If you track churn by cohort — grouping customers by when they started and watching how many remain over time — a pattern appears with uncomfortable regularity. Month 1 and 2 are fine. Then month 3 arrives and a cluster of customers cancel, go quiet, or simply fail to renew. The cliff is real, and it shows up across industries, product types, and business models.
Understanding why it happens — and specifically what you can do to prevent it — is one of the highest-leverage retention conversations a small team can have. Because by the time a customer churns in month 3, the decision was usually made in month 2. And the root cause was planted in month 1.
The chart below shows the typical engagement curve for a subscription product or recurring service. The month-3 drop is not random — it follows a predictable sequence of psychological and behavioural events.
The timing is not accidental. It reflects the convergence of three things that happen simultaneously around the 10–12 week mark.
When customers first start using a product or service, there's an activation energy — the newness of it, the optimism of having made a decision, the intention to get value. This energy is powerful but temporary. By month 3 it's gone, and the customer is now evaluating the relationship based purely on what it actually delivers week to week.
If the habit of use hasn't formed by this point — if the product hasn't become genuinely embedded in their workflow — the calculation tips toward "is this worth continuing?" And without an emotional hook or a clear recent win, the answer is often no.
Customers start with a goal in mind. By month 3, one of two things has happened: they've made progress toward the goal (and may feel they no longer need you), or they haven't made progress and are quietly disappointed. Either scenario can trigger churn — and neither is visible to you unless you're actively checking in.
Many businesses do informal quarterly expense reviews. Month 3 is exactly when a subscription or retainer first appears in three consecutive months of bank statements — making it the first time a decision-maker might scrutinise it. If that scrutiny isn't met by a clear, recent demonstration of value, the subscription gets cancelled.
The strongest predictor of long-term retention is whether a customer achieves a meaningful result in the first two weeks. Not a complete transformation — a first win. A report generated. A workflow automated. A result measured. Something concrete that makes the value real and tangible before the novelty wears off.
This means your onboarding process needs to be designed around this outcome, not around showing features. Map the shortest possible path from signup to first meaningful result, and actively shepherd every new customer down that path rather than leaving them to explore.
Six weeks in, proactively reach out with something useful — a usage summary, a tip based on how they've been using the product, a case study from a customer in a similar situation. The purpose is twofold: it demonstrates that you're paying attention, and it surfaces any concerns before they become decisions.
This touchpoint is often where you learn that a customer is struggling with something they haven't told you about. A simple "how's it going?" with genuine curiosity gets information you can act on. Waiting for them to raise it first means you'll often hear about it in a cancellation notice.
Define what low engagement looks like for your product or service — login frequency below a threshold, features going unused, reports not being opened, invoices going unpaid. Then build a trigger: when those signals appear, someone on your team gets a task to reach out personally.
This is not automated marketing. It's a human reaching out because they've noticed something. "I can see you haven't logged in in a couple of weeks — is there anything I can help with?" is a message that gets replies. An automated "We miss you!" email gets ignored.