Hiring · August 2, 2026 · 9 min read

Time to Hire vs Time to Fill: Formulas + Example

Time to hire vs time to fill for employers: both definitions, both formulas, one worked example, and why your own trend beats any published benchmark table.

← Part of The five pillars of hiring: what assessments measure

If you are the hiring manager or recruiter reporting on how quickly roles get staffed, the difference between time to hire and time to fill is not pedantry — it decides whether your dashboard tells you the truth. The two metrics sound interchangeable and get quoted as if they were, which is how a talent team ends up celebrating a fast time to hire while roles sit open for two months and a frustrated department head wonders what the number even measures. They answer different questions. One tracks the candidate through your pipeline; the other tracks the requisition from the moment it is approved. Confuse them and you will optimise the wrong stage, benchmark against numbers that were never comparable, and miss where the delay actually lives. This guide lays both definitions side by side, gives you both formulas, walks one requisition through the calendar so the gap is concrete, and then makes an unpopular argument: the published benchmark table you are looking for is worth less than your own trend.

What is the difference between time to hire and time to fill?

Time to hire measures the candidate's journey — from the day a person enters your pipeline to the day they accept the offer. Time to fill measures the requisition's journey — from the day the role is approved and opened to the day it is filled. The overlap is real, but time to fill also counts sourcing and any delay before the first strong candidate even applies, which is why it is almost always the larger number.

Think of it as two clocks that start at different moments and, often, stop together. The time-to-fill clock starts when the business commits to hiring: the requisition is approved, budget is signed off, the role is live. The time-to-hire clock starts later, when a specific candidate becomes real to you — they apply, or you source and first contact them. Both clocks usually stop on the same event: offer acceptance. The distance between the two start points is the sourcing gap, the stretch where the role is open but no eventual hire has entered the picture yet. That gap is invisible if you only look at time to hire, and it is frequently where the weeks hide.

The distinction matters because the two numbers implicate different people. A long time to hire points inward, at your own process: screening queues, slow panels, sign-off that waits on a calendar. A long time to fill with a short time to hire points upstream — at sourcing, or at a manager who let an approved requisition sit for three weeks before doing anything. Report only one and you hide half the story.

How do you calculate time to hire and time to fill?

Both are subtractions between two dated events, then a median across recent hires. Time to hire is the offer-acceptance date minus the date the candidate entered the pipeline. Time to fill is the offer-acceptance date minus the date the requisition was opened. The only hard part is deciding your start and end events once and then never quietly changing them — consistency is what makes the number mean anything.

Write the definitions down where the whole team can see them, because the failure mode is not the arithmetic — it is drift. One recruiter counts from application date, another from first recruiter contact, a third from the screen. Each is defensible in isolation; together they produce a metric that wobbles for reasons that have nothing to do with hiring speed. The two formulas are simple enough to state plainly:

• Time to hire (one candidate) = offer-accepted date − date the candidate entered the pipeline (application or first contact).
• Time to fill (one role) = offer-accepted date − date the requisition was approved and opened.
• Reported metric = the median of those per-hire or per-role numbers across a defined period. Prefer the median over the mean; one dragged-out executive search will skew an average and hide your typical experience.

A word on start events. For time to hire, application date is the cleanest choice for inbound roles because it is unambiguous and sits in your applicant tracking system already. First-contact date suits sourced or headhunted candidates who never formally applied. Pick the rule that fits how you actually source, document it, and apply it to everyone. For time to fill, resist the temptation to start the clock when the first candidate applies — that quietly deletes the sourcing gap and flatters the number. Start it when the role was approved.

A worked example: one requisition through the calendar

The cleanest way to feel the difference is to walk a single hypothetical requisition through the calendar. The dates below are illustrative — invented to show the arithmetic, not drawn from any survey or dataset. Imagine a mid-level analyst role at a company that documents its metrics properly.

• 1 March — the requisition is approved and opened. The time-to-fill clock starts here.
• 1 March to 12 March — sourcing and advertising. The role is live but no eventual hire has entered the pipeline yet. This eleven-day stretch is the sourcing gap.
• 12 March — the candidate who will eventually be hired applies. The time-to-hire clock starts here.
• 12 March to 4 April — screening, a work-sample assessment, and two structured conversations.
• 4 April — the candidate accepts the offer. Both clocks stop.

