I Built an App With AI. Now What? Your First 30 Days, In Order

You built an app with AI, it works, and now you are asking the question that comes right after shipping: I built an app with AI, now what? For the next 30 days, stop adding features and work through four steps in order. First, check that the app is safe to show people and that you can see who is using it. Second, talk to real people who have the problem it solves. Third, test whether they want it and will pay for it. Fourth, pick one channel to reach more people like them, and at the end of the month decide what comes next.

Shipping used to be the finish line. Now it is day one. Building got fast, but finding customers, setting a price, and getting in front of the right people still move at the same speed they always did, and that gap is why the dashboard feels so quiet right now.

Below is a week by week plan with the tools to use and a clear signal for each week that tells you when you are done and can move on.

What should I do in the first 30 days after I built an app with AI?

The table below lays out the first 30 days after you build an app with AI, week by week, with the question each week answers and our rule of thumb for when you have enough of an answer to move on. The tools are cheap or free.

Moving on means you have an answer to that week's question, even a rough one. It does not mean you finished building anything.

Why does launching feel so much quieter than building?

While you were building, every evening produced something you could see. A screen came together, a bug went away, a feature shipped. Launching takes that feedback loop away. You post the link, a few friends sign up, and then the numbers sit still.

That silence means you have no data yet. It says very little about whether the idea is any good, because most of the people who would want your app do not know it exists. Nobody has told them.

The business side of the work runs at human speed. Conversations need to be scheduled, people take days to reply, and trust builds one exchange at a time. Thirty days is enough to get real answers, as long as you spend those days on the right questions in the right order.

How do I check my app is safe to show people?

Before you invite strangers in, spend part of week 1 making sure nothing embarrassing or harmful happens when they arrive. This is a short checklist, and most of it takes minutes.

• Logins. Open a private browser window and sign up with a fresh email. Log out, log back in, and reset your password. Every step should work without you touching the database.
• Who can see whose data. Create two test accounts and confirm that account A cannot see or edit anything that belongs to account B, including when you change an ID in the URL. If your app runs on Supabase, confirm that Row Level Security is turned on for every table that holds user data and that the policies limit each row to its owner.
• Secrets. API keys and database passwords should live in server-side environment variables, never in front-end code or a public repository. If a key ever ended up in a commit, replace it with a new one.
• Emails the app sends. Trigger the welcome email, the password reset, and any notifications. Check that they arrive, that they stay out of spam, and that they come from your own domain.
• Backups. Find out where your data lives, confirm that backups are switched on, and learn how you would restore one before you ever need to.
• Errors. Install Sentry, which has a free plan, so you hear about crashes from a report instead of from an annoyed user.

If you find items on this list you cannot answer, that is useful to learn now. Some founders fix everything in an afternoon. Others decide the app needs a professional review before real users touch it. If you go that way, it helps to know how to vet an AI development company before you pay for one and what custom development work actually costs.

How do I track whether anyone is really using my app?

Install simple analytics during week 1 so that everything you try later shows up somewhere. PostHog's free tier covers far more than an app needs at this stage and includes session replays, which let you watch how real people move through your screens. Plausible is a simple paid option with a free trial if all you want is clean traffic numbers.

Watch three numbers and ignore the rest for now:

1. Visits. How many people land on the app or its landing page.
2. Signups. How many of them create an account.
3. People who come back. How many return on a later day without you reminding them.

Do not set a target for any of these yet. The numbers will be small, and that is expected. What matters is that you can see them move when you do something, so that in week 4 you can tell whether your channel worked.

Before the week ends, write down one sentence: "This is for [person] who [problem]." For example, "This is for independent physical therapists who lose hours every week rescheduling patients." Your first version will probably be wrong or too broad, and that is fine. Week 2 exists to correct it.

Who are my first 10 users, and how do I get them without ads?

Your first users come from a list you write by hand. Open a spreadsheet and write down 10 specific people by name who have the problem from your sentence. Look through your contacts, former colleagues, clients from past jobs, and the online communities where people complain about this exact problem. Next to each name, note where you found them and why you think the problem applies to them.

Then reach out to each one personally. A short message works best: say why you thought of them, describe the problem in one line, and ask for 20 minutes to hear how they handle it today. Offer to set them up in the app yourself at the end of the call if they are curious.

