Your go-to tech blog

Why Your Time to Hire and Time to Fill Numbers Disagree, and How to Calculate Both

The head of engineering says the last backend hire took 74 days. The recruiter says it took 23. They are both looking at the same person, who accepted an offer on 12 May, and both numbers are correct. The engineering lead is counting from the day the headcount was approved in early March.

The recruiter is counting from the day the eventual hire applied. One is time to fill, the other is time to hire, and the quarterly review is about to spend twenty minutes arguing about which one is “the real number”.

That argument fails because the two metrics answer different questions. Here is where each clock starts and stops, how to calculate them, a calculator you can run on one hire and the reporting mistakes that make either number meaningless.

Two clocks on one hire

Time to hire usually runs from the moment the eventual hire enters your pipeline (they apply, or a recruiter first contacts them) to the moment they accept the offer. It measures how fast your process moves a candidate once you have found them.

Time to fill usually runs from the moment the requisition is approved, or in some companies the moment the job is posted, to the moment an offer is accepted. It measures how long the team goes without the person it decided it needed.

The word “usually” matters. Definitions vary by organisation, by tracking system and even between teams in one company. Some count time to fill until the start date rather than offer acceptance. Some start time to hire at the first interview rather than at application. Some use business days, others calendar days.

None of these is wrong. What is wrong is comparing your number with a benchmark, a previous quarter or another team without checking that the start and stop events match. Before reporting either metric, write down the exact two events you use, and put that definition in the report footer.

Both clocks normally stop at the same event, offer acceptance. The difference between them is everything that happens before your eventual hire shows up: approvals, writing the job description, posting, sourcing and waiting.

Why the two numbers drift apart

Take the 74 versus 23 days. Counted without the first day, the figures are 73 and 23, and the 50-day difference sits in two periods the recruiter’s metric never sees, plus one effect of how the metrics are defined.

First, the gap between approval and posting. Requisition approved on 28 February, job description agreed and posted on 14 March: 14 days. This is pure internal delay.

Second, the sourcing gap. The job was live from 14 March, but the person eventually hired did not apply until 19 April, 36 days later. Other candidates filled that period, but none of it shows in the hired person’s clock.

Third, the order of events. Time to hire is a per-candidate number, while time to fill is a per-requisition number. If your first-choice candidate declined after six weeks and you then hired someone fast, time to hire looks excellent and time to fill looks terrible. Both are honest.

That is why time to hire alone can make a struggling hiring process look healthy. A short time to hire combined with a long time to fill usually points to slow approvals, slow posting or weak sourcing rather than a slow interview process.

How the metrics compare

AspectTime to hireTime to fill
Typical start eventEventual hire applies or is first contactedRequisition approved, or job posted
Typical end eventOffer acceptedOffer accepted (some use start date)
UnitPer candidatePer requisition
Main ownerRecruiter and interview panelHiring manager, recruiting and approvers together
What it revealsSpeed of screening, interviews, decisions and offersTotal cost of the vacancy to the team
What it hidesApproval delays, sourcing time, failed earlier candidatesWhether the delay was internal process or a thin market
Best used forCandidate experience, interview loop efficiencyWorkforce planning, setting expectations with managers

A useful rule: report time to fill upward to leadership and finance, because it reflects how long the business carries an empty seat. Use time to hire inside recruiting, because it is the part of the clock the interview process actually controls. If you can only afford to track one, track time to fill with the stage split, since time to hire can be derived from the same four dates.

Calculating both, step by step

  1. Fix your definitions. Choose the start event for each metric and the stop event. Write them down. For this example: time to fill starts at requisition approval, time to hire starts at application, both stop at offer acceptance, calendar days.
  2. Collect four dates per hire. Requisition approval, job posting, the hired candidate’s application or first contact date, and offer acceptance. Add the start date if you also track time to start.
  3. Subtract. Time to fill equals offer acceptance minus requisition approval. Time to hire equals offer acceptance minus application date.
  4. Split the gap. Approval to posting, and posting to application, explain where the extra days went.
  5. Aggregate with medians. For a quarter, report the median of each metric across hires, plus the count of hires.
  6. Segment. Split by role family or level. A single company-wide figure blends senior searches with high-volume roles.

Using the backend example: approval 28 February, posting 14 March, application 19 April, offer accepted 12 May, all in the same year. Time to fill is 73 days on a calendar count. Time to hire is 23 days. Approval to posting is 14 days, posting to application is 36 days. The lead’s 74 counted both the first and last day; settle that in step 1.

Time to hire and time to fill calculator

Enter the dates for one hire. The calculator uses calendar days and does not count the start date, so a hire approved on the 1st and accepted on the 2nd is 1 day.

With the default dates it shows a time to fill of 73 days, a time to hire of 23, and tells you that only about 32% of the vacancy was spent on the hired person’s own process. The other 50 days are approvals, posting and sourcing. The formulas are the ones from the previous section: time to fill is acceptance date minus approval date, time to hire is acceptance date minus the candidate’s entry date. Run it on your last 5 hires before trusting any quarterly average. If the share spent on the candidate’s own process is consistently under 40%, the delay is mostly upstream of the interviews. If you use posting date as the start of time to fill, simply enter the posting date in the approval field.

Mistakes that make the numbers meaningless

  • Mixing definitions across teams. One team stops at start date, another at acceptance. The company average then measures nothing.
  • Using the mean on a small sample. In a quarter of 8 hires where 7 took about 50 days, one executive search of 180 days pulls the mean up by more than 15 days. Report the median.
  • Using the job posting date for sourced candidates. A candidate first contacted on LinkedIn three weeks before they formally applied has a time to hire that starts at first contact, if that is your definition. Use the earliest date you actually touched them.
  • Counting cancelled or paused requisitions. A requisition frozen for two months during a budget review should be reported separately or excluded with a note.
  • Ignoring the offer stage. Days between final interview and offer often hide the biggest controllable delay, usually waiting on approvals for compensation.
  • Treating speed as the goal. A falling time to hire with rising early attrition is not an improvement.

What the gap between them tells you

Once both numbers are reliable, the gap is diagnostic. A rough reading:

If time to hire is long, the interview process is slow. Look at days between stages, especially scheduling delays and decision time after the final round.

If time to hire is short but time to fill is long, the delay is before the hired candidate arrived. Check approval to posting first, because it is entirely internal and often the cheapest to fix. If posting to application is long, the problem is sourcing: the role is poorly advertised, the band is below market, or the requirements are narrower than the market can supply.

If both are long, start with the interview process anyway, since a slow process also causes strong candidates to drop out, which lengthens sourcing.

Track the ratio over time as well. In the example, 23 of 73 days is about a third. If that share keeps shrinking quarter after quarter while time to fill stays flat, recruiting is getting faster and the rest of the organisation is not.

How these metrics come up in interviews

Recruiter, talent acquisition partner, recruiting operations and HR analytics interviews often include “What’s the difference between time to hire and time to fill?” or “Our time to fill is 60 days; how would you bring it down?” A strong answer defines both with start and stop events, says that definitions vary and should be written down, and then breaks the problem into stages instead of jumping to a solution.

Hiring managers get a version too: “How long does it take you to fill a role, and what slows it down?” Here the strong answer separates what you control (writing the job description quickly, holding interview slots, deciding within 48 hours of a final round) from what you do not.

The trade-off is straightforward. Time to hire is the more actionable metric for recruiters and the more flattering one. Time to fill is less actionable and closer to what the business feels. Report time to fill to leadership so nobody mistakes a fast interview loop for a fast hire, manage the interview process with time to hire, and keep the definitions in writing so the next quarterly review argues about the delay instead of the arithmetic.