Customer churn is the share of customers (or recurring revenue) that leaves over a period. It is the single most-watched number in subscription businesses because retained revenue compounds and lost revenue rarely comes back. Teams usually track two distinct cuts of it, and confusing them leads to bad decisions.
- Logo churn counts accounts. If you start a month with 200 customers and 6 cancel, logo churn is 3%.
- Gross revenue churn counts dollars lost to cancellations and downgrades, ignoring any expansion. It can never be negative.
A handful of small accounts leaving looks fine on a logo basis but can be brutal on a revenue basis if one of them was a whale. That is why most boards look at gross churn alongside net revenue retention, which folds expansion back in.
Why churn is a lagging metric
By the time a cancellation lands, the decision was usually made weeks or months earlier. The signals that predict it (a quiet champion, falling adoption, a frustrated support thread, a missed QBR) are visible long before the renewal date. Merrily exists to surface those signals so a churn becomes an at-risk account you can still save, not a number you explain after the fact.
Reducing churn is almost always cheaper than replacing the lost revenue with new sales, because the acquisition cost of a saved customer is effectively zero.