Time-to-fill only becomes a number you can act on when your Applicant Tracking System records the same start moment, the same end moment and every pause in the same way for each vacancy. Start the clock when the vacancy is approved and opened in the Applicant Tracking System, stop it when the chosen candidate accepts the offer, and record any period on hold with a date and a reason. Track the time until the candidate actually starts as a separate figure. Report the median per segment and role type, show how many vacancies sit behind each figure, and compare against your own target fill dates rather than against benchmarks nobody can verify.
The definition stays the same across segments, but what matters most differs. Recruitment agencies want to know how long a client order takes to become an accepted offer. Staffing desks often care more about the days until someone is actually at work. Secondment teams measure towards the start of an assignment. Executive search firms run fewer and longer searches, where a single outlier moves the average. In-house recruitment teams depend on hiring managers whose response times are part of the clock.
Why time-to-fill numbers are often unreliable
The same problems appear on many desks:
- One recruiter starts the clock when the client calls, another when the vacancy is published, a third when the first candidate is added
- Some vacancies end at the accepted offer, others at the start date, and a few only when someone remembers to close them
- Vacancies that were on hold for weeks count as slow, while the real delay was a budget freeze at the client
- Cancelled vacancies are left open or closed without a reason, so they distort the average in both directions
- A single average hides the difference between a two-week staffing order and a five-month executive search
None of this is solved by a nicer dashboard. It is solved by agreeing on definitions and making the Applicant Tracking System record them.
Agree on one start moment
Choose one event that starts the clock for every vacancy and make it a required date in the Applicant Tracking System. For most desks the best start moment is the date the vacancy is approved and opened: for an agency, the date the client order is confirmed; for an in-house team, the date the requisition is approved. That date should be set when the vacancy is created together with a structured brief and a target fill date. Do not let the start date be edited later without a recorded reason, otherwise the number can be improved by moving the start instead of moving faster.
Agree on one end moment, and measure the start date separately
End time-to-fill when the chosen candidate accepts the offer and that acceptance is recorded in the Applicant Tracking System. An acceptance deadline on every offer keeps that date precise. Then measure the period from acceptance to the confirmed start date as a second figure, often called time-to-start. Keeping the two apart shows whether the delay sits in your process or in notice periods, contracts and onboarding at the client.
Pauses, cancellations and reopened vacancies
- On hold — when a client or hiring manager pauses a vacancy, record the start and end of the pause with a reason; report time-to-fill with and without pauses, so both views stay honest
- Cancelled — a vacancy that closes without a hire gets a documented close reason and is reported separately, not mixed into time-to-fill
- Reopened — when a placed candidate drops out before or shortly after starting, open a new interval linked to the original vacancy instead of rewinding the old one
- Several hires on one vacancy — record a fill date per position, so the first hire does not hide how long the last one took
How to report time-to-fill without vanity metrics
- Report the median, not only the average, and always show the number of vacancies behind each figure
- Split by segment, role type and client, because a combined figure answers no practical question
- Compare each vacancy with its own target fill date: how many were filled on time, how many late, and by how much
- Break the total down per stage, so you can see whether time is lost before the first review, while a client submission waits for a reply, while hiring-manager feedback is outstanding, or while candidates are stuck in a stage
- Combine time-to-fill with the candidate source recorded at first application only when the source field is filled reliably
- Publish no figure you cannot trace back to dates in individual vacancy records
An honest median based on twelve vacancies is more useful than a precise-looking average that mixes definitions.
Who owns the dates
Each date needs an owner. The single owner of the role confirms the start date, records pauses and closes the vacancy with the right outcome. Stage changes already record who moved the candidate, so the stage breakdown can be traced without extra work. Repeated steps such as reminders on an approaching target fill date can be handled by the system, as described in which recruitment steps to automate, but the dates themselves are confirmed by people. Keep the client contact person and their agreed response times in the candidate and client relationship records beside the pipeline, so delays on the client side are visible as such.
Where artificial intelligence may help — and where it must not
Artificial intelligence may point out vacancies with missing or inconsistent dates, summarise where time was lost in a long search, or draft the explanation that accompanies a monthly report. It must stay secondary. Artificial intelligence does not replace recruiters, does not fill in or adjust start and end dates on its own, and does not invent market benchmarks to compare against. The Applicant Tracking System holds the recorded dates; the people who own the vacancies confirm them.
A practical week to get started
- Write down the start moment, the end moment and the pause rules in one short definition the whole desk agrees on
- Make the start date, the target fill date and the offer acceptance date required fields in the Applicant Tracking System
- Check the vacancies that are open today and correct missing start dates, with a note where the date is reconstructed
- Record current pauses with a reason and an expected restart date
- Produce a first report per segment with the median, the number of vacancies and the share filled on time, and agree when you will review it next
The first report will be smaller than expected. That is the point: it only contains numbers you can defend.
Measure the same way every time
If every recruiter starts and stops the clock differently, time-to-fill becomes an opinion. Agree on one start moment and one end moment, measure time-to-start separately, record pauses and cancellations with reasons, and report medians per segment against your own target fill dates. Then explore the product or book a demo — we walk through time-to-fill reporting in BoldATS for your segment.