How to get your first users for a mobile app

Go and be a real participant in the thing your app is for - the hobby, the sport, the profession, the condition - and fix your store metadata while you do it. What you must not do is treat the app stores as distribution on day one. A listing with no installs and no ratings ranks for nothing anyone competes over.

That is the mobile-specific trap. Store search really is the dominant discovery surface at scale, which makes founders wait for it - but it is a surface you earn onto with installs and ratings, and those come from somewhere else first. The pillar covers the manual stage; this page is about what changes when your product ships through someone else's storefront.

Two short text fields decide your organic surface

On iOS the app name and subtitle carry ranking weight, plus a keyword field the user never sees. On Android the title and the descriptions are indexed. That is the whole lever. Nowhere else in software does so much organic discovery hang on so few characters - and most founders spend those characters on a brand name and a slogan, which are the two least searched things they could have written.

The first screenshot is the headline

Search results show your icon, name and the first screenshot or two. Most people never swipe. So the first tile has to carry the entire proposition: one clear claim in large text over a real screen, not a lifestyle photo and not screen four of an onboarding flow. This is the highest-yield hour of design work available to a new app and it is routinely treated as an export chore.

Ratings gate conversion, and the OS limits how often you can ask

A visible star rating sits next to you in every result. The platforms provide a native review prompt and cap how often it may be shown, so you get very few asks per user - which means the timing of the ask is a real decision, not a detail. Ask immediately after the user has succeeded at the thing the app is for, never on first launch, and never after an error.

You cannot easily contact your users

Unless someone hands you an email address, your only channel to an installed user is push - and push requires a permission the user can refuse permanently on iOS. Which means the permission prompt is itself a conversion event: ask after the user understands what notifications would be for, not on the splash screen, because a refusal is close to irreversible.

Almost never in app-focused places. The people who discuss apps as a category are reviewers and builders; the people who install yours are doing something else and want it to be easier.

• The activity, not the software. A climbing-log app belongs in climbing forums and gyms, not in app subreddits. This is the single most reliable rule on this page and the most commonly ignored.
• Reviews of the apps people already tolerate. One and two-star reviews of the incumbent are a free, structured list of unmet needs, written by people who have already proven they will install something.
• Video, because mobile is visual and vertical. A phone screen recording of the app doing its job is native to every short-video surface. No other product type on this cluster demonstrates so cleanly in fifteen seconds.
• "Best app for X" articles and AI answers. These are answered from the open web, not from the stores, which is why an app with no website is invisible to a growing share of the people looking for it.
• In person, where the activity happens. Unfashionable and effective: the club, the market, the clinic waiting room, the conference. Mobile is the one type where you can install it on a stranger's phone in front of you and watch them use it.

Ranked for a consumer or prosumer app in its first month. Store search sits mid-table on purpose: it is the biggest channel eventually and the weakest one now.

• Do not ask for the push permission on the splash screen. A refusal is close to permanent, and you have just spent your only owned channel to save fifteen seconds of explanation.
• Do not trigger the rating prompt on first launch. You have very few asks per user. Spend them right after a success, never after an error.
• Do not buy installs to "seed" the listing. Users acquired that way churn immediately and rate badly, and both effects persist long after the spend stops.
• Do not localise before you have retention in one language. Localised metadata is a genuine ASO lever, and it multiplies a broken first session just as efficiently as a good one.
• Do not treat downloads as the number. Downloads are the cheapest action a human can take on a phone. They tell you your screenshot worked and nothing else.

Mobile gives you an unusually clean funnel through the store dashboards. Use it, then ignore most of the rest.

• Impressions, product page views, installs. Both stores report a version of this. Weak impressions is a ranking and metadata problem; strong impressions with weak installs is a screenshot, icon and rating problem. These need completely different fixes.
• Day-one retention, before anything else. If people do not open the app again the next day, no channel will help. This is the number that decides whether your first hundred installs are worth acquiring.
• Install to first successful action, unaided. Not signup. The moment the app did its job once, with nobody coaching the user through it.
• Rating volume as well as average. A near-perfect score from a handful of reviews is read as untested, not as excellent - visitors read the count alongside the stars, so volume is its own signal.
• What the two-star reviews say, verbatim. One-stars are often rage; two and three-stars are people who wanted it to work and can tell you exactly where it did not.

Mobile is the least forgiving type in this cluster. You rent distribution from two companies, you cannot reach your own users, you cannot ship a fix in an hour, and the one asset that compounds - your rating - is set early and moves slowly afterwards. So sequencing matters more here than anywhere else: retention first, metadata second, acquisition third. Reversed, you spend money to permanently damage the only asset that would have made the money work.

If you are unsure which community is genuinely yours, the channel picker narrows it to two or three.

The channel picker asks what your app does, who it is for and how it makes money, then names the two or three channels worth your first month. For mobile it usually comes down to which is genuinely searched more - your category, or the hobby your app serves.

Or get the 30-day post-build playbook Send it