Time to close, also reported as sales cycle length, measures how many days a deal takes from a defined starting point to closed-won. It tells you how long capital, attention, and forecast risk sit inside the pipeline before revenue arrives.
The common formula averages the closed-won cohort:
Time to close = sum of days from stage start to closed-won ÷ number of closed-won deals
Salesforce describes the same method: add up the days each closed deal took, divide by the number of deals. The formula is trivial. The two decisions that make it comparable are the starting stage and the average you report.
Fix the stage-start definition
A clock that starts at lead creation measures marketing response plus selling. A clock that starts at qualification measures the selling system. Both are valid; they are different metrics. Document one starting stage, apply it to every deal, and annotate the series if it changes. HubSpot's own cycle framework runs from prospecting through close, which is exactly why two teams quoting "sales cycle length" can differ by months without either being wrong.
Renewal and expansion deals need their own series. Blending a 14-day renewal motion into a new-logo cycle metric makes both unreadable, and the blend shifts every quarter as the base grows.
Median versus mean
Cycle-length distributions are right-skewed: most deals cluster, a few crawl for quarters. The mean chases the crawlers. The median describes the deal you will most likely see next.
A worked example
Five deals close with cycle lengths of 20, 35, 40, 55, and 210 days.
Mean: (20 + 35 + 40 + 55 + 210) ÷ 5 = 72 days
>
Median: 40 days
One 210-day deal nearly doubles the mean. A forecast built on 72 days misplans four of the five deals. Report both numbers, and when they diverge, investigate the tail instead of averaging it away. The long deals usually share a cause worth knowing: a missing stakeholder, a procurement path nobody mapped, or a segment the team should qualify differently.
Definition pitfalls
- comparing series that start the clock at different stages;
- measuring only closed-won deals and calling it pipeline health: losses and
stalls have cycle times too, and aging open deals are invisible in a won-only view;
- blending new logo, renewal, expansion, and segment mix into one number;
- chasing a shorter cycle by skipping security review or stakeholder work,
which trades cycle time for win rate;
- celebrating a falling average that came from a mix shift toward small
deals;
- ignoring stage-level dwell time, where the actual waiting lives.
Where time to close fits
Time to close is the denominator of pipeline velocity and belongs in the Measurement pillar of the Tenbound Pipeline Architecture Standard.
To act on it, break the cycle into stage dwell times and separate buyer work from avoidable waiting: unclear owners, late legal and security review, proposals sent without an agreed decision process. Shorten the waiting. Leave the buying work alone.