How to Get Users for Your App: What Actually Works After You Ship

A designer I know built a meal-planning app in two weekends. She described what she wanted to an AI coding agent, reviewed what came back, shipped it to the App Store, and posted a link on X. Fourteen likes, three signups. Two of the signups were friends. Then nothing, a flat line of nothing, for three weeks. The app worked. The problem it solved was real; she had it herself. And it sat there with five users while she refreshed the analytics and wondered what she had missed.

What she had missed is the half of the job nobody's excited to talk about. Building was supposed to be the hard part, and it used to be. Now you can describe an app into existence in a weekend, which means the number of apps shipping every week has exploded, which means the scarce thing is no longer the ability to build. It is the ability to get anyone to care. If anyone can code anything, distribution is what decides who wins.

So here is the honest playbook for getting users, written for the person who just shipped: what to do first, what actually moves numbers, and why the question you are really asking (which channel works for my app?) has no general answer, only an empirical one.

Shipping is not launching, and launching is not distribution

The quiet assumption behind "I shipped it and nothing happened" is that publishing an app puts it in front of people. It does not. An app store listing and a live URL have no audience attached; they are shelf space in a warehouse nobody visits. Even a proper launch (a Product Hunt day, a well-received post) is a spike, not a system: it produces a burst of curious visitors and then decays back toward zero within a week.

The launch spike is worth dwelling on for a second, because it fools almost everyone once. A good launch day brings a few hundred visitors, a burst of signups, maybe a nice comment thread. It feels like the start of a curve. It is almost always the whole curve: those visitors came from the launch platform's audience, not yours, and when the post scrolls off the front page the flow stops with it. Launches are useful (they produce first users and feedback fast — here is how to run one with no audience), but they answer a different question than the one you have. The question is not "can I get attention once." It is "where do users come from on a random Tuesday in month three."

Distribution is the system that answers that: it produces users on purpose, week after week. It has exactly three parts, and everything you will ever do fits one of them: be where your users already are, say something that makes them care, and do more of whatever measurably works. Every channel, tactic, and growth thread you have ever read is one of those three wearing different clothes.

Start where your users already gather

Before anything scalable, do the manual thing: find the rooms where your users already talk about the problem, and join the conversation. For the meal-planning app that meant a handful of nutrition subreddits, two Discord servers, and the comment sections of meal-prep creators. For a developer tool it is different rooms, same move. The people with the problem congregate somewhere, and they are already describing the pain in their own words.

Contribute honestly (answer questions, share what you learned building), and invite people directly, one at a time. This does not scale, and that is fine, because scaling is not the current job. The current job is two things: your first fifty real users, and the exact words those users use for their problem. The second one matters more than it looks. It is the raw material every later channel runs on, and no tool can generate it for you. The full version of this motion is the first customers playbook, and it is manual on purpose.

The message decides whether anything else works

Here is what most app marketing gets backwards: founders describe what the app does, and users only care about what it fixes. "AI-powered meal planning with macro tracking" is a feature list. "Stop deciding what to eat when you're already hungry" is a reason to tap. The gap between those two sentences is the difference between a channel that converts and a channel that seems broken.

You find the second sentence by listening, not brainstorming. The communities from the last section, the replies to your invites, the one user who texts you unprompted — they will hand you the phrasing. Your job is to notice which pain, said which way, makes people stop scrolling, and then to lead with it everywhere. A mediocre channel with a proven message beats a perfect channel with a guess. And when a channel fails, the first suspect is always the message, not the channel.

Nobody can tell you which channel is yours

Now the question you actually googled. Should you do SEO, ads, cold outreach, LinkedIn, short-form video? And the honest answer is the one no listicle will give you: nobody knows, because it depends on what your app is, and the dependence is total.

A B2B compliance tool charging $500 a month can afford cold outreach; a $4.99 consumer app cannot, because the math of writing to one person at a time never closes. A product people already search for ("meal planning app") fits SEO; a product nobody knows to search for needs to interrupt feeds instead. Visual, personal products travel on social; boring, expensive ones travel on referrals and outbound. The channel lists you read are survivor stories: true for the app whose founder wrote them, useless as prescriptions, because the advice is downstream of the business.

So treat channels the way you treated your product: as hypotheses to test cheaply, not commitments to make on faith. A few hundred dollars of ads buys you a week of message data. Twenty researched outbound emails cost an afternoon. One month of honest posts in the right community costs consistