# How to build an MVP: sign up, get value, pay > How to build an MVP for a subscription product: the three things a first version needs, how much can stay manual behind the scenes, and when a manual step has to be automated. - Course: [From Idea to Profitable Product: Building Without Investors](https://productovi.com/courses/bootstrapped-product.md) - Module: The smallest thing that works - Lesson 14/30, ~26 min - Source: https://productovi.com/courses/bootstrapped-product/mvp-sign-up-get-value-pay ## The MVP feature list: four weeks to 3 March **Alex Tutor:** Back to GarageDesk, whose founders you and I advise: launch is on 3 March, with the five shops that prepaid. Today's question is how to build an MVP in time. Nina, what's on your list? **Nina:** Sign-up with email confirmation, password reset, a dashboard with charts, booking, repair history, a billing page with plan changes. And the reminders. **Sofia:** That's a lot for four weeks. **Alex Tutor:** It's a list for a year. An **MVP** needs three things: the customer can **get in**, **gets the result** they paid for, and **pays**. Everything else waits. ![Three stepping stones across a puddle in front of a garage hold a key, a car driving back into the bay and a coin.](https://productovi.com/courses/bootstrapped-product/inline/mvp-sign-up-get-value-pay-illustration-2.webp) *Get in, get the result, pay: anything else can wait.* **Alex Tutor:** "Minimum" means you can't cut more without losing the result. "Viable" means the result is real: a car owner actually gets the reminder and comes back. > **Exercise:** Which items belong in GarageDesk's first version for the five shops? Choose all that apply. ## Manual behind the scenes: a concierge MVP **Sofia:** But the shops still have to sign up somehow. And their client lists have to get in. **Alex Tutor:** Not necessarily through code. There are five shops. Nina can create their accounts by hand and send each owner a link. **Nina:** And the lists? Dennis's niece has a notebook with hundreds of regulars. **Alex Tutor:** Type them in yourself. It's slow, but you'll see every odd phone number and every car with no mileage. That's a **concierge MVP**: the shop sees a working service, and part of the work behind it is yours. ![A tidy shop front with one simple button, while behind a half-drawn curtain the founder types client lists by hand from a stack of notebooks.](https://productovi.com/courses/bootstrapped-product/inline/mvp-sign-up-get-value-pay-illustration-3.webp) **Oscar:** The shops will ask what happens to those lists. In January one owner said no for exactly that reason. **Alex Tutor:** Then answer in writing before launch: the client list stays the shop's. Only the shop sees it, its car owners get nothing but that shop's reminders under its name, and the shop can download or delete the list at any time. **Alex Tutor:** Two possible replies to that worried owner. Pick the one you'd send. > **Exercise:** A shop owner asks what happens to the client list Nina will type in. Which reply should Oscar send before launch? **Oscar:** Payment works the same way: an invoice or a payment link from a ready-made provider. The five have paid for three or twelve months anyway. > **Exercise:** Match each part of GarageDesk with how it works on launch day. ## What can't be manual: automate the core **Sofia:** Then why not send the reminders by hand too? Nina could write them in the evenings. **Alex Tutor:** Because that's the promise itself. A reminder that arrives late, or not at all, breaks the result the shops paid for. Keep the edges manual, automate the core. ![The MVP has three parts: getting in and paying stay manual behind the scenes for five shops, while the reminders, the promise itself, are automated.](https://productovi.com/courses/bootstrapped-product/diagrams/m4-l2-mvp-core.png) *Keep the edges manual, automate the core.* **Nina:** And if I send them by hand, I learn nothing about whether the automatic version works. **Alex Tutor:** Right. Manual work is fine while it teaches you something and fits into your week. When it stops doing either, it goes to the top of the list for automation. > **Exercise:** Why do the reminders have to be automatic from day one, while accounts and client lists can be manual? ## When manual gets too slow **Nina:** Typing in one shop's list takes me about three hours. For the first five, that's 15 hours. Fine. **Alex Tutor:** Now count April. Oscar, how many new shops do you expect? **Oscar:** About nine, if the shops that are asking around actually sign. That would take us from 5 to 14. **Alex Tutor:** Then set the rule in advance: when a manual task takes more than half a day a week, about four hours, it's next in line to automate. **Alex Tutor:** Before April, check the launch month against that rule. Nina spread the first five lists over four weeks. > **Exercise:** Nina spent 15 hours typing in the five launch shops' lists, spread over the four weeks before 3 March. How many hours a week is that? > **Exercise:** Typing in one shop's client list takes Nina about 3 hours. If 9 new shops join in April, how many hours will she spend typing in lists that month? ## What waits: features outside the MVP **Sofia:** And booking? Eleven of fourteen shops complained about phone chaos. **Alex Tutor:** Booking solves a different problem. Starting two at once doubles the work and halves the attention. The five paid for reminders. **Nina:** So booking, repair history, charts, plan changes and my old idea of an app for car owners all go on a "later" list. **Alex Tutor:** Keep that list, and note who asked for each item and how often. Early users ask for lots of things; the list shows which requests repeat. **Oscar:** And the five know it's rough. They're **early adopters**: they forgive gaps as long as they see quick progress. **Alex Tutor:** Nina is starting that "later" list today. What should each line on it say? > **Exercise:** What should Nina note next to each item on the "later" list? > **Exercise:** **True or false:** booking should go into the first version, because 11 of the 14 shops Nina and Oscar checked in December mentioned phone chaos. ## Launch day **Alex Tutor:** 3 March. Five shops, accounts made by hand, lists typed in by Nina, reminders going out on their own. **Nina:** Honestly, it's a bit embarrassing. No charts, no settings page. **Alex Tutor:** Embarrassing is fine. The only question that matters this month: does a car come back? **Oscar:** Two days later Dennis calls me: a regular replied to the reminder, "Thanks, I'll come on Friday." His niece booked him in. **Sofia:** And Dennis wants more already? **Oscar:** Of course. He wants a report: how many cars came back this month thanks to GarageDesk. > **Exercise:** Dennis, who owns one of the five repair shops on GarageDesk since its launch on 3 March, asks for a monthly report of how many cars came back thanks to the reminders. The report isn't in the first version. In 3–5 sentences, write Nina's reply: keep the MVP scope, but don't lose Dennis. > **Key ideas** > > - An **MVP** for a subscription product needs three things: the customer can **get in**, **gets the result** they paid for, and **pays**. > - **Minimum**: nothing more can be cut without losing the result. **Viable**: the result is real. > - Around the core, work can stay manual (**concierge MVP**): accounts created by hand, lists typed in by the founder, payment by invoice or link. The core promise, the reminder, is automated from day one. > - Answer data questions before launch: the client list stays the shop's, and the shop can download or delete it. > - Set a limit in advance: when a manual task takes more than about four hours a week, automate it next. Nina: 9 × 3 = 27 hours in April → import from a file in May. > - Keep a "later" list with who asked and how often. **Alex Tutor:** Next time: SMS, payments, sign-in, hosting. Which parts should a two-person team rent instead of writing themselves? - Previous lesson: [Problem-solution fit: a solution is not a product](https://productovi.com/courses/bootstrapped-product/a-solution-is-not-a-product.md) - Next lesson: [Build vs buy: ready-made instead of your own](https://productovi.com/courses/bootstrapped-product/ready-made-instead-of-your-own.md) --- Productovi Learning. Free, no sign-up needed to read.