BuildInSeven
← All articles

What to Do the Week Your App Goes Live

7 min read

Your app just launched. Here's a practical day-by-day plan for onboarding first users, catching bugs early, collecting feedback, and deciding what to build…

When your app goes live for the first time, the most useful thing you can do in the first seven days is resist the urge to add features. Instead, focus on getting a small number of real users through the product, watching what happens, and fixing what breaks. Everything else can wait.

Why the First Week Matters More Than the First Month

Most founders spend months building, then treat launch day as the finish line. The first week after launch is actually when you find out whether the assumptions behind the product hold up. Bugs surface. Users get confused in places you never expected. Features you thought were essential get ignored.

That feedback loop, compressed into seven days with real users, is worth more than any amount of internal testing. The goal is not to get a thousand sign-ups. The goal is to get ten people through the product and understand exactly what happened.

Day 1: Confirm the Basics Work

Before you invite anyone, run through the product yourself as if you are a brand-new user.

  • Sign up from scratch using a new email address
  • Complete every core action the product is supposed to support
  • Try it on mobile if your users will use mobile
  • Check that confirmation emails, notifications, and any payment flows actually fire

If you built with an agency or AI dev service, this is also the day to confirm you have access to everything you need: the codebase, the hosting dashboard, the database, and any third-party service credentials. If ownership or access is unclear, review what was agreed in your contract. The article Who Owns the Code From an AI Dev Service? covers what founders should expect to receive and control.

Make a short list of anything that feels broken or confusing. Send it to your developer before you invite a single external user.

Day 2: Set Up the Simplest Possible Feedback Channel

You do not need a sophisticated analytics stack on day two. You need a way to hear from users directly.

A single form linked from inside the app, a dedicated email address, or a short Typeform asking three questions is enough. The questions that matter most at this stage:

  1. What were you trying to do when you got stuck?
  2. What did you expect to happen that didn't?
  3. Is there anything you wanted to do that you couldn't find?

Keep the form short enough that someone will fill it in right after hitting a problem, not after thinking about it for three days.

Also set up error logging if your developer hasn't already. Tools like Sentry (for web apps) or equivalent mobile crash reporters will show you what is breaking in the background without users needing to report it. Ask your developer to confirm this is in place.

Days 3 and 4: Onboard Your First Five to Ten Users

Do not post on Product Hunt. Do not send a mass email. Start with five to ten people you know personally or have already been talking to, people who actually have the problem your app solves.

Onboard them manually. Walk each one through the product over a video call or in person if you can. This is not a demo. You want to watch them use the product while narrating their thoughts. Where do they hesitate? Where do they click the wrong thing? What do they say when they get stuck?

Manual onboarding at this stage is not inefficient. It is the fastest way to find out what your automated onboarding will need to say. You cannot write good onboarding copy until you have watched real people get confused.

The table below shows how to categorise what you observe:

ObservationWhat it usually meansWhat to do
User can't find a core featureNavigation or labelling problemFix copy or move the element
User completes the action but says it felt slowPerformance issueLog it and escalate to developer
User skips a step you expected them to takeStep may be unnecessary, or unclearTest removing it or rewriting the prompt
User asks for a feature that doesn't existPossible roadmap itemLog it, don't build it yet
User completes the core flow without helpGood signalNote exactly how they described the product

Day 5: Triage What You've Found

By day five you will have a list of issues. Not all of them deserve immediate attention. Sort them into three buckets:

Blockers. Bugs or confusion points that prevent users from completing the core action the app exists to support. These get fixed before you grow the user base.

Friction points. Things that slow users down or cause confusion but don't stop them entirely. Queue these for the next development cycle.

Nice-to-haves. Feature requests and preferences that are genuinely interesting but not critical. Keep a log. Do not act on them yet.

If you worked with a developer on a fixed-price contract, understand what your agreement covers before sending a list of changes. Some fixes will fall under a bug-fix period; others may be out of scope. The articles Fixed Price vs Hourly: Which Is Safer for Founders? and What Is Included in a Software Dev Quote? explain what is typically covered and what counts as new work.

Day 6: Fix Blockers, Document Everything Else

Send your developer the blocker list with specific reproduction steps. For each bug: what the user did, what they expected, what actually happened, and whether you can reproduce it consistently.

For everything that is not a blocker, keep a running document rather than a scattered set of messages. Group issues by area of the app. Include screenshots where relevant. This document becomes your brief for the next build cycle.

Avoid the temptation to expand scope based on day-one feedback. Every feature request from your first ten users sounds urgent. Most of them reflect that person's specific workflow, not a universal need. Wait until you have heard the same request from at least three or four independent users before treating it as a priority.

Day 7: Decide What Comes Next

At the end of the first week, you have real data. Use it to answer three questions:

Is the core flow working? Can users complete the main action without hand-holding? If not, that gets fixed before anything else.

Is there demand signal? Did any of your first users come back on their own, share the product, or ask when a specific feature would be ready? Organic re-engagement, even from five people, is a meaningful signal.

What is the one thing that would make this product significantly more useful to the people who already like it? Not the list of forty requests. The one thing that keeps coming up.

That one thing becomes the brief for your next development cycle.

What Not to Do in Week One

A few patterns trip up first-time founders in this window:

Chasing large user numbers too early. Ten engaged users who tell you exactly what is wrong are more valuable than five hundred who sign up and never return. Growth comes after the core experience is solid.

Building in real time. The feedback will create urgency. Resist it. Log everything, fix blockers, and plan the next cycle properly rather than patching things reactively.

Rewriting the whole product. One week of feedback does not mean the original concept was wrong. Separate signal from noise before making structural decisions.

Ignoring quiet users. The people who say nothing and never come back are as informative as the ones who send detailed feedback. If you can, reach out and ask what happened.

Setting Up for the Next Cycle

By the end of week one, your job is to have a working product that ten real users have touched, a clear list of what broke, a short brief for what to build next, and a feedback channel that will keep running.

That is a foundation. Supporting articles in this cluster go deeper on specific parts of this process: collecting structured feedback, writing a development brief, and deciding when incremental improvements are worth another build cycle versus when you need something more significant.

Frequently asked questions

How many users should I aim for in the first week after launch?
Five to ten is enough for week one. The goal is depth, not volume. You want to watch real people use the product and understand where they get confused, not accumulate sign-ups that tell you nothing about behaviour.
What counts as a bug versus a feature request?
A bug is anything that prevents the app from doing what it was built to do: a broken flow, a missing confirmation, a crash, or data that doesn't save. A feature request is something the app was never designed to do. The distinction matters because bug fixes are often covered under a development agreement, while feature requests usually are not.
Do I need an analytics tool before I launch?
Basic error logging and a way to contact users directly are the minimum. A full analytics stack is useful but not urgent in week one. Manual observation of real users tells you more than dashboards at this stage.
What should I send my developer after the first week?
A short, specific list of blockers with reproduction steps, followed by a separate document of lower-priority issues grouped by area of the app. Clear reproduction steps (what the user did, what was expected, what happened) make fixes faster and cheaper.
How do I know if the product is working well enough to start growing the user base?
The clearest signal is whether users who are not friends or family complete the core flow without your help and come back on their own. If both of those are happening, even with a small number of users, the product is ready for broader outreach.