Time to hire vs time to fill: two clocks, what each measures, worked with dates

Time to fill and time to hire are two different clocks. Time to fill counts the days from when a job is approved or opened to when a candidate accepts the offer, so it measures how long the seat stays empty. Time to hire counts the days from when the person you eventually hired entered your pipeline to when they accepted, so it measures how fast your process moves once the right candidate is in it. The difference between the two is the time you spent before that candidate showed up.

This page gives both formulas, works them through on one requisition with real dates, shows a five-role report, and lists the definition choices that make the numbers drift. To shorten the clock once you have measured it, see how to reduce time to fill.

The two formulas

Time to fill  = offer accepted date − requisition approved date   (calendar days)
Time to hire  = offer accepted date − hired candidate's entry date (calendar days)
Sourcing gap  = time to fill − time to hire

Some teams stop both clocks at the start date instead of offer acceptance. That folds the notice period into the number, which neither the recruiter nor the hiring manager controls. It is cleaner to stop at acceptance and report time to start as its own line.

Worked example: one requisition, with dates

Read together, the three numbers tell a story the single 42 would hide. Seventeen days went on finding the candidate. Twenty-five went on moving them from application to acceptance, of which nineteen (March 22 to April 10) sat between the screen and the offer. That middle stretch is interviews, feedback and approvals, and it is usually where the fixable days are.

Reading the gap between the two

The value of tracking both metrics is the comparison. Four patterns come up again and again:

Time to hire also matters to candidates. A strong candidate who applied on day 1 does not care that the requisition opened three weeks earlier. From their side, the clock started when they applied, and every week without a decision is a week in which another employer can make an offer.

A five-role report

Here is an invented quarter for one role family, with both clocks and the gap between them.

The medians come from sorting each column: time to fill 28, 35, 42, 61, 88, so 42; time to hire 18, 21, 25, 30, 34, so 25. The mean time to fill would be (42 + 35 + 61 + 28 + 88) ÷ 5 = 50.8 days, pulled up almost nine days by the controller search.

The report points in two directions at once. Roles C and E had quick processes once the candidate appeared (18 and 34 days) but long searches before that, so the fix for them is sourcing lead time, not interview speed. Role B had almost no sourcing gap but the slowest time to hire in the group: a referred candidate waited 30 days for a decision, which is the kind of delay that loses referrals.

Is there a benchmark?

SHRM's 2026 Recruiting Executives Benchmarking summary reports a median time to fill of 39 calendar days for nonexecutive positions, based on data from over 4,600 organizations. The public summary does not spell out where the clock starts and stops, and it blends every industry and role level, so treat it as context rather than a target. There is no comparable widely published figure for time to hire, partly because the start date depends on how each organization records sourced candidates.

Your own trailing median, by role family, is the more useful comparison. A payroll specialist and a controller should never share a target.

Definition choices that make the numbers drift

Most arguments about time to fill are arguments about definitions. Settle each of these once, write the rule at the top of the report, and do not change it mid-year.

• Approval date or posting date. Starting at approval captures the days a requisition sits unassigned; starting at posting hides them. Approval is the more honest start.
• Sourced candidates. Start time to hire at first two-way contact (a reply or a call), not the day a recruiter viewed a profile. Otherwise sourced hires look slower than inbound ones for no real reason.
• Silver medalists. A candidate who interviewed for an earlier role and was hired for this one: start their clock when they were first considered for this requisition.
• Reopened requisitions. If an accepted offer falls through and the role reopens, either restart the clock and note it, or keep the original date and flag the row. Mixing the two produces numbers nobody can explain.
• Requisitions with several openings. Count each hire separately: three hires against one approval date give three time-to-fill values, not one.
• Evergreen requisitions. High-volume roles that are always open have no meaningful approval date. Report time to hire for them and leave time to fill out.
• Paused requisitions. A hiring freeze of six weeks is not recruiting time. Either subtract documented pause days or report the paused roles separately.

Calculating both in a spreadsheet

You need one row per hire, not per requisition, with four dates. Most ATS exports have them; if yours does not, add the fields now and backfill what you can.

HIRES (one row per hire)
ReqID | Role | RoleFamily | ReqApproved | CandidateEntered |
OfferAccepted | StartDate | Source | Reopened (Y/N)

TimeToFill     =OfferAccepted - ReqApproved
TimeToHire     =OfferAccepted - CandidateEntered
SourcingGap    =TimeToFill - TimeToHire

Median time to fill for one role family (Excel 365, Google Sheets)
=MEDIAN(FILTER(TimeToFill, RoleFamily="Finance"))

Check for negative sourcing gaps. They mean a candidate's entry date is earlier than the requisition's approval date, which happens with pipelined or silver-medalist candidates. Apply the rule you chose above rather than leaving the negative number in the median.

How to report it

A useful monthly or quarterly view has one line per role family and five columns: hires, median time to fill, median time to hire, median sourcing gap, and the slowest role with a one-line reason. Show the count of hires next to every median; a median of three hires is a list, not a trend.

TIME TO FILL AND TIME TO HIRE, Q2
Definitions: fill = req approved to offer accepted; hire = candidate
entered (applied, or first reply if sourced) to offer accepted.
Medians, calendar days.

Role family   Hires  Fill  Hire  Gap   Slowest (reason)
Finance         5     42    25    17   Controller, 88 (sourcing)
Engineering     8     51    29    22   Staff engineer, 97 (pay band)
Support        14     23    12    11   Team lead, 41 (panel scheduling)

Put the process portion in front of the people who own it. Time to hire is largely made of scheduling and feedback waits, which belong to hiring managers as much as to recruiters; a recruiting SLA with hiring managers gives those waits an agreed limit. The sourcing gap belongs to the recruiting team and to whoever sets the pay range. For weekly reporting, the recruiter weekly report template has room for both.

Common mistakes