Blog · Finance metrics and formulas · Sales teams
Sales cycle length measures elapsed days from a defined starting event to close. Calculate the mean and median, reconcile stage time and keep open-deal aging separate from completed durations.
Sales cycle length is the number of elapsed days from a defined starting event to a deal's close. Calculate the average and median for a stated group of deals, then separate changes in the selling process from changes in the mix of deals closing. Keep open opportunities visible because completed deals cannot describe their eventual timing.
Choose a start that is recorded consistently: first lead creation, accepted qualification or opportunity creation. Each answers a different question. A company that changes from lead creation to qualification will shorten its reported cycle even if buyers move at the same speed.
Choose the endpoint too. This example uses the closed-won date under the sales team's bookings policy. Invoice date and cash receipt date belong to later processes and should not replace it halfway through a comparison.
Document whether the cohort consists of deals that closed during the period or deals that started during it. A close-date cohort describes completed work. A start-date cohort follows a group forward and must retain the unresolved opportunities. For the broader process and its handoffs, see sales life cycle.
Cycle days per closed deal = Close date - Defined start date
Average sales cycle length = Sum of cycle days / Number of included closed deals
Median cycle length = Middle duration after sorting; for an even count, average the two middle durations
Use calendar days here, with a same-day start and close equal to zero. Export opportunity ID, start date, actual close date, outcome and segment. Use stage histories for the later breakdown; the current stage alone cannot reconstruct past residence time.
Check that each opportunity appears once in the completed-deal calculation. Reject a close date earlier than the start, and investigate missing dates rather than substituting zero. Keep lost records in a separate outcome view unless the stated measure explicitly includes them.
An illustrative US business services company has these six wins. Every start date is June 1, 2026; all dates below are in 2026. The observation cutoff is September 30, so each listed deal is complete.
| Opportunity | Segment | Start date | Closed-won date | Elapsed days |
|---|---|---|---|---|
| A | Small business | June 1 | June 11 | 10 |
| B | Small business | June 1 | June 21 | 20 |
| C | Small business | June 1 | July 1 | 30 |
| D | Enterprise | June 1 | July 11 | 40 |
| E | Enterprise | June 1 | August 10 | 70 |
| F | Enterprise | June 1 | August 30 | 90 |
| Total | 260 |
With start dates in C and close dates in D, the Excel formula in E2 is:
=DAYS(D2,C2)
Fill it through E7. Microsoft documents that DAYS uses the ending date minus the starting date. Store complete dates including the year, even when the display shows only month and day.
The summary formulas are:
=SUM(E2:E7)/COUNT(E2:E7)
=MEDIAN(E2:E7)
The mean is 260 / 6 = 43.3 days, rounded to one decimal place. Sorted durations are 10, 20, 30, 40, 70 and 90 days, so the median is (30 + 40) / 2 = 35 days. Microsoft's MEDIAN documentation explains the middle-pair calculation, and notes that empty referenced cells are ignored while zeros are included. Covirage is not affiliated with Microsoft.
That last distinction matters. A missing start date must become a visible data-quality exception before the summary is trusted. A total that looks plausible is not evidence that all dates were valid.
The same histories provide the following intervals. For this simplified process, qualification-to-proposal and proposal-to-close are consecutive, nonoverlapping intervals that cover the whole cycle.
| Opportunity | Qualification to proposal, days | Proposal to close, days | Check: sum equals cycle days |
|---|---|---|---|
| A | 3 | 7 | 10 |
| B | 6 | 14 | 20 |
| C | 10 | 20 | 30 |
| D | 10 | 30 | 40 |
| E | 20 | 50 | 70 |
| F | 30 | 60 | 90 |
| Total | 79 | 181 | 260 |
The identity is 79 + 181 = 260 days. Mean stage times are 79 / 6 = 13.2 days and 181 / 6 = 30.2 days, rounded separately. Their unrounded means sum to the unrounded overall mean; the displayed rounded values need not add to the displayed total.
Proposal-to-close accounts for more elapsed time in these rows. That identifies where to inspect next, but does not establish avoidable delay. Separate time awaiting a buyer's approval from time awaiting an internal action before assigning an intervention.
For repeated stage visits, include every residence interval once. Gaps in the recorded stages must remain visible, and overlapping intervals must be resolved before stage days can reconcile to total elapsed time.
From the first table, small-business mean cycle length is (10 + 20 + 30) / 3 = 20 days; its median is also 20. Enterprise mean cycle length is (40 + 70 + 90) / 3 = 66.7 days, while its median is 70 days. Those are descriptions of these deals, not benchmarks for other companies.
Even the date filter changes the answer. Restricting the close-date cohort to July through September includes only C, D, E and F. Their total is 30 + 40 + 70 + 90 = 230 days, giving a mean of 230 / 4 = 57.5 days and a median of (40 + 70) / 2 = 55 days.
No duration changed. The later closing cohort simply contains a different set of deals. When comparing periods or peers, align start event, outcome, sales motion, segment and date filter before interpreting the gap.
The sales velocity formula uses mean cycle days alongside opportunity count, win rate and deal value. The worked comparison across two quarters shows how to separate those effects instead of attributing every revenue change to timing.
Suppose the same export also contains these unresolved opportunities at September 30, 2026:
| Opportunity | Defined start | Current stage entered | Total age, days | Current-stage age, days |
|---|---|---|---|---|
| G | July 1 | July 10 | 91 | 82 |
| H | September 1 | September 1 | 29 | 29 |
These ages are September 30 minus the stated dates. G is already older than any completed deal in the example, but its final duration remains unknown. Adding its current age to completed durations would treat an unfinished observation as finished.
Use this view for next-action reviews and follow the original starting cohort until more outcomes are known. A falling mean among completed deals can coexist with an aging unresolved population.
Bring an opportunity export and its dated stage history to Covirage. Covirage maps the columns, and its tools compute and check mean days, medians and stage totals; the external AI model explains the results with every figure traced to its rows. Create a free Covirage account and bring a sample of your own export.
See sales insights for the supporting workflow.
Either convention can answer a useful question, but label it clearly. Won-only duration measures the timing of successful deals. A separate loss-duration measure shows how long unsuccessful work remains active. Combining outcomes can obscure both patterns when the win rate changes.
The median describes the middle completed deal and is less affected by a few long durations. The mean uses every deal's days and supports aggregate calculations such as sales velocity. Report both, with deal counts and a consistent cohort definition.
Calendar days are a clear default for elapsed customer buying time. Business days can help analyze internal working capacity, but require a defined weekend and holiday calendar. Use the same convention across periods and state it next to the result.
Their final duration is not yet known, so do not treat their current ages as completed cycle lengths. Show open-deal age separately and retain the original starting cohort. Revisit that cohort as outcomes arrive, particularly when older unresolved opportunities are material.