What it means
A company created 100 qualified opportunities in January and 100 in April; by June, January has 30 wins and April has 10, but comparing both as of June unfairly gives January more time, so compare them at two months of age instead. Define the start event, because a lead creation date, qualification date and opportunity creation date produce different cohorts, and choose the one tied to the decision.
Define the endpoint as well, since a demo booked, proposal accepted and deal won are different conversions, and a team may need several stage metrics, each with its own denominator. Metabase's conversion-rate guidance describes overall won deals divided by created deals in a cohort and stage-to-stage measures based on stage history, and it warns that a current CRM snapshot can miss deals that already passed through a stage.
Stripe's cohort-analysis guide explains how shared start dates reveal patterns that aggregate totals can hide, which is especially important in sales when the buying cycle spans months. Keep original membership stable, so a deal created in January does not move to April simply because its close date changed, and retain a separate current status field.
Choose equal exposure windows, comparing 30-day outcomes for each cohort or waiting until both mature, and mark recent groups incomplete rather than assigning a final low rate. An illustrative two-month win rate is deals won within two months divided by deals originally created in the cohort, so if 20 of 100 qualify, the rate is 20%.
Keep lost deals in the denominator, define withdrawn, duplicate and test records separately, and say how reopened deals count; track the denominator even when an opportunity is merged or reassigned, so a CRM cleanup does not look like an improvement that did not happen. Track deal value as well as counts, since ten small wins and one large win produce different revenue, and segment cautiously by lead source, product, region and deal size, showing the count beside every percentage.
Check CRM hygiene, because duplicates, imported historical leads and missing stage dates can distort the cohort. Distinguish conversion from causation: a better April rate after an onboarding change may reflect more qualified leads or a different market, and a campaign that floods the pipeline with low-intent records may lower the percentage while adding some valuable sales.
A two-month window can penalise enterprise deals that normally take six months, and a new qualification rule can make later cohorts look better by excluding weaker leads earlier, so keep a dated snapshot of the stage definitions. For an owner, sales cohorts show whether newer groups move through the funnel more effectively at comparable ages, and they support better decisions when entry rules, sample sizes and deal values are visible.
In practice
Real-world examples.
Example
January and April opportunities are compared at two months after creation. The analyst takes the January count of wins as of the end of March, not the end of June. The comparison now shows how the two groups behaved over the same elapsed time.
Example
A stage conversion uses historical stage-change records, not only current status. A deal that reached proposal in February and was lost in March still counts as a proposal-stage entrant. The report therefore shows how many deals actually passed through each stage.
Example
A recent cohort is marked incomplete while its deals remain open. The dashboard shows the count of open deals beside the current rate. Leaders avoid declaring a final low conversion before the buying cycle has run its course.
Formula
Calculation
Two-month conversion = original cohort deals won within two months / original cohort deals x 100.
Worked example: the January cohort has 100 opportunities and 20 are won within two months, so its two-month conversion is 20 / 100 x 100 = 20%. The April cohort also has 100 opportunities and 10 are won by the end of June, exactly two months later, so its rate is 10%. Compared as of June, the snapshot shows 30% against 10%, a 20-point gap; compared at equal age, it is 20% against 10%, a 10-point gap, which is the fair figure.Case study
Seen in the real world.
In this entirely fictional example, Cedar Systems sees a lower win rate for a recent cohort. It compares both periods at 60 days and finds the gap smaller, then checks lead source and deal size. The team does not discard open recent opportunities or declare the new process a failure without evidence.
The fictional numbers show why. The older cohort of 200 opportunities has 60 wins today, 30%, but only 44 of them closed within 60 days, which is 22%. The recent cohort of 200 has 36 wins at 60 days, 18%, so the fair gap is 4 points rather than the 12 points the raw comparison suggested.
Watch out
Common mistakes.
- Comparing new cohorts with mature ones at the same calendar date.
- Removing lost deals from the original denominator.
- Using only current CRM stages and losing past stage movements.
Questions
People also ask.
What defines a cohort?
A shared entry period and event, such as opportunity creation.
Why compare equal age?
Older deals had more time to convert.
Does a higher rate prove more profit?
No. Deal value, acquisition cost and service economics also matter.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%