After 15 years and around 120 startups, we've reviewed a lot of churn data, which lines up with the broader state of product UX. The pattern is that founders who can't move their churn rate are almost always doing everything right on paper — shipping features, redoing onboarding, cleaning up flows — and nothing changes. That's usually because they're treating churn as one problem when it's two: voluntary and involuntary. Each has a different root cause and a different fix. And then there's activation failure, which isn't technically churn but makes both harder to fix.

Involuntary churn has nothing to do with your product

This is when a subscription lapses because a payment failed, not because someone chose to leave. It accounts for more of your churn than you'd expect, and most of it stays invisible until you pull your data and separate failed payments from real cancellations. The fixes are mechanical, not strategic. Add a reminder flow a few days before billing. When a payment fails, notify by email and in-app, give a grace period, and follow up two or three times so the user actually has a chance to act. Let users save multiple payment methods as a backup, and set up automatic card updaters that refresh expired card details in the background. None of this needs a redesign — it just needs to exist.

Voluntary churn needs a diagnosis before a fix

This is a deliberate decision to cancel — a bug, a missing feature, pricing that didn't feel worth it, or a better option elsewhere. Different users leave for different reasons, so there's no single fix. Start with an exit form when someone cancels or downgrades: capture the reason with specific categories ("no longer needed," "too expensive," "a bug," "switched to a competitor"). The more specific the options, the better your data gets over time. But exit forms only go so far — people click an answer in 30 seconds without much thought. Getting churned users on an actual call is far more valuable when you can manage it; a $25–$50 gift card is usually enough of an incentive, and the conversation gives you the real reasoning.

What you fix depends entirely on what the data tells you. UX friction? Prototype and test a fix with real users before spending dev resources. A feature you don't have? Weigh it against your roadmap — not every request is worth building. A feature users can't find? Update the flow that surfaces it, not the feature itself. Lost to a competitor? Map what they offer that you don't, and decide which gaps are actually worth closing.

The activation failure underneath

Activation failure is the related problem that doesn't always show up in your churn numbers: users who sign up, never get activated, and never come back. Free users who never see why they'd upgrade. Trial users who don't return after the first session. A paying customer has to get activated first, so every user lost at that step is a customer you'll never get. Diagnose it with product analytics running inside the tool — Hotjar, Microsoft Clarity, PostHog, Mixpanel — to see where people struggle before they leave. Then map what your retained users actually did early on, compare it against churned users, and find where the two paths split. That divergence is where the friction lives. From there, redesign onboarding to guide new users down the same path your best users took — account setup and creating their first asset is usually the fastest route to value.

If you're working through a churn problem and want help running this kind of diagnosis, reach out — we'll figure out which type is behind your churn and what to fix first.