Operations
Part of Sales funnels that hold up: buyer-verifiable stages, not vanity counts
B2B sales funnels questions on conversion and coverage math
Common B2B sales funnels questions on conversion and coverage: worked cohort arithmetic, why rates swing between quarters, and how to size open pipeline.
What to take away
- Stage conversion is each stage's count divided by the count that entered it.
- The same process can move between about 4 and 8 percent inquiry to win across quarters, so one quarter's rate is a reading, not a benchmark.
- Coverage comes from the inverse of the segment's observed win rate, then gets checked against evidence.
- Small cohorts need a stated sample size, because a rate from a few wins is noisy.
- Record the unit, cohort, and rule next to every rate.
How do you turn counts into stage conversion rates?
Divide each stage's count by the count that entered it, then multiply the stage rates for the end-to-end figure. Keep the unit and the cohort fixed, so the same 400 inquiries are followed from first contact to a closed outcome.
| Stage | Count | Rate from previous stage |
|---|---|---|
| Inquiry | 400 | not applicable |
| Discovery call | 240 | 60 percent |
| Qualified opportunity | 96 | 40 percent |
| Closed won | 24 | 25 percent |
Multiplying 60, 40, and 25 percent gives 6 percent inquiry to win. Treat those figures as arithmetic from one illustration, not a benchmark. The method carries over, but it only compares like with like when each stage has stable definitions.
Why does inquiry to win move between quarters?
Small counts and shifting inputs both move it. Across quarters the same process can sit anywhere between about 4 and 8 percent inquiry to win. On the same 400 inquiries, that range is 16 wins at one end and 32 at the other.
Part of the swing is plain noise from small counts. NIST's engineering statistics handbook warns that confidence limits approximated with the normal distribution may not be accurate enough when the sample is very small, so show the sample size beside every rate.
The rest is real change. Rates vary with the unit, stage rule, segment, source, value, product, sales motion, cycle length, geography, and data quality. A team planning a test can use the same handbook's method for the sample size required when testing a proportion.
How do you work out the coverage you need?
Take the inverse of the segment's observed win rate. A target of 100 closed deals at a 25 percent win rate needs about 400 open opportunities, and at a 40 percent win rate it needs about 250. Build the figure from relevant cohorts, as a range.
| Win rate | Open opportunities for 100 deals | Multiplier on target |
|---|---|---|
| 25 percent | about 400 | 4 times |
| 40 percent | about 250 | 2.5 times |
The multiplier sits between roughly 2.5 and 4 times the target in this illustration. A segment with a different observed win rate gets a different multiplier, which is why a pipeline review reads coverage next to cycle length and deal age.
What should you check before trusting the coverage number?
Check whether the open deals contain credible buyer action before you count them. Coverage depends on win evidence, stage, age, segment, value, seller capacity, timing, and forecast reliability, so a large number of records proves little on its own.
- Each opportunity has a named owner and an agreed next buyer action.
- The next decision is dated, and the close date has a stated basis.
- The buying group covers the roles the deal needs.
- Delivery capacity exists if the deals in the range close.
- No record was opened only to meet a pipeline target.
If one item fails, the record may still be a lead rather than an opportunity. A qualification checklist helps settle which side of that line a record sits on, and the lead generation questions answers cover the same boundary.
Should stage probability drive the forecast?
It can inform a forecast, but it should not replace current evidence. Keep historical stage probability, seller judgment, manager judgment, model output, and forecast category in separate columns, then compare each with the cohort figures above.
Examine the next buyer action, stage age, buying-group coverage, commercial and technical risk, and the basis of the close date. Preserve uncertainty rather than presenting false precision. The forecast method itself, with stage tables, is worked through in the pipeline model for B2B demand.
Record the evidence behind each number
For each figure above, record the commercial unit, cohort, entry and exit criteria, supporting evidence, and decision. Review definitions after material process changes, and keep internal rates tied to their stated population.
Decision record checklist
- Connect evaluation questions to evidence needs
- Apply design discipline to funnel questions
- Match funnels to available evidence
- Review governing law and configuration
- Avoid shifting privacy work to individuals
Review definitions quarterly and after a material offer, territory, process, incentive, or system change. A funnel is a management model, not an unchanging law, and the marketing strategy choices upstream of it change what the numbers mean.
Common questions
Is 6 percent inquiry to win a good rate?
There is no universal rate. The 6 percent here is arithmetic from one illustration, so compare a team's rate only with its own history under the same definitions.
Can I compare my rates with another company's?
Only when the unit, stage rule, segment, source, and sales motion match. Otherwise the comparison measures a different funnel.
When should the arithmetic be redone?
Review definitions quarterly and after a material offer, territory, process, incentive, or system change, then rebuild the cohort figures from current records.
