How to Get Your First 100 Users for an App

A hundred users is a small number, which is why most plans for getting them are too big. You do not need a viral loop or a press push. You need about a hundred people to find the app, understand the listing and press Get.

Those are three separate steps, and your store console already counts each one. Most makers skip the console and go straight to posting, then wonder why 2,000 views produced 11 downloads. This plan starts with the console and spends two weeks on the step that is leaking.

Day 1: read the console before you post anywhere

Open App Store Connect (or Google Play Console if you ship on Android) and find the acquisition report. Apple calls it App Analytics, and the metric names shift between releases, so read what your console calls them. You are looking for three figures and the ratios between them.

On Google Play the same three steps read as store listing visitors and store listing acquisitions, with the acquisition channels broken out beside them.

Write the three ratios on a page. This is your baseline. We do not quote an industry average here, because conversion rates vary by category, price and country, and a number borrowed from another app tells you little. Your own figure on day 1 is the one to beat on day 14.

If the console shows fewer than a few dozen impressions in total, you have no ratio to read yet. Skip ahead to day 4 and spend the first days on listing work and outreach, then come back to the console once traffic exists.

Why a post cannot fix a listing

Take a hypothetical. You post somewhere, 500 people reach your product page, and 5 per cent of them download. That is 25 downloads. If you improve the page so 8 per cent download, the same 500 visitors give 40. The second post you write for the same audience would need to reach 800 people to match what the page change did on its own.

This is the reason to read the console first. A maker who sees 100,000 views and 35 downloads has two possible problems: the post sent the wrong people, or the listing lost the right ones. The console's source split tells you which. If web referrer traffic reaches your product page and does not download, the problem is the page.

Days 1 to 3: fix the weakest step, only that step

Pick the one step from the table with the lowest ratio, and work on it for three days.

If the weak step is Found. Your keywords are the first thing to check. The App Store has a 100-character keyword field and the title and subtitle count towards search too. Find the phrase a stranger would type when they have the problem your app solves, and check that it appears in your title, subtitle or keyword field exactly once. Repeating a word across fields wastes characters. The aso skill is an open, readable file that walks through the audit order Kwern's agents use, and the ASO agent page says what that agent drafts each week.

If the weak step is Understood. Look at your icon, name and subtitle at thumbnail size, next to five competitors, on your phone. One of them will read at a glance. If yours does not, change the subtitle first, since it costs nothing and goes through with your next version.

If the weak step is Chosen. Rebuild the first three screenshots so each one carries one sentence of benefit, written the way the customer would say it. Move the strongest line of your description to the top, because that is the part shown before the "more" tap. If you have no ratings yet, that is the next job (day 8).

Keep a record of what you changed and the day you changed it. A console report cannot credit a change you did not write down.

Days 4 to 7: pick one channel, and pick it by where the question is asked

Spread across five channels, a first hundred users takes months. On one channel it takes days, because you learn what works while you are still doing it.

Choose the channel by a test: where do people type out the problem your app solves, in words, to strangers? For many app makers that is Reddit, a Hacker News thread, a niche forum or an X search. Search for the problem in the words a customer would use, not your app's name.

Then work it like this.

1. Find 10 threads from the last two weeks where someone describes the problem.
2. Answer the question fully in the reply, so it stands alone without a link.
3. Mention your app only where it is the honest answer, say you made it, and add the store link as a second line.
4. Read the community's rules first. A subreddit that bans self-promotion gets a helpful answer with no link, or no reply.

You post these by hand. That is the right way to do it on Reddit and Hacker News, where an account that posts links on a schedule gets removed. Kwern's Reddit agent drafts the replies and finds the threads, and you paste and post them yourself from your own account. The community-marketing skill covers the etiquette in detail.

Give the channel a kill rule before you start. Two removed comments, or seven days of replies with no clicks, and you move to the next candidate. A rule written in advance saves you from the sunk-cost argument on day 6.

Days 4 to 7, in parallel: 20 people, one at a time

Make a list of 20 people who have the problem and who you can reach by name. They might be people from a waitlist, a previous project's users, a community you already belong to, or friends of friends. Send each a short message: what the app does in one sentence, why you thought of them, the link, and a request for one honest reaction.

Twenty messages will not give you a hundred users. They give you your first ten, and ten people who tell you why they did not download are worth more than a hundred silent installs. Write their words down. They will sound different from your own description, and the difference is your next screenshot caption.

Day 8: ask for ratings at the right moment

A listing with no ratings loses people at the Chosen step. Ask inside the app, once, right after a moment the user liked: a finished task, a streak, a saved result. Apple limits how often the system rating prompt can appear, so do not spend it at launch before anything good has happened.

Reply to every review, including the bad ones, in plain words and without promises you cannot keep. A developer response is public, and a prospective user reads it.

Days 9 to 12: a small, capped Apple Ads test

If your Found step is weak and posting has not moved it, a small paid test gives you data in days instead of weeks. Apple Ads (Search Results) puts your app at the top of the App Store search for a keyword you choose.

Set a budget you can lose. Start with a handful of exact-match keywords, the ones nearest to the problem a customer types, and leave broad match off. Read the cost per install after a few days, then do the arithmetic. As an illustration, if one install costs $2, a hundred installs cost $200. If your app earns revenue, compare that figure with what an installer is worth to you, which RevenueCat can tell you per campaign.

If the numbers do not work, you have learned that for $50 or so, not after a quarter of guessing. The Apple Ads agent drafts keyword and bid changes against a monthly budget that you set, and you approve each one. See the Apple Ads connector and the RevenueCat connector for what it reads.

Day 13: tag every source so the console can tell them apart

From day 4 onwards, every link you post should be a different link. Apple lets you generate campaign links in App Store Connect, each with its own campaign name, and the Sources report then separates them. Use one for Reddit, one for your outreach list, one for each post. Otherwise, when you reach 100, you will not know which of the five things you tried did it.

Day 14: count, and decide

Return to your day 1 page and fill in the same three ratios, plus the download count by source. Then answer three questions in writing.

• Which source gave the most downloads, and did those users open the app a second time? Download counts are the console's figure. Opens and retention come from your own analytics.
• Which step improved, and by how much compared with day 1?
• What is the single next experiment?

If you are at 100 downloads, you have a baseline and a working channel to keep feeding. If you are at 40, you have a clear read on which step is still leaking, and the next two weeks go there. Both are a better position than 100 downloads from a lucky post that you cannot repeat.

Where Kwern fits, and where it does not

Everything above can be done by hand with a spreadsheet and the two consoles, and for a first hundred users that is a reasonable way to do it. Kwern is for the weeks after, when the same audit has to run every week and nobody has read the listing since last month. Its ASO agent reads App Store Connect and drafts listing and review-reply changes that wait in your approval feed. It never publishes without your click, and the model runs on your own provider key. Plans start at $29 a month for one product; the pricing page lists all three.

If you are launching this week, the last-seven-days checklist covers the waitlist and the listing before anything else. If your product is a website and not an app, the same console-first logic applies with Search Console in place of App Store Connect.