SaaS Adaptation
How RFM Changes for B2B SaaS
Recency → Risk
Account-Level Recency
Track at account level, not user level. A single active user can mask an otherwise churned account.
Aggregate by account_id
Frequency → Depth
Feature Breadth
Count features used, not just sessions. Frequency alone misses whether they're using core value features.
Core vs peripheral feature split
Monetary → Expansion
Expansion Potential
Monetary in SaaS is about NDR potential: seat growth, upgrade likelihood, and cross-sell readiness.
Seat count, tier, usage headroom
The Failure Mode
User ID vs Account
Most RFM projects fail because they classify noisy user IDs instead of real accounts.
Aggregate to account_id first