about hiring quality. A 12-day process that brings people who leave in three months is more expensive than a 40-day one that brings people who stay. Which is why time to hire is always read alongside first-year turnover — separately, either can be optimised until it breaks the other.

How you cut the waiting days, concretely

Once you have the breakdown by stage, the interventions that work are few and mundane. They have nothing to do with technology and everything to do with the calendar:

• Two fixed windows a week, reserved in the interviewing managers’ calendars whether used or not. Typically cuts 5–8 days from the total and is the only intervention you can make without anybody else’s agreement.
• The decision is taken within 48 hours of the last interview, with a clear rule for what happens if an assessor does not respond. Ambiguity here costs, on average, more than the whole screening stage.
• Parallel rather than sequential interviews, where the role allows: two assessors in the same hour instead of two bookings a week apart.
• The job description and the interview scorecard prepared before the advert is published, not after the first applications — otherwise the first days go on preparation, with candidates already waiting.

What does not work, though it is often tried: pressure on the recruiter. The stages they control are usually already the shortest in the whole process, and speeding them up produces superficial screening and more pointless interviews — the opposite of the intended effect.

A last note, for comparisons between periods: recruitment has strong seasonality. July and August add days to almost every stage, because assessors and candidates are on holiday, and December does the same for another reason. An eight-day increase in August compared with June says nothing; the same increase compared with August last year says everything.

Do I measure in calendar or working days?

Calendar days, because that is how the candidate experiences it and how the waiting department feels it. Working days are useful only when you analyse a particular stage internally.

What do I do about positions I close without hiring?

They are reported separately, as a rate of positions closed without a hire. Included in time to hire they would distort the figure; ignored entirely, they hide exactly the positions where the process failed.