Pipeline conversion rates by stage: how to calculate them and find the leak

Pipeline conversion rates by stage show what share of candidates move from each stage of your hiring process to the next: candidates who reached stage N+1 ÷ candidates who reached stage N, measured on the same cohort. Laid out side by side, they show where candidates leave the process and, if you record why they left, whether the leak is your screening, your speed or your offer.

This page defines the stages, works a quarter's funnel through, compares it with the previous quarter to find the stage that changed, and gives a report layout. The same rates run backwards become a sourcing plan; the recruiting forecast model shows how.

Define the stages first

Conversion rates are only comparable if every recruiter moves candidates through the same stages the same way. A typical in-house funnel:

Define each stage by an event that happened ("interview held"), not an intention ("interview scheduled"). Scheduled interviews that never take place otherwise inflate the later stage. Agencies have extra stages, such as submitted to client; the stage from submittal to client interview is covered in how to improve submittal-to-interview ratio.

Stage conversion and cumulative conversion

Stage conversion      = candidates reaching stage N+1 ÷ candidates reaching stage N
Cumulative conversion = candidates reaching stage N ÷ candidates at stage 1

Both on one cohort: every candidate who entered stage 1 in the period,
followed to an outcome.

Stage conversion tells you where the funnel narrows. Cumulative conversion tells you how many people you need at the top to get one hire at the bottom, which is the number for planning.

Worked example: one quarter's funnel

Invented data: account executive roles, everyone who applied or was sourced in Q2, followed to an outcome by the end of August.

Four hires from 600 candidates: 150 candidates at the top for every hire. The stage that stands out is passed screen to first interview, at 66.7%. A third of the candidates the recruiter advanced never had a first interview. Either the hiring manager declined them on paper, or they dropped out while waiting. The rate alone cannot tell which.

Split every exit into rejected or withdrew

A conversion rate counts everyone who did not move forward. The fix depends on why. Record an exit type for every candidate who leaves: rejected by us, withdrew, or no response.

Now the leak is clear. Of the twelve candidates lost between the screen and the first interview, nine withdrew or went silent; only three were rejected by the hiring manager. That is a speed and communication problem, not a screening one. Looking at the dates would likely show a long wait between passing the screen and being offered an interview slot. The fixes are in how to reduce time to fill (the scheduling wait) and how to handle candidate ghosting.

Compare that with the screen itself: 46 of 54 exits were rejections, which is the screen doing its job.

Compare two quarters to find what changed

A single funnel shows where candidates leave. Two funnels side by side show what changed.

Every transition is roughly stable except one, which fell by 22 points. Q1 had 5 hires from 520 candidates; Q2 had 4 from 600. Had Q2's screen-to-interview rate matched Q1's, about 32 candidates would have reached a first interview instead of 24, and at the later rates that is roughly one more hire. Something changed between Q1 and Q2 at that step: a new hiring manager, a holiday period, a new scheduling process. That is where to ask.

With counts this small, a 3-point move at the later stages is one candidate and means nothing. The 22-point drop is twelve candidates lost at a stage that lost four the quarter before, which is worth a conversation.

Cohort counting and candidates still in process

Count on a cohort: take the candidates who entered in the quarter and follow each of them, however long it takes. If you count stage events that happened in the quarter instead, Q2's offers partly came from Q1's applicants, and rates can exceed 100% on small funnels.

The cost is that a cohort is not final until its candidates reach an outcome. When you report, mark candidates still active at each stage and leave them out of the conversion for that stage until they move or exit. Reporting a quarter four to eight weeks after it closes, depending on your typical time to hire, gives a nearly complete picture.

Other pitfalls

• Skipped stages. An internal candidate who goes straight to a final interview has no screen. Count them as having passed through the skipped stages, or exclude them from the funnel; do not let them make a later stage larger than an earlier one.
• Candidates considered for several roles. Count each candidate once per requisition.
• Stage statuses not updated. If recruiters move candidates straight from "applied" to "rejected" without recording the screen, the screen stage is undercounted. Audit a sample each quarter.
• Blending role families. An engineering funnel and a support funnel have different shapes. Combining them hides both.
• Applied-to-screened on high-volume roles. This rate mostly reflects how many applications the team has capacity to review, not candidate quality. Read it next to recruiter load.

How to report it

PIPELINE CONVERSION, ACCOUNT EXECUTIVE, Q2 COHORT (as of Aug 31)
Stages by event held; exits typed rejected / withdrew / no response.
Still active: 0.

Stage                     n     Stage %   Cum. %   Q1 stage %
Applied or sourced       600      —       100       —
Screened                  90    15.0      15.0     16.2
Passed screen             36    40.0       6.0     41.7
First interview           24    66.7       4.0     88.6   <-- 9 of 12 exits
Final interview           10    41.7       1.7     38.7       withdrew or
Offer                      5    50.0       0.8     50.0       no response
Accepted / started         4    80.0       0.7     83.3

Median days, passed screen to first interview: 11 (Q1: 5)
Action: hiring manager holds two interview slots per week from Sept 8.

The median wait line turns the conversion drop into a cause. For the stage right before the offer, the offer acceptance rate page covers how to count and report it in more detail, and the interview-to-hire ratio turns the interview stages into interviewer hours.

Common mistakes