This is the advice Paul Graham gives in his essay Do Things that Don't Scale. He describes how the Stripe founders handled anyone who agreed to try their product. They would say "Right then, give me your laptop" and set them up on the spot. His summary is blunt: "You can't wait for users to come to you. You have to go out and get them."

Use Calendly or Cal.com so people can book a time without back and forth, and keep a log in Notion or a Google Sheet with the name, date contacted, call date, what they said, and the next step. That log becomes the most valuable document you own this month.

How do I talk to users without getting polite lies?

People are kind. If you describe your idea and ask whether they like it, most will say yes, and that answer tells you nothing. Rob Fitzpatrick's book The Mom Test is built around avoiding this trap, and its core rules fit on an index card. Ask about their life and what they have actually done in the past, not about your idea or what they might do someday. Get specifics instead of opinions. Talk less than they do.

Here is a script you can use on every call this week:

1. "Tell me about the last time you dealt with [problem]."
2. "What did you do about it?"
3. "What did that cost you, in time or money?"
4. "What have you tried or paid for to fix it?"
5. "Can you show me how you handle it today?"
6. "Who else do you know who deals with this?"

Show the app only at the end, and only if the conversation gets there. Write your notes within the hour, while the exact words are still fresh, because the way people describe their problem is the language your landing page should use later. The last question matters more than it looks, since every call can hand you one or two more names for your list.

How do I know if people would actually miss my app?

In week 3, measure how much people would care if the app disappeared. The simplest tool for this is the Sean Ellis survey. Build it in Tally or Typeform and send it to everyone who has actually used the app. It asks one question:

"How would you feel if you could no longer use this product?"

• Very disappointed
• Somewhat disappointed
• Not disappointed

According to First Round Review's account of how Superhuman found product-market fit, Sean Ellis benchmarked nearly a hundred startups and found that products where at least 40% of users answered "very disappointed" were the ones that tended to grow. Superhuman's CEO Rahul Vohra ran the survey and started at 22%. When the team segmented the results and focused on the type of user who loved the product most, the score rose by about ten points. Within about three quarters of work aimed at those users, it reached 58%.

If you land under 40%, that does not mean you should quit. Look closely at the people who said "very disappointed." Find what they have in common, such as their job, their situation, or the specific problem they use the app for. Rewrite your "this is for" sentence around them, and spend the rest of the month on people like them. Narrowing usually feels like giving something up, but it is how a vague product becomes one a specific group can't do without.

When should I start charging for an app I built with AI?

Start in week 3, at least with the people who lit up in your survey and interviews. Paying is a much stronger signal than saying you like something, and you only find out whether someone will pay by giving them the chance.

A Stripe Payment Link is the fastest way to do this. You create a product and a price in your Stripe dashboard, and Stripe gives you a link you can send in a message or put on a page, with no billing code to write. You need an activated Stripe account for it to accept real payments.

Keep this step small. The goal for the month is only to put a real price in front of real people and see what happens. If they pay, you have learned something important. If they hesitate, ask what would make it worth paying for, and listen closely to the answer. Working out the right price, the right plans, and how pricing fits your costs is a bigger piece of work, and it deserves its own look once you know who your buyer is.

How do I get my first customers without a marketing team?

In week 4, pick one channel and commit to it for the whole week. The right channel is wherever the people from your week 2 calls already spend their time, so look back at your notes before you choose. Common options include:

• Founder-led LinkedIn posts, if your users are professionals who read LinkedIn. Write about the problem in their words, using what you heard on calls.
• One niche community, such as a subreddit, Slack group, or forum where the problem gets discussed. Be useful there first, and mention the app when it actually answers someone's question.
• Direct outreach, continuing what you did in week 2 at a steadier pace, with a message refined by everything you learned.
• A single focused landing page, built around your "this is for" sentence and shared wherever your people are. If you go this way, these ideas on getting more leads from the visitors your website already gets apply directly.

Product Hunt and similar launch sites can be one option inside this plan, but they work best once you already know who you are talking to. They should not be the whole plan.

Run your channel every day of the week and measure it in the analytics you set up in week 1. You want to see whether visits, signups, and returning users moved, and whether the people who came in look like the ones you interviewed.

What should I decide at the end of the first month?

