Loading
Because grouping by close date mixes cohorts. A strong quarter can be a weak quarter that happened to close deals started nine months earlier, and the report will call it growth.
Cohort analysis groups deals by the period they were created and follows each group forward through conversion, cycle length and value.
Each cohort is followed forward, so conversion at 30, 60 and 90 days is comparable between periods, and a change in cycle length is visible while the cohort is still open rather than two quarters later.
Q1 closed 18% above Q4 and looked like improvement. By cohort, Q1's own creations were converting slower than Q4's — the quarter was living off an older cohort, and pipeline creation had been falling for months.
It shows differences between groups; it does not explain them. A slower cohort might be a market change, a pricing change or a new rep ramping, and the report cannot tell you which.
Books reads it for revenue by acquisition period, source tracking supplies the channel dimension, and win/loss reasons explain the differences between cohorts rather than just showing them.
Creation period by default. You can cohort by source, product or first-touch campaign instead.
Yes, cohort against equivalent cohort, which is the only comparison that means much.
Yes. A cohort that never converts is the finding.
Fourteen days, every module, no card. Or half an hour with someone who will run it on your own records and tell you where it does not help.