The time from an opportunity's creation to its signature, in days. Read as a median, because a few very long deals drag the average.
In practice
Sales cycle length says how long money spent waits before it becomes a customer. That is why it matters well beyond the sales team: while it runs, the spend is committed and the revenue has not started, and a lengthening cycle weighs on cash before it weighs on revenue. It also sets the lag of the other numbers: a customer signed this quarter was born from an opportunity opened weeks earlier, and from spend made earlier still. Two decisions make the figure, and both should be written down: where it starts — at the opportunity's creation, not at first contact — and which statistic you publish. The median, almost always: out of 18 deals, a single one that drags on for another year moves the average by three weeks while nothing else has changed.
The formula
Sales cycle = median of (signature date − opportunity creation date), in days, over the new-customer deals won in the period
- The start
The day the opportunity was created in the CRM. Not the first website visit, not the first email: those belong upstream. What matters is that the starting point stays the same every quarter — a team that starts opening opportunities only after the demo shortens its cycle on paper, and nowhere else.
- The finish
The signature of a new-customer contract. Lost deals stay out of the median: they would mix how long it takes to win with how long it takes to give up. Their duration is worth reading separately.
- The median
The middle deal once they are sorted by duration — with an even count, the average of the two middle ones. It ignores how far the extremes go, which is exactly the property wanted here: sales cycles have a long tail, and the tail is not the typical deal.
Most CRM reports give an average by default. The median usually takes an export, and is worth the export.
Worked example
64 days, and why the median
- A B2B software company signs 18 new customers from June to August. Sorted by duration, the median is the average of the 9th and 10th deals: 64 days.
- Stretch the longest of the 18 deals by a full year. The average rises by 365 ÷ 18 ≈ 20 days. The median does not move by a single day: the longest deal was above it, and still is.
- What the 64 days do to the other numbers: a contract signed on 1 June after 64 days was opened at the end of March, before the June-to-August window over which the €19,000 customer acquisition cost (CAC) is computed. As long as spend is steady, that lag cancels out; the quarter spend doubles or halves, the CAC describes the previous quarter as much as this one.
- And after the signature, the clock the customer cares about keeps running: here, the time to go live is not measured at all. The buyer waits 64 days to sign, then an unknown time to get what they bought.
Publish the median, say where the clock starts, and read the acquisition cost with the cycle beside it: the longer the cycle, the more this quarter's CAC is a statement about the last one.
Orders of magnitude
- Cycle lengths published by vendors vary with deal size, with the number of people who have to sign, and with where each company starts its clock — and the surveys rarely say which. Two teams quoting the same number of days may not be measuring the same span.
- The order of magnitude that matters is internal: compare the cycle with the window over which you compute your acquisition cost. Under three months, a quarterly CAC is a measure; beyond it, the customers signed this quarter come from spend made in earlier ones, and the CAC is an order of magnitude.
- Your own trend, with the same starting point and the median, is the comparison that holds. When it lengthens, ask three questions before changing the process: did the deals grow, did the buyers change, or did the clock move?
How to improve it
- Find where the days go. Split the cycle by stage — first meeting, proposal, security review, legal — and work on the longest stage, not the one people complain about most.
- Prepare what every buyer asks for: the security questionnaire, the standard contract, the price list. A document sent the same day instead of drafted on request takes that wait off every deal that needs it.
- Close out the deals that no longer move. They do not lengthen the median of won deals, but they eat the time that would shorten it.
- Do not shorten it by opening opportunities later. Moving the starting line improves the number without changing the business — and breaks the comparison with every previous quarter.
In the Tour
Do you know your customer acquisition cost, even roughly?
- Yes, a solid number20 pts
- A rough estimate7 pts
- No idea0 pts
The Tour asks whether you know your customer acquisition cost, even roughly — 20 points for a solid number, 7 for a rough estimate, 0 for no idea. In a sales-assisted business, the sales cycle decides which of the first two answers is honest. A CAC computed over one quarter assumes the customers it counts were won by that quarter's spend; with a cycle longer than the quarter, they were won by the previous one's. A team that knows its median cycle can say whether its "solid number" is a measure or an order of magnitude — which is exactly the gap between 20 points and 7.
Questions people ask
How do you calculate sales cycle length?
For each new-customer deal won in the period, count the days between the opportunity's creation and the signature, then take the median. With 18 deals, the median is the average of the 9th and 10th once they are sorted. Most CRMs export both dates; few compute the median for you.
Median or average?
The median. Cycle durations have a long tail — a handful of deals that take a year — and the average follows the tail: stretching one deal out of 18 by a year adds about 20 days to the average and nothing to the median. Publish the average too if you like, but steer by the median.
What is a normal sales cycle in B2B?
No normal holds across companies: the length follows the size of the deal, the number of people who have to agree, and where the clock starts. The numbers to compare against are your own last quarter, and the window of your acquisition cost — which the cycle should not exceed if that cost is to be read as a measure.
Where does the sales cycle start?
Wherever you decide, as long as it is written down and never moves. The common choice is the opportunity's creation, because every CRM keeps that date. Starting at first contact gives a longer, more complete number that is harder to measure; both are fine, and switching between them is not.