How to turn a side project into a real business

A side project becomes a real business when a stranger pays for it, and then another one does. Getting there from a working project takes three jobs the building never touched: the admin that makes money legal to receive, finding out what people will actually pay for, and growing a customer base past the first happy user. If you built the project with a vibe coding tool like Lovable or Base44, the good news is that the product job, the one that used to stop everyone, is the one you already did. This guide walks the remaining jobs in order, including the parts that are slow.

What separates a side project from a business?

Payment, repeated. Not stars, not signups, not a waitlist, not friends saying they would definitely use it. A business is the loop where someone pays, the thing gets delivered, and enough of them come back or send others that the loop keeps turning. Everything in this guide serves that loop. The full framework behind it is our four-jobs guide, written for people starting from zero: starting a business with no experience: the four jobs. A side builder starts from job 2 already done, which changes the order of the work but none of the jobs.

You built the product first. What does that change?

Two real advantages and one real risk. The advantages: you have a demo instead of a pitch, which makes every conversation about the thing rather than the idea of the thing, and you have proof you finish what you start, which most people never get to show. The risk: you built before asking. A project built for yourself has exactly one confirmed user, and the feature list grew from your taste, not from anyone's wallet. That is not a reason to throw it away. It is a reason to treat the next step, finding out what people pay for, as genuinely open, including the possibility that what they pay for is a smaller piece of what you built than you hoped.

The other thing building first buys you is the possibility of the accidental customer, and it is real. One builder described his own version of this moment:

I had an application I had written for my own personal use. I was talking to a guy who asked how I did something. I told him I had written this program.

The reply, in the builder's own telling:

I'll pay you $200 for a copy of that.

From a thread on Hacker News. Notice what happened there: the product existed, but the business started at the conversation.

What is the admin that makes it real?

The unglamorous afternoon, and it is genuinely an afternoon to a few days, not a project. Somewhere for the money to land: depending on your country, receiving repeated payments means registering something, often a simple sole-operator status that is cheap or free and done online. The reality behind the checkout button: whichever tool renders it, the payment provider underneath will ask you to verify who you are and, past certain thresholds, what legal entity is selling. Terms, a privacy page and a refund policy: templates exist, the decisions inside them are yours. A domain that is yours. None of this finds you a customer, which is why it should take days and not become the place you hide. The actual costs of this stage are small and we itemized them line by line: do you need money to start a business?

How do you find out if anyone will pay?

By asking real people for real money, earlier than feels polite. This is the job no tool does, ours included, and the job most side projects quietly skip by shipping features instead. The mechanics are unglamorous: name the person your project helps, find ten of them you can already reach, in your network, in the communities where they complain about the problem, and put a price in front of them. A price, not a survey. "Would you use this" produces compliments; "it costs this much, want in" produces information. Expect weeks to months, expect most asks to land nowhere, and expect the offer to change at least once under contact with real customers. That is the job working. We went through it ourselves this summer and published the record, including the slow parts: we built a real business in 7 days with baget

How long can it take when it is going fine? One founder, posting the week he finally crossed the line:

It took me 7 months of different ideas, marketing methods, product changes, and working my ass off just to get my first paying customer.

From a thread on Hacker News. Seven months, while working at it, is inside the normal range, not a verdict.

When do you charge, and what?

From the first real user, and one clear offer at one clear price. Charging early feels like it will scare everyone away; what it actually does is sort the polite from the real, which is the entire point of this stage. Package what the project does as an outcome someone buys, a subscription, a one-time purchase, a service delivered with the tool, rather than a feature list. Then let the price be a test you adjust, not a verdict on your worth. We are not quoting figures, here or anywhere: what works depends on what it is and who it serves, and this audience is sold made-up numbers constantly.

What about the day job, or the rest of your life?

Keep it, and design around it. A side business run on evenings, weekends or school hours is not a lesser version of a business; it is the version where consistency matters more than intensity, because the weeks compound while the heroic sprints do not. The trap to refuse in advance: quitting on the strength of a waitlist. Let the paying loop, not the enthusiasm around it, tell you when the project deserves more of your week. For the honest arithmetic of small weekly hours, we did the math here: how much time does it take to start a business?

The business around your project can exist this week; the customers still take the asking. Describe the offer in one sentence and see. It costs nothing to look.