# Why customers churn and how to reduce it > How to reduce churn by its cause: failed payments, customers who never got value, owners who close, and customers who were never a fit, with a different fix for each. - Course: [From Idea to Profitable Product: Building Without Investors](https://productovi.com/courses/bootstrapped-product.md) - Module: Numbers every month - Lesson 21/30, ~27 min - Source: https://productovi.com/courses/bootstrapped-product/why-shops-leave ## Why customers churn: three calls, three stories **Alex Tutor:** Back at GarageDesk, the founders ask you and me for advice about a less happy number: some shops have cancelled. **Oscar:** I called the three shops that left in October. Three different stories. **Oscar:** The first one's card expired and nobody noticed. The second never loaded its client list, so no reminder ever went out. The third moved to a cheap general CRM. **Sofia:** So one churn number, three different problems. **Alex Tutor:** Exactly why the names matter. Churn is a sum of very different reasons, and you reduce it one reason at a time, each with its own fix. Let's sort them before anyone starts fixing. **Nina:** I went back through everyone who left since the March launch: eleven shops. Four failed payments, four never loaded their clients, one owner sold his shop, two went to a general CRM. > **Exercise:** Match each departed shop's story to the kind of churn it is. ## Involuntary churn: leaving by accident **Alex Tutor:** Start with the cheapest fix. A shop whose card fails didn't decide to leave. That's **involuntary churn**, and in small subscriptions it's often a big share of all churn. ![A cracked bank card slips out of a wallet onto a garage floor while the owner, busy under a car, doesn't notice.](https://productovi.com/courses/bootstrapped-product/inline/why-shops-leave-illustration-2.webp) **Nina:** Four of eleven here. The payment provider retries twice and then just gives up. **Alex Tutor:** So you add the human part. A notice a week before the card expires. A Viber message to the owner with a payment link, not an email to an inbox nobody reads. Then a call. And never delete a shop's data on day one. **Alex Tutor:** Which of those messages actually reaches a shop owner? Compare two. > **Exercise:** A shop's payment has failed after the provider's retries. Which message should GarageDesk send? **Oscar:** The October shop would have paid. The owner just didn't know. > **Exercise:** Put GarageDesk's steps for rescuing a failed payment in order. ## Customers who never reach the aha moment **Sofia:** And the four shops that never loaded their clients? **Alex Tutor:** They paid for a promise and never saw it kept. What's the moment a shop owner thinks “this works”? **Oscar:** When a regular he'd forgotten about drives in and says “I got your reminder”. That's the day. **Alex Tutor:** That's your **aha moment**. Everything before it is a cost for the shop: typing in clients, entering mileage. A shop that doesn't reach it within a month or two leaves. ![A repair-shop owner looks up delighted as a familiar regular drives back into his garage holding up a phone with a bell icon.](https://productovi.com/courses/bootstrapped-product/inline/why-shops-leave-illustration-3.webp) *Everything before this moment is a cost for the shop; shorten the road to it.* **Nina:** The spreadsheet import only helps shops that keep a spreadsheet. Most keep a notebook, and nobody wants to type in five years of it. **Alex Tutor:** So start small. On the first call, enter only the cars due for service in the next two months. Twenty cars, and the first reminders go out that week. **Nina:** And I'll add a flag for any shop with no reminder sent after two weeks. > **Exercise:** Which moment is the best candidate for GarageDesk's aha moment, the one to get every new shop to as fast as possible? ## Poor fit: not every departure is a failure **Sofia:** And the owner who sold his shop? **Alex Tutor:** Before I answer, what would you do with that one? > **Exercise:** One shop left because its owner sold the business. What should GarageDesk do? **Alex Tutor:** Nothing to fix in the product. But the shop still exists: call the new owner. It's the warmest lead you'll get this month. **Oscar:** The two that went to a general CRM wanted a warehouse and invoicing. Should we build that? **Alex Tutor:** Ask first whether they're your niche. You chose shops that live on repeat service. A shop that mostly needs parts accounting was a poor fit from day one. Chasing it turns GarageDesk into a worse copy of the CRM. **Nina:** So we learn from them, and fix who we sell to, not the product. ![Of 11 departures, 4 were failed payments, 4 never loaded clients, 2 went to a general CRM and 1 shop was sold; each reason has its own fix, failed payments the cheapest.](https://productovi.com/courses/bootstrapped-product/diagrams/m6-l2-churn-reasons.png) *Sort churn by reason before fixing anything.* > **Exercise:** Two of eleven departed shops moved to a general CRM because they needed parts and invoicing. **True or false:** the right response is to add a warehouse and invoicing module to win shops like them back. ## Churn warning signs before the cancel button **Alex Tutor:** Most shops go quiet long before they cancel. If you see the silence, you still have time for a call. ![A row of small garages at dusk; one has dimmed lights and an empty forecourt, and a founder calls it on the phone from across the street.](https://productovi.com/courses/bootstrapped-product/inline/why-shops-leave-illustration-4.webp) **Nina:** I can list signals from the data: no reminders sent for 30 days, nobody from the shop logged in for three weeks, a card expiring next month. **Oscar:** And every Monday I call the shops on that list. A call before they decide is worth ten after. **Alex Tutor:** Good. And if October's two avoidable losses, the expired card and the empty client list, had been saved, churn would have been 1 of 58, under 2%, instead of 5.2%. **Alex Tutor:** Check my arithmetic on that. > **Exercise:** GarageDesk started October with 58 shops and lost 3: one whose expired card nobody noticed, one that never loaded its client list, and one that moved to a general CRM because it mainly needed parts accounting. If the first two had both been saved, what would October's churn have been, in %? Round to one decimal. > **Exercise:** Which of these are early warning signs that a GarageDesk shop may leave? Choose all that apply. ## Your turn: a message that prevents churn **Sofia:** Can I write the message for a shop that's stuck before the aha moment? **Alex Tutor:** Go ahead. A new shop in Khmelnytskyi signed up ten days ago and its client list is still empty. Write to the owner in Viber. **Oscar:** Keep it short. Owners read Viber between two cars. > **Exercise:** A shop in Khmelnytskyi signed up for GarageDesk ten days ago, paid, and hasn't added a single client. Write a Viber message (3–5 sentences) to the owner that helps the shop reach its first returning customer. > **Key ideas** > > - Churn is a sum of different reasons. Write down who left and why, then sort them before fixing anything. > - **Involuntary churn** (failed payments) is the cheapest to fix: expiry notices, retries, a message with a payment link, a call, pause instead of delete. > - Shops that never reach the **aha moment** leave. For GarageDesk it's the first regular who comes back after a reminder, so remove the chores before it. > - A sold or closed business isn't a product failure; call the new owner. > - Customers outside the niche are a lesson about whom you sell to, not a feature request. > - Watch the silence: no reminders, no logins, an expiring card. **Alex Tutor:** Next time: the founders want to charge new shops more. How do you raise a price without breaking trust? - Previous lesson: [MRR calculation and monthly churn](https://productovi.com/courses/bootstrapped-product/mrr-and-churn.md) - Next lesson: [Raising prices without losing customers](https://productovi.com/courses/bootstrapped-product/raising-the-price.md) --- Productovi Learning. Free, no sign-up needed to read.