How to Launch an App: Your First Week, Day by Day

To launch an app, tell a small group of people who have the problem first, watch whether they finish the one thing the app is for, ask them what got in the way, and fix the most damaging problem each day. Widen the circle only after the last one is fixed. The first week is for learning, not for volume.

A lot of launch advice is about the announcement. This guide is about the 7 days after it, when a few real people meet your app and show you things you could not see while building. It is for web apps that people open from a link. If your domain, HTTPS, and email are not set up yet, start with the website launch checklist. We pick up where it ends, with a simple app launch plan for the days that follow.

What does launching mean in week 1?

It means real people, outside your head, using the app for the thing it promises. It does not mean a post that goes everywhere. A small, quiet launch is easier to learn from, because you can tell who said what and fix a problem before the next group meets it.

Decide one sentence before you start: "A launch week success is ___." Something you can observe, such as "6 parents finish setting up a pickup group without asking me for help." Not a visitor count, which flatters you and teaches nothing.

What belongs on your app launch checklist for day 1, before you tell anyone?

Look at the live link the way a customer will. Open it in a private window, on your own phone, and click your main path from the first screen to the finished action. Then get a view that is not yours.

Our audit does that for the public side: it clicks through public pages and gives a score out of 100 across six areas (Clarity and Promise; Craft and Design Integrity; Trust and Legitimacy; Flow and Usability; Mobile and Access; and Conversion and The Ask), with the fixes ranked. 2 parts matter most on day 1:

• The evidence block. It lists links checked, broken links, buttons clicked, and dead buttons, plus a Not confirmed list for anything it could not verify. Test those items yourself, or ask one person to try them.
• The fingerprint check. The report checks for the marks AI builders leave behind and says so in "Slop tells," so you can remove the obvious ones before anyone judges you by them.

It reads public pages only, so anything behind a login is outside it, and it is not a security review or a full accessibility audit. Fix the top item in the lowest-scoring area, then announce.

Who should you tell, and in what order?

In rings, one ring at a time.

1. Ring 1: 5 to 10 people who have the problem. Ask them directly, by name, in a message they can answer in 1 line. Tell them what the app does and who it is for, nothing more.
2. Ring 2: one community where those people already gather. A group chat, a forum, a newsletter you belong to. Post only after ring 1 has finished the main task at least once.
3. Ring 3: everyone else. Social posts, directories, press. Do this after the first 2 rings have stopped finding the same problems.

Each ring should fix what the last one found, so the next group meets a better app. How to ask people for their time, and how to read what they say, is covered in our guide to honest website feedback before you launch. We will not repeat it here.

What should you watch every day?

3 things, in this order.

1. Did anyone finish the main action? Not "did anyone visit." Count finished sign-ups, bookings, or whatever your sentence says. If your app cannot tell you, that is today's fix.

2. What arrived in your inbox? Messages, reply emails, error alerts, and form submissions. Read all of them, including the angry ones.

3. Where are people coming from, and what are they doing right now? Your analytics tool answers this, and tools count differently. Google's Realtime report shows active users over the last 5 and 30 minutes (page undated, read 8 October 2026; Google Analytics Help). Lovable's analytics counts visits (one browsing session each), not people, and does not read UTM tags (page undated, read 8 October 2026; Lovable docs). Read your own tool's page before you trust its numbers, and see Beyond the MVP for what to measure.

With a handful of users, numbers are stories. "3 of 6 stopped at the same screen" tells you where to look. A percentage does not.

How do you collect first feedback without a survey?

Make it easy to reach you: put a visible contact on every page. How to ask, and what to ask, is covered in the feedback guide. Keep a plain list of what people said, each line dated, in their own words. The wording is evidence.

What do you fix first?

Start with anything that stops the main action or loses someone's work, then anything that makes people hesitate to trust you (the Trust and Legitimacy area), then confusion that 2 or more people share. One person's taste goes in the log. Our guide on how to know when your MVP is good enough owns the full sort of what blocks a launch and what can wait.

Change one thing at a time, then click the main path again on a phone. If a change breaks something, go back to the last version that worked.

What does a week look like for a real app?

Hypothetical example: A founder builds a carpool coordination app for school pickups. Parents join a pickup group, say which days they can drive, and see who is next. The founder's week 1 success sentence: "6 parents in one class set up a group without asking me for help."

Notice what the founder did not do: add a feature, chase press, or spend a day on colors.

What if nobody shows up?

A quiet week can mean the telling needs work, not the app. First, check that the link opens for someone else, on their phone, off your Wi-Fi. Then ask 5 specific people, by name, to try one task. If they come and leave without finishing, the problem is in the app. If they do not come, it is in the asking. Both are fixable, and neither is a verdict on whether the idea is good.

When is week 1 over?

When one ring has finished the main task without you, your top fix list is short, and you can say what you will learn in week 2. After the fixes, click the main path again on a phone and re-read your day 1 list, crossing off what is fixed.

Launching is not one day. It is a week of small, specific fixes with real people watching.