How to Calculate Average Sales Cycle Length Without Distorting It

What is the sales cycle length formula?

Sum the days between the start date and the close date for every deal in your sample, then divide by the number of deals.

``` Average sales cycle length = Total days across all deals / Number of deals ```

If five deals took 61, 74, 88, 95, and 210 days, the total is 528 and the mean is 106 days. That number is already misleading, and the reason is the 210-day deal. This is the central problem with cycle length. The arithmetic is trivial and the sample construction decides whether the output is useful.

Cycle length matters because it sets the boundary on what can still close inside a period. A deal qualified in week eight of a quarter, in a motion with a 90-day cycle, is next quarter's revenue no matter what the close date field says.

Which dates should you measure between?

Measure from the stage where a deal becomes qualified to the close date, and apply the same start point to every deal.

The three common start points produce very different numbers.

Qualification stage entry is the right default for forecasting. Lead creation includes nurture time that no seller controls, so a cycle built on it will overstate how long an active deal takes and understate what can close this quarter. Opportunity creation sits in between and is defensible if your reps create records only after real engagement, which is worth verifying before you rely on it.

Whichever you choose, apply it uniformly. A sample that mixes start points is not a measurement.

Should you use the mean or the median?

Use the median as your headline number and report the mean beside it. The mean of this sample is 99 days. The median is 71. Six of the seven deals closed in under 85 days, so the median describes the motion and the mean describes an outlier. Forecasting close dates off 99 days would push nearly every deal a month later than it will actually land.

A wide gap between mean and median usually means two motions are hiding inside one number. Split by deal size or segment and recalculate. If enterprise deals run at 180 days and mid-market runs at 60, neither is served by a blended figure of 99.

Should lost deals be included?

Calculate both versions, because they answer different questions.

Won-deal cycle length is the forecasting number. It tells you how long a deal takes when it works, which is what you need to judge whether an open opportunity can close in the period.

Won plus lost cycle length is the capacity number. It tells you how long deals occupy rep time regardless of outcome, which is what you need for headcount and territory planning. Lost deals typically run longer than won deals, because deals that stall eventually get marked lost after months of no movement.

Never blend the two into a single reported figure. A cycle length that mixes them will be longer than the truth for forecasting and shorter than the truth for capacity.

How should cycle length vary by segment?

Split by segment, deal size, and lead source, then use the specific number rather than the blended one.

Deals are not uniformly distributed around a single average. Opportunities can be grouped by their shared characteristics, and each group carries its own close curve. In ORM models those curves run from 1 to 80 weeks, with most of the expectation concentrated before week 12 and very few groups showing expectation past 52 weeks. That structure is the practical argument against a single company-wide cycle length. A deal in a fast group and a deal in a slow group are not both going to close in 74 days because the average said so.

The operational version is straightforward. Calculate median cycle length per segment, per source, and per deal size band. Then judge each open opportunity against the cycle of its own group. Cycle length is also one of the four inputs in the sales velocity formula, so a blended figure there corrupts that calculation too.

What makes sales cycle length grow?

Changes in buying conditions, changes in deal mix, and stalled deals that nobody has closed out.

Market conditions are the most common cause and the least visible in the data. Uncertainty produces fewer decisions, which stretches the time from qualified to closed. The same effect appears when a new competitor enters and buyers start running additional evaluations. Neither shows up as a CRM field change. Both show up as a cycle that is materially longer than last year with no obvious internal cause.

Deal mix is the second. If your segment mix shifted toward enterprise, your blended cycle length grew without any individual motion slowing down. Check the mix before you conclude that execution declined.

Stalled deals are the third and the easiest to fix. Across ORM customers, 10% or more of open pipeline has not been touched in 12 months. Those records eventually close as losses, and when they do they enter the cycle length calculation with enormous day counts. Treat a change in stage, close date, or amount as meaningful activity, and close out anything that has shown none of the three for a year. Persistent close-date movement is a related signal, covered in the deal slippage definition.

How does cycle length change what you forecast?

It sets a hard filter on which open deals can realistically land in the current period.

Take the median cycle for the deal's own segment, subtract the days it has already been open, and compare the remainder against the days left in the quarter. A mid-market deal with a 68-day median that was qualified 12 days ago needs roughly 56 more days. If 40 days remain in the quarter, that deal is not this quarter's revenue regardless of what the close date says or how confident the rep sounds.

Run that filter across the pipeline and the in-quarter picture usually shrinks. That is the point. It is better to know the shape of the quarter on day one, when there is still time to create pipeline or accelerate what is already in flight, than to discover it in the final two weeks. The process for building that view is covered in how to create a sales forecast.