Deal cohort analysis groups opportunities by the period they were created, then follows each group forward through its life. A cohort of deals created in March is tracked as a unit: how many advanced, how many closed, at what value, and how many weeks it took. The result is a conversion and timing profile per vintage rather than one blended average that mixes deals created two years ago with deals created last week.
The blended average is the problem it solves. A single win rate calculated across all closed deals assumes pipeline quality has been stable. It rarely has been.
What a cohort curve shows
Each cohort produces a curve of cumulative closes over time. Early weeks are flat, the curve steepens through the middle of the typical cycle, then flattens as the remaining deals age out.
ORM builds this at the group level. Each opportunity is grouped by a machine learning model, and for each group ORM predicts a curve for how long it will take to close. Those curves run from 1 to 80 weeks, with most of the expectation landing before week 12 and very few groups carrying expectation past 52 weeks.
That shape has a direct operational reading. A deal well past its group's expected window is not simply late. It is sitting in the tail of a distribution where closure rarely happens, and it should be valued accordingly rather than carried at full amount.
Reading cohorts against the quarter
Cohort thinking also applies to the pipeline standing in a quarter on day one. ORM reports that roughly 20% of the pipeline carrying in-quarter close dates on the first day of the quarter usually closes, which means 80% of that visible value does not land in the period.
| Question | What the cohort view answers |
|---|---|
| Is pipeline quality improving? | Compare conversion rate across consecutive creation cohorts |
| Are cycles lengthening? | Compare weeks to 50% closure across cohorts |
| Is the carry-in pipeline realistic? | Compare current in-quarter pipeline to prior cohorts' realized share |
Where it changes the forecast
Cohort curves give the forecast a time dimension that stage probability alone does not carry. Two deals in the same stage with the same amount can sit at very different points on their group's curve, and treating them identically overstates the near-term number.
For sales forecasting, the practical use is calibration. Take the cohort created six months ago, compare its realized conversion to what was assumed when those deals were created, and correct the assumption. Repeating that check each period keeps forecast accuracy tied to observed behavior rather than to a stage table someone set years ago.
Frequently Asked Questions
What is deal cohort analysis?
Deal cohort analysis groups opportunities by creation period, usually the month or quarter they entered the pipeline, then tracks each group's conversion rate and time to close as it ages. Instead of one blended win rate, you get a win rate per vintage and can see whether recent pipeline is converting better or worse than older pipeline.
How is deal cohort analysis different from customer cohort analysis?
Customer cohort analysis groups accounts by when they started paying and tracks retention and expansion afterward. Deal cohort analysis groups opportunities by when they were created and tracks them until they close or die. One measures what happens after the sale, the other measures what happens before it.
What cohort size do you need for the analysis to be usable?
Enough deals per cohort that a handful of outcomes does not swing the conversion rate. Teams closing a small number of large deals should cohort by quarter rather than month, and should read the pattern across several cohorts rather than reacting to any single one.
What does deal cohort analysis reveal that a pipeline snapshot cannot?
A snapshot shows what the pipeline looks like today with no memory of how it got there. Cohort analysis shows whether pipeline created six months ago converted at the rate that was assumed at the time, which is the only way to tell whether current pipeline assumptions are calibrated.
Put these metrics to work
ORM builds custom revenue forecast models that turn concepts like deal cohort analysis into prescriptive action for your team.
Schedule a Demo