On day 30, look at your log, your survey, your pricing test, and your channel results together, and make one of three calls. This is our rule of thumb, meant as a decision aid rather than a benchmark:

• Keep going if a specific group answered "very disappointed" at or near 40%, at least one person paid or committed to pay, and your channel brought in people who signed up and came back. Do more of what worked.
• Narrow if a few people clearly love the app but most are lukewarm. Rewrite your sentence around the people who lit up, cut or pause features they do not use, and run the month again for that group.
• Change if almost nobody would be very disappointed, nobody would pay, and the conversations showed the problem is not painful enough to act on. Change the problem or the audience you are solving for rather than adding more features.

The month is meant to give you clarity, and it can't promise success. Knowing which of the three you are in is a real result, and it is far better than another month of building in silence. The questions you answered this month, who the app is for, whether they will pay, and how to reach them, are the core of a go-to-market strategy for founders, and the next layer of that work is pricing in depth, operations, and first hires.

Common Questions About Building a Business After Vibe Coding

I built an app with AI and nobody is using it. What do I do?

Pause new features for a month and work through the business questions in order. In week one, set up simple analytics and make sure the app is safe to show strangers. In week two, reach out personally to a short list of specific people who have the problem and ask how they deal with it today. In week three, run a one-question survey to see who would miss the app, and put a real price in front of the people who are most interested. In week four, pick one channel where those people spend time and run it for a full week. A quiet dashboard this early usually means nobody has been asked yet, so the fix is to go and find people one at a time.

How do I get my first users for an app I vibe coded?

Write down 10 specific people by name who have the problem your app solves. Look in your own network, among former colleagues, and in the online communities where that problem gets discussed. Message each one personally, say why you thought of them, and offer to set them up yourself on a short call. Ads and launch sites rarely work at this stage because you do not yet know exactly who the app is for or what to say to them. Going one by one teaches you both, and what you learn in those first conversations becomes the way you describe the product.

How do I know if my AI built app has product-market fit?

Look for a specific group of people who keep coming back without being reminded and who would be upset to lose the app. Two checks work well in the first month. Watch how many people return in your analytics, and run the Sean Ellis survey, which asks users how they would feel if they could no longer use the product. Sean Ellis found that products where around 40% of users answer very disappointed tended to grow well. If you are below that, study the people who did answer very disappointed and narrow the product toward people like them.

What is the Sean Ellis survey and how do I run it?

It is a one-question survey created by growth marketer Sean Ellis that asks: How would you feel if you could no longer use this product? The answer options are Very disappointed, Somewhat disappointed, and Not disappointed. Build it in a form tool like Tally or Typeform and send it to people who have actually used the app, ideally more than once. Add a couple of follow-up questions, such as what kind of person would benefit most from the app and what the main benefit is for them. Then read the very disappointed answers closely, because those people describe the customer you should build for.

Should I charge for my vibe coded app right away?

Put a real price in front of your most interested users within the first month, even if the app stays free for everyone else for now. A Stripe Payment Link lets you do this without writing any billing code, as long as your Stripe account is activated. Someone paying, or clearly turning down a specific price, tells you far more than someone saying they like the app. You do not need a finished pricing strategy to run this test, and the first price can change later. Pricing deserves its own careful look once you know who is actually buying.

Do I need a business plan or a go-to-market strategy for an app I built with AI?

You do not need a long business plan in the first month, but you do need the core of a go-to-market strategy: who the app is for, what problem it solves for them, whether they will pay, and one repeatable way to reach more people like them. A focused 30 days of user conversations, a simple survey, a real price, and one channel will give you a first draft of exactly that. Once you have those answers, the next layer is pricing in more depth, operations, and eventually your first hires, which is where a fuller go-to-market strategy becomes worth the time.

If you would rather not do this month alone

Everything in this plan can be done by one founder with a calendar, a spreadsheet, and a few free tools. Plenty of founders do it exactly that way. Others would rather have someone beside them who has worked through the same questions before.

The people behind sagulabs run businesses themselves, and we have sat on this side of the table, with a product that works and no business around it yet. Finding fit, naming the customer, setting a price, building one repeatable channel, and then putting operations and first hires in place is what our go-to-market work with founders covers.

If that sounds like where you are right now, start a conversation with us.