Customer churn is five different problems that look identical in your dashboard. Sort your last twenty cancellations before you change anything.
A churn reduction framework is a diagnosis before it is a fix. Customer churn is not one problem with one solution — it is five different problems that look identical in your dashboard and respond to opposite treatments.
Diagnose first. The tactics that fix onboarding churn actively worsen price churn.
The most common mistake is applying a fix before naming the type — usually a discount, which converts a price problem into a margin problem and does nothing for the other four.
You need two pieces of data, both of which you already have. When each churned customer cancelled, measured from signup. And whether they were active in the two weeks before they left.
Those two numbers separate the five types cleanly.
| Cancelled at | Active before leaving? | Type | Fix |
|---|---|---|---|
| Day 0–60 | No | Onboarding churn | Activation moment |
| Day 60+ | No | Value churn | Recurring use case, not a one-off job |
| Any | Yes | Price churn | Value metric or tier structure |
| Any | No cancellation event | Involuntary churn | Dunning and card retries |
| Day 0–90 | Briefly, then nothing | Fit churn | Positioning |
Sort your last twenty cancellations into those rows. One type usually accounts for most of them, and that is the only one worth working on this quarter.
They signed up, never reached the activation moment, and cancelled within the first two months. In most early-stage products this is the largest single category, and it is frequently misread as a pricing problem because the customer says "it wasn't worth it".
The fix is upstream of churn entirely. Nothing you do at cancellation time recovers a customer who never saw the product work.
They activated, used the product properly, and then stopped needing it. This is the honest one, and it is a product structure issue rather than a retention issue.
Products that solve a one-time job churn by design. Migration tools, launch checklists, setup utilities. If your customers finish, no retention tactic will hold them — the question becomes whether there is a recurring use case adjacent to the one-off one.
They were active right up to the day they cancelled. The product was working. They decided it was not worth the money.
This is rarely solved by lowering the price. Active users who leave over cost are usually telling you the value metric is wrong — they are paying for something that does not scale with what they get out of it. The value metric framework covers the reset, and the pricing framework covers the method.
Discounting here trains customers to threaten cancellation and cuts margin on the customers who were happy. Consider a cheaper tier with genuinely less in it before considering a discount on the same thing.
The card expired or the payment failed and nobody chased it. The customer did not decide to leave — they were removed by your billing system.
This is the cheapest churn to fix and the most commonly ignored by solo founders, most of whom have never looked at their failed payment rate. Card retry logic, a dunning sequence and an expiry warning are a weekend of work that recovers revenue every month afterwards.
Check this first regardless of what your diagnosis says. It takes ten minutes in Stripe, it requires no product change, and any recovery you find is money you have already earned and simply failed to collect.
They were never the right customer. They came from a channel that attracts the wrong people, or a landing page that promised something adjacent to what you do.
Fit churn is a positioning and acquisition problem wearing a retention costume. Fixing it means changing who arrives, not what happens after they arrive — which is why the retention versus acquisition question matters here.
Context first: monthly churn for small self-serve SaaS is typically higher than the enterprise benchmarks quoted in most articles. Comparing yourself to a 1% enterprise figure is not useful.
| Monthly churn | Average customer lifetime | Reading |
|---|---|---|
| 3% | About 33 months | Strong for self-serve |
| 5% | About 20 months | Workable |
| 8% | About 12 months | You replace your base annually |
| 12%+ | About 8 months | Growth is arithmetically capped |
The number that matters more than the rate is the shape. Churn concentrated in the first sixty days is an onboarding problem you can fix. Churn spread evenly across all cohorts is a value or fit problem that runs deeper.
This page diagnoses. For the tactics that fix each type once you have named it, use the guide to reducing SaaS churn.
If your churn is fine but growth has stopped anyway, the constraint is elsewhere — breaking the MRR plateau covers that diagnosis. And if customers churn before ever paying, the problem is in trial-to-paid conversion rather than retention. If you're building a subscription app rather than browser-based SaaS, reducing churn in a subscription app covers the app-store-specific mechanics that don't apply here.
A churn reduction framework is a diagnostic method that sorts customer churn into types before applying a fix. The five types are onboarding churn, value churn, price churn, involuntary churn, and fit churn. They look identical in a dashboard but respond to opposite treatments.
Five: onboarding churn where the customer never reached value, value churn where they stopped needing the product, price churn where the value is real but not worth the cost, involuntary churn from failed payments, and fit churn where they were never the right customer.
You need two data points per cancelled customer: how many days after signup they cancelled, and whether they were active in the two weeks before leaving. Cancelling early while inactive indicates onboarding churn; cancelling while active indicates price churn.
For small self-serve products, 3% monthly is strong and 5% is workable. At 8% the average customer lasts about a year, meaning you replace your entire base annually just to stay flat. Enterprise benchmarks near 1% are not a useful comparison.
Rarely. Discounting is the right fix for one churn type out of five, and it trains customers to threaten cancellation while cutting margin on happy ones. Active users who leave over cost usually indicate the value metric is wrong rather than the price.
Involuntary churn is revenue lost to failed or expired card payments rather than a customer decision. It is the cheapest churn to fix — card retry logic, a dunning sequence and expiry warnings are a weekend of work — and most solo founders have never checked their failed payment rate.
Bring your last twenty cancellations with dates. Marcus names the type and gives you one fix to run this month.
Try GhostCoach free →14-day free trial · cancel anytime · 30-day money-back on Lifetime