Monthly Recurring Revenue (MRR) is the normalized monthly value of every active subscription at a point in time. The normalization is the load-bearing step: a brand running weekly, monthly, quarterly, and annual cadences in the same subscriber base has to convert each one to its monthly equivalent before summing. Weekly billings multiply by 52/12 ≈ 4.33 (some operators use a flat 4 for simplicity, which understates MRR by roughly 8%). Quarterly plans divide by 3, semi-annual by 6, annual by 12. The sum across the active base is the brand’s MRR for that snapshot.
A single MRR number is the topline; the diagnostic value sits in how it moves. The month-over-month decomposition is:
MRR (period-over-period decomposition)
Current MRR = Prior MRR + New + Expansion − Contraction − Churned
Topline can mask rising churn offset by stronger acquisition, or expansion masking a leaking base.
New MRR is first-time subscribers in the period; expansion MRR is existing subscribers upgrading tier, adding SKUs, or moving to a higher frequency; contraction MRR is downgrades and frequency reductions; churned MRR is cancellations and dunning failures that aged out. A finance lead or subscription PM reads MRR through that five-component view because a flat headline can hide rising churn offset by stronger acquisition, or expansion masking a leaking base. The components answer questions the topline can’t.
DTC replenishment introduces a behavior SaaS doesn’t have to model: skip, pause, and snooze. A customer who skips a month is still active — the subscription continues, that month’s revenue is zero. That belongs in contraction MRR, not churn. A pause is the same shape with no defined resume date. Per brand policy, a skip or pause that doesn’t resume within a window — often somewhere in the 60–90 day range, though it varies by platform and brand — ages out and becomes churn at that aging moment, not at the skip moment. Conflating skip with churn overstates churn early and understates it late; conflating skip with steady-state contraction misses that paused cohorts historically resume at lower rates than active ones — a leading indicator worth tracking on its own.
What counts as “active” determines whether the MRR figure is honest. If paused and skipped accounts are included at their would-be cadence value, the brand reports phantom MRR against a base that isn’t paying. If they’re excluded entirely, headline MRR is clean but the pause cohort disappears from view. The operator move is to report both: headline MRR on truly active subscriptions, with a paused-MRR sidecar showing what’s at risk. That definition also anchors the relationship to adjacent retention metrics. MRR is a point-in-time monthly run rate of the active subscription base. NRR is closed-cohort retention over a defined period. Churn rate is the share of subscribers lost over a period. The three answer different questions, and an MRR trend built on a loose “active” definition distorts every retention number that reads from it.