Most SaaS founders learn they have a churn problem the same way: they look at one month's numbers, panic, then go back to ignoring it for another quarter. By that point the problem has compounded into something much harder to fix.
Here's the thing: your monthly churn rate is the worst possible early warning system. By the time it spikes, you're already 6 months into a problem. The signals that actually matter show up earlier and are more specific.
Want to skip the diagnosis and get a fast answer? The Churn Health Check scores your retention setup in 60 seconds and tells you if you have a problem and where it lives.
The 8 warning signs (in priority order)
1. Net Revenue Retention is below 100%
This is the single most important number in SaaS. NRR measures how much revenue you keep from existing customers including expansions, contractions, and churn. Below 100% means your customer base is shrinking. Above 110% means it's growing on its own.
Track this monthly. It's more important than monthly churn rate because it accounts for downgrades, upgrades, and the actual revenue impact, not just customer count.
2. Involuntary churn is more than 25% of total churn
If a quarter or more of your churn is failed payments, you don't have a product problem. You have a billing problem disguised as one. The fix is fast and high-ROI: AI dunning recovers 30-50% of failed payments within weeks of implementation.
Pull the data. Most teams are shocked when they see how much of their "churn" is actually involuntary.
3. Activation rate is dropping
Activation rate (% of new signups who reach your aha moment) is the leading indicator of next quarter's retention. If activation is dropping today, your churn rate is going to spike in 30-60 days when those unactivated users decide they don't need you.
If you don't even know your activation rate, that's a bigger problem. Define your aha moment first using the activation audit.
4. First-90-day churn is rising
New customer cohorts churning faster than old ones is a marketing/positioning/onboarding problem. You're acquiring customers who don't fit. This is sneaky because your blended churn rate looks fine while the underlying business is getting worse.
Track each monthly cohort separately. Compare the first 90 days of cohort N to cohort N-3. If the line is going down, you have a problem brewing.
5. Expansion revenue is flatlining
Even if churn is fine, lack of expansion is a future churn problem. Customers who don't expand are customers who don't deepen their dependence on your product. The next budget cut, leadership change, or competitor pitch and they're gone.
Healthy SaaS sees 15-30% of customer base expanding annually. If yours is below 10%, your customers aren't growing into your product, which means you're a tool not a system.
6. Support tickets per account are increasing
Rising support ticket volume per active account is one of the strongest leading indicators of churn. It means users are struggling with things that used to work, or hitting edges they didn't hit before. Either way, frustration is building.
Track this monthly. A 30%+ increase in tickets per account over 90 days predicts a 2-3x increase in churn over the following 90 days.
7. NPS is dropping (especially in long-tenure customers)
NPS isn't a perfect metric, but its trajectory is meaningful. NPS dropping in customers who've been with you 12+ months means you're losing your most valuable advocates. These are the people who refer others. Their decline kills your acquisition flywheel before it kills your retention.
8. Time-to-value is increasing
If new customers are taking longer to reach their aha moment than they did 6 months ago, your product complexity is outrunning your onboarding. This is especially common after major feature releases or pricing changes.
Time-to-value drives both activation and long-term retention. AI-personalized onboarding is the highest-impact lever to fix it.
The 60-second diagnosis
Tracking all 8 signals manually is a project. You probably won't do it. The Churn Health Check is a faster shortcut: 8 questions that cover the same ground, output a score and tier, and tell you which signals are most likely problems for your specific situation.
Take it now and find out where you stand. If you score below 50, you almost certainly have a churn problem hiding somewhere. If you score above 70, you're in better shape than most SaaS companies, and the score will tell you what to optimize next.
What to do once you know you have a problem
The right next step depends on which signal triggered the alarm:
- If NRR is the problem: Build expansion. Product-led expansion is usually the highest-leverage move.
- If involuntary churn is the problem: Implement smart dunning. This is the fastest, highest-ROI fix.
- If activation is the problem: Run an activation audit and define your aha moment with milestones.
- If first-90-day cohorts are getting worse: The acquisition channel is broken or the onboarding is broken. Probably both.
- If support tickets are spiking: Your product complexity has outrun your education. Triage docs, in-app help, and onboarding before adding new features.
For the full picture of where to start, see our guide on how to use AI to reduce churn, or take the Health Check for a personalized starting point.