The hiring funnel numbers worth tracking
Article by Emoti Research. For talent leaders. 5 minute read. Updated .
The short answer
The hiring funnel numbers worth tracking are time to first contact, reply rate, screen pass-through, interview-to-offer, offer acceptance, time to fill, time to hire and source of hire, each with a written definition. Read them together, one role at a time, and be wary of small numbers, which swing widely.
Which numbers are worth tracking?
Eight numbers cover most of the funnel: time to first contact, reply rate, screen pass-through, interview-to-offer, offer acceptance rate, time to fill, time to hire and source of hire. Each needs a written definition before it is worth comparing.
| Metric | Formula | What it tells you |
|---|---|---|
| Time to first contact | Date of the first message to a candidate minus the date they entered the pipeline | How quickly you act on the people you find or who apply |
| Reply rate | People who replied divided by people contacted | Whether outreach reaches the right people with the right message |
| Screen pass-through | People advanced after the first screen divided by people screened | Whether sourcing and screening agree about who fits |
| Interview-to-offer | Offers made divided by people who reached the final interview | Whether the final stage confirms what earlier stages found |
| Offer acceptance rate | Offers accepted divided by offers made | Whether the offer, the pay and the process win people over |
| Time to fill | Date an offer is accepted minus the date the role was opened | How long the team waits for the hire |
| Time to hire | Date an offer is accepted minus the date the eventual hire entered the pipeline | How long the successful candidate's journey took |
| Source of hire | Hires from each source divided by all hires | Where hires actually come from |
Use medians for the time measures. One role that stays open for months will pull an average far away from what is typical.
Write the edge cases into the definitions. Does a bounced email count as contacted? Do offers still waiting for an answer count in the acceptance rate? Is "entered the pipeline" the day someone applied, was found in sourcing, or was first contacted? Any choice is fine, as long as it doesn't change halfway through the year. In Emoti's outreach pages, for example, reply rate is replies divided by everyone whose first message has gone out, including bounces; people still waiting for their first message are left out.
How do time to fill and time to hire differ?
Time to fill measures the role, from opening to an accepted offer. Time to hire measures the person, from the day the eventual hire entered the pipeline to the day they accepted.
The gap between them shows where the time went. A long time to fill with a short time to hire means the right person arrived late: the problem is sourcing, or the role was opened before anyone was ready to search. If both are long, the process itself is slow. Definitions vary between organizations, so say which one you use whenever you share either number.
An example with invented dates: a role opens on March 3. The eventual hire is found in sourcing on April 14 and accepts an offer on May 19. Time to fill is 77 days and time to hire is 35 days. The six weeks before the right person entered the pipeline are a sourcing question, not an interviewing one.
How should you read the numbers together?
As a chain, in order, one role at a time. Each stage's output is the next stage's input, so improving one number often just moves a problem along.
| Pattern | Likely reading | What to check |
|---|---|---|
| Reply rate rises, screen pass-through falls | Outreach is reaching more people who don't fit | The brief and the search criteria |
| Pass-through high, interview-to-offer low | The screen and the final interview test different things | Whether both use the same must-haves |
| Offers made, acceptance low | Pay, speed or the experience lost people at the end | The pay range, and time from final interview to offer |
| Time to first contact long | A backlog between finding people and contacting them | Recruiter capacity and approval steps |
| One source produces most hires | Either a strong source or a narrow one | Pass-through by source, as well as volume |
Read stage counts next to rates. A 50 percent pass-through rate means something different at 4 of 8 than at 40 of 80.
What are the common pitfalls?
- Small numbers. With five candidates at a stage, one person moves the rate by 20 points. Wait for enough people, combine periods, or report counts instead of percentages.
- Mixing roles. A funnel for senior engineers and one for sales development representatives have little in common. Average them and you learn about neither.
- Mixing channels. Email and LinkedIn messages can draw replies at different rates. Report reply rate per channel before you combine them.
- Moving start dates. If time to fill starts when the role is approved for one team and when the job is posted for another, the comparison means nothing.
- Survivorship. Measuring only the people who reached a stage hides everyone who dropped out before it. Track withdrawals and their reasons.
- Targets that bend behavior. Push pass-through up and screens get easier; push time to fill down and offers go to the first acceptable person. Watch the next stage whenever you push on one.
- Ignoring who passes. A stage can look healthy overall and still filter out one group. Compare pass-through rates by group. Under the US Uniform Guidelines on Employee Selection Procedures, a selection rate for any race, sex or ethnic group below four-fifths of the rate for the group with the highest rate is generally regarded as evidence of adverse impact, though differences based on small numbers may not be (29 CFR 1607.4(D)).
How do you set this up without a data team?
Agree on the definitions, record six dates for every candidate, and review the numbers monthly, per role.
- Write the definitions on one page and share it with every recruiter.
- Record six dates per candidate: entered the pipeline, first contact, first screen, final interview, offer, and the offer's outcome.
- Record the source for everyone who enters the pipeline, not only for hires, so you can compare pass-through by source.
- Review monthly per role and quarterly across similar roles.
- Note every change. When you change the brief, the outreach or the screen, write down the date, so later shifts in the numbers have an explanation.
Emoti keeps a count for each role at each stage: sourced, contacted, replied, screened, interviewing and offer. Each outreach sequence shows its own reply rate, and each voice screen shows who is still to finish, who is waiting for review and what has been decided.
What should you not do with these numbers?
Don't compare yourself with someone else's benchmark until you know their definitions, their roles and their market.
We don't publish benchmark figures, because a number without its definition and context invites the wrong comparison. Your own history, for the same kind of role, measured the same way, is the most useful baseline you have. Use the numbers to find questions, such as why acceptance fell for one team this quarter, rather than to grade recruiters. A recruiter on a hard role with a thin market will look slow on every chart, and the chart won't say why.
Key terms
- Time to fill
- The time from opening a role, usually the day the requisition is approved, to the day a candidate accepts an offer.
- Time to hire
- The time from the day the eventual hire entered the pipeline, by applying or being sourced, to the day they accepted the offer.
- Pass-through rate
- The share of candidates at one stage of hiring who move to the next, for example people advanced after a screen divided by people screened.
- Offer acceptance rate
- Offers accepted divided by offers made in a period, leaving out offers still waiting for an answer.
- Source of hire
- Where each hire came from, such as sourcing, referrals, the careers page, job boards or agencies, usually reported as each source's share of hires.
- Reply rate
- Replies divided by people contacted in an outreach sequence or campaign. Teams should decide whether bounced messages count as contacted and use the same rule every time.
More in the glossary.
Questions
What is a good reply rate for recruiting outreach?
It depends on the role, the market, the channel and how you define it, so we don't quote one. Compare a sequence with your own earlier sequences for the same kind of role, measured the same way.
Should we measure quality of hire?
Yes, carefully. Pick one or two signals you can collect consistently, such as the hiring manager's rating at six months or retention at a year, and read them per role, with small numbers in mind.
How often should we review funnel numbers?
Monthly per role is usually enough to act on, and quarterly across roles for trends. Weekly reviews mostly show noise unless volumes are large.
Do candidate withdrawals count against pass-through?
Track them separately. A candidate who withdraws didn't fail a stage, and mixing the two hides why people leave.