What variance tells you that the average hides
Sales cycle variance measures how widely your deals scatter around the average time to close. Two teams can report the same 60 day sales cycle length and run completely different businesses. One closes almost everything between 45 and 75 days. The other closes half its deals in three weeks and drags the rest past six months. The first team can forecast. The second cannot, and the average gives no hint of the difference.Cycle length distributions are rarely symmetric. A deal can close much later than the average but never much earlier than zero, so the tail runs to the right and pulls the mean with it. The mean ends up describing a deal that almost never happens.
How to measure the spread
| Measure | What it gives you | How to read it |
|---|---|---|
| Median days to close | The typical deal | The center of the distribution, unmoved by outliers |
| 75th percentile | The slow but normal deal | Your realistic inspection threshold |
| 90th percentile | The tail | Deals that need a documented reason to stay open |
| Standard deviation | Overall dispersion | Useful for comparing cohorts against each other |
Where variance comes from
Segment mix is the largest source, followed by deal size, since bigger deals carry more approvals and more calendars to align. Market conditions widen the spread on top of both. ORM sees uncertainty in the market translate directly into longer time from qualified to closed, because buyers make fewer decisions when the outlook is unclear. That shift widens the distribution before it moves the average, which is why variance is the earlier warning.
At ORM each opportunity is grouped by a machine learning model, and every group gets its own predicted closing curve. Those curves run from 1 to 80 weeks. Grouping is the practical answer to variance: instead of forcing one average onto a mixed pipeline, each group carries the timing pattern its own deals actually follow.
Why forecasts care about spread
A forecast is a claim about timing as much as about value. High variance means the same pipeline can produce a strong quarter or a weak one depending on which deals happen to land, and no amount of coverage fixes that. Narrowing the distribution improves forecast accuracy more reliably than adding pipeline does, and it makes every downstream commitment easier to defend. Start by splitting the cohorts, then attack the tail.
Frequently Asked Questions
Why is the average sales cycle misleading?
Cycle length distributions have a long right tail. A handful of deals that took 250 days pull the mean well above the point where most deals actually close, so the average describes a deal that rarely exists. Report the median alongside the 75th and 90th percentiles and the shape of the quarter becomes visible.
How do you measure sales cycle variance?
Two methods work. Calculate the standard deviation of days to close within a cohort, or read the percentile spread by comparing the 50th, 75th, and 90th percentile values. The percentile version is easier to act on because each number maps to a real deal you can pull up and inspect.
What causes high sales cycle variance?
Mixing motions inside one number. An SMB deal and an enterprise deal in the same cohort produce a distribution with two peaks and an average that sits in the empty space between them. Segment by deal size, by product, and by lead source before you conclude that the process is inconsistent.
Is low variance more valuable than a short cycle?
For forecasting, yes. A predictable 90 day cycle produces a more reliable quarter than an erratic 60 day one, because the timing of revenue is what a forecast commits to. Compression is worth pursuing after the spread is under control.
Put these metrics to work
ORM builds custom revenue forecast models that turn concepts like sales cycle variance into prescriptive action for your team.
Schedule a Demo