Your NRR Is 94.6%. That Sounds Fine. It Isn’t.
On paper you’re retaining 94 cents of every revenue dollar. In practice you’re adding 250 customers a month and losing 207. The cohort model shows exactly where — and when — the exit happens.
Four metrics. All look fine. None tell the truth.
These are the numbers most mid-market SaaS teams report to leadership. Each one is technically accurate. Together they hide a slow, compounding revenue bleed.
1,247
250
8.3%
94.6%
NRR of 94.6% means your revenue base erodes 5.4% per month net. Gross churn is 8.3% — expansion revenue is masking the problem, not solving it. You are losing 103 customers every single month.
None of these metrics tell you when customers leave, which segments churn fastest, why they cancel, or who is next. That is what is missing.
67%
of churned accounts showed disengagement signals 30+ days before cancelling
$167K
in ARR recoverable by improving NRR from 94.6% to 98%
0 accounts
proactively flagged as at-risk before cancellation this quarter
Four questions your summary metrics never answer.
We connect your billing system, CRM, product usage events, and support data into a single retention model — then surface each answer as an actionable deliverable.
Cohort Retention
When exactly do you lose customers?
NRR Decomposition
Is expansion hiding gross churn?
Segment Analysis
Which channels produce loyal customers?
Health Scoring
Who is about to leave right now?
The retention picture, fully assembled.
Every panel below is derived from the same connected data model — billing, CRM, product events, and support — with zero manual work.
NRR Decomposition — Current Month
Gross Revenue Churn
−8.3%/mo
103 customers · $35,020 MRR
Expansion MRR
+4.1%/mo
Upsells · seat expansions
Contraction MRR
−1.2%/mo
Downgrades · plan changes
Net Revenue Retention
94.6%
Losing ground. Target: ≥ 100%
Cohort Retention Heatmap — % of Original Customers Still Active
| Cohort | M+0 | M+1 | M+2 | M+3 | M+6 | M+12 |
|---|---|---|---|---|---|---|
| Jan '24 | 100% | 73% | 59% | 51% | 39% | 27% |
| Feb '24 | 100% | 70% | 55% | 46% | 36% | — |
| Mar '24 | 100% | 75% | 62% | 53% | 42% | — |
| Apr '24 | 100% | 68% | 54% | 44% | — | — |
| May '24 | 100% | 77% | 64% | 55% | — | — |
| Jun '24 | 100% | 80% | 68% | — | — | — |
Read across (horizontal): Most cohorts lose 25–30% of customers in Month 1 alone. That is an onboarding problem, not a product problem.
Read down (vertical): M+1 retention improved from 73% (Jan) to 80% (Jun). Retention is getting better cohort-over-cohort — there is a product story here worth surfacing.
Retention Curves by Acquisition Channel
Curves that flatten signal product-market fit with that segment. Curves that keep declining indicate wrong-fit acquisition.
Key insight: Referral customers are 2.6× more likely to be active at Month 6 than cold outbound. Your cheapest-looking acquisition channel is your most expensive when measured against retention.
At-Risk Account Monitor — Intervention Required
Acme Manufacturing
Usage −82% MoM
BrightPath Solutions
Usage −61% MoM
Veritas Group
3 open tickets, no response
NorthStar Retail
Login frequency declining
Cascade Digital
New team, not yet onboarded
Revenue Impact Model — What Fixing Churn Is Worth
Current State
Target: 5% gross churn
At 5% Gross Churn
Stop guessing which customers are about to leave.
We build the cohort models, NRR decomposition, health scores, and at-risk alerts your CS team needs to intervene 60 days before it's too late.