Run the two subtractions. Time to hire is 4 April minus 12 March: twenty-three days — the candidate's journey through the pipeline. Time to fill is 4 April minus 1 March: thirty-four days — the role's journey from approval to acceptance. The eleven-day difference is entirely the sourcing gap, the period the time-to-hire number never saw. Report only time to hire here and you would tell the department head the role took twenty-three days to fill, when they experienced thirty-four days of an empty seat. Neither number is wrong; they simply answer different questions, and the honest report shows both.

You do not need a spreadsheet to reproduce this. You can build and compute the same calculation in the browser with our time-to-hire calculator — enter your own start and end dates, and it does the subtraction and the median for you so the definition stays fixed even as the team changes.

What is a good time to hire or time to fill?

There is no universal good number, and anyone who quotes one without asking about your role mix and your definitions is guessing. Published benchmark tables conflict because they draw their start and end lines in different places and blend roles, markets and seniority into a single average. A figure that reads as fast for a graduate scheme reads as slow for a senior engineering search. The useful benchmark is your own trend, measured the same way every time.

For rough orientation only, SHRM's annual data puts the median time to hire somewhere around 24 to 30 days and time to fill nearer 44 days. Notice that the shape matches the example above: time to fill runs meaningfully longer than time to hire, because of the sourcing gap. Use those figures to sanity-check that you are in the right postcode, not as a target. If your numbers are wildly outside that range, the first question is usually whether you are measuring the same thing the benchmark measured — and you almost never are.

This is the honest position most benchmark articles will not take: chasing someone else's average is a distraction. What moves hiring outcomes is watching your own median month over month, with a frozen definition, and asking why it changed. Did a stage get slower? Did the role mix shift toward harder searches? Your trend answers that. A stranger's table cannot. If you want the deeper argument about compressing these numbers without lowering standards, reduce time-to-hire with AI works through which stages actually shrink.

Where the days actually go — and what compresses them

Once you can see both numbers, the next question is where the days pile up. The instinct is to blame the interviews, but interviewing is rarely the slow part. In most pipelines the delay sits in the screening queue: a stack of applications waiting for a human to read them, a role-relevant assessment that has to be built by hand, a panel waiting to align on who is even worth a call. That queue grows with the size of your applicant pool, which is why adding interviewers rarely speeds up a slow pipeline — the jam is forming upstream of the interview slot.

This is where the AI-era shift matters for these metrics. When first-round evaluation is a serial human task, time to hire scales with volume: the more people apply, the longer the queue. An AI-native skills assessment platform changes the shape of that stage by evaluating the whole pipeline against a consistent, job-relevant assessment in parallel rather than one CV at a time. The screening queue stops being the bottleneck, and the days it used to consume come off your time to hire directly. Crucially, the compression comes from the mechanical stage, not from cutting interviews — the human judgement stays where it earns its keep.

The wider point is that speeding these metrics up is a question of evidence per stage, not effort per stage. Teams often add interview rounds out of fear that a CV cannot be trusted — a fear that is more acute now that so many CVs are AI-polished. The fix is better signal earlier, not more conversations later; a single work-sample test run early gives the panel something real to decide on and lets you justify fewer rounds. If you are weighing exactly how many stages a role needs, how many interview rounds makes the case for a defensible default, and the structured interviews guide covers keeping the conversations you do run consistent and fair.

The other direction: when fast becomes reckless

It is worth saying plainly, because a post about hiring speed can read as if faster is always better: it is not. Both metrics measure duration, and duration is only good news when the signal underneath it holds. A pipeline that shaves a week off time to hire by skipping the assessment or collapsing the interview to a single gut-feel chat has not got faster in any way worth celebrating — it has moved the cost downstream to the day the wrong hire underperforms.

That downstream cost is not trivial. A mis-hire carries a real, well-documented price — the US Department of Labor puts the cost of a bad hire at least 30 percent of the person's first-year earnings, and other estimates run to one or two times salary. Weighed against that, shaving a few days off a metric by lowering the bar is a bad trade. The discipline is to compress the stages that were only ever mechanical — the screening queue, assessment setup, scheduling — and leave the judgement stages their room. The cost of a bad hire is the counterweight worth keeping on the same dashboard as your speed metrics, so that fast never quietly becomes reckless.

Time to hire tells you how good your process is at deciding. Time to fill tells you how good your organisation is at committing. Fix the one you can actually influence this quarter, measure it the same way every month, and stop chasing a stranger's average.