Sales Cycle Length Calculator

What you're calculating

Your true average sales cycle = the number of days between a deal being qualified (SQL / Stage 2, whichever your team uses as the true "sales cycle start") and the day it closes-won. Use closed-won deals only — closed-lost deals don't tell you how long a successful sale takes.

Step 1: Pull the data

From your CRM, export closed-won opportunities from the last 4 full quarters with these fields:

• Opportunity ID
• Close date
• Date entered Stage 2 / SQL date (your qualified-pipeline start point)
• Deal amount (ACV or TCV — pick one and be consistent)
• Segment (SMB / Mid-Market / Enterprise, or your equivalent)
• Rep / team

Step 2: Calculate cycle length per deal

`` Cycle Length (days) = Close Date − SQL Date ``

Do this for every row. Keep cycle length as its own column — don't blend deal amounts and dates into one average.

Step 3: Segment the data

Split into a grid — deal size band × customer segment (adjust bands to your ACV distribution):

For each cell, calculate:

• Count of closed-won deals in that cell
• Median cycle length (not mean — see Step 4)
• 10th percentile (your fastest realistic close)
• 90th percentile (your slowest realistic close)

If a cell has fewer than 5 deals, merge it with the nearest adjacent cell — the sample is too small to trust.

Step 4: Use median, not mean

Formula for median: sort all cycle-length values in the cell low to high; the median is the middle value (or the average of the two middle values if the count is even).

Why median wins here: sales cycle data is right-skewed — a handful of 300-day enterprise deals will drag a mean upward and make every rep's forecast look artificially slow. Median resists that distortion.

Worked example

Mid-Market, $10k–$50k ACV, last 4 quarters, 22 closed-won deals. Cycle lengths in days, sorted: 38, 41, 44, 47, 49, 52, 54, 55, 58, 60, 61, 63, 65, 67, 70, 72, 75, 80, 88, 95, 110, 145

• Median (11th/12th values averaged) = (61 + 63) / 2 = 62 days
• 10th percentile ≈ 47 days
• 90th percentile ≈ 110 days

Reading this: a typical Mid-Market $10-50k deal takes 62 days from SQL to close. A rep forecasting a 30-day close on a deal in this band should be challenged — it's below your 10th percentile. A rep forecasting 150 days should also be challenged — check what's actually blocking it.

Step 5: Confidence banding for forecast dates

For any open deal, calculate an expected close-date range:

`` Earliest realistic close = SQL date + 10th percentile (for that segment/size cell) Most likely close = SQL date + median (for that segment/size cell) Latest realistic close = SQL date + 90th percentile (for that segment/size cell) ``

Use "Most likely close" as the default forecasted close date unless the rep has specific stage-based evidence (signed mutual close plan, verbal + contract out) to override it.

Step 6: Stage-by-stage breakdown (optional, higher-value diagnostic)

Repeat the same median calculation for the gap between each stage transition, not just start-to-finish, to see where time is actually spent:

If "Verbal commit → Contract signed" is your longest stage, your bottleneck is legal/procurement, not sales — route effort accordingly.

Outlier exclusion rule

Exclude any deal where cycle length exceeds 3× the segment's median (these are usually re-opened, paused, or data-entry errors, not real cycle behavior) — but log them separately; a cluster of true outliers is itself a signal worth investigating.

Refresh cadence

Recalculate every quarter using a rolling 4-quarter window. Sales cycles shift with pricing changes, new competitors, and macro conditions — a stale benchmark produces confidently wrong forecast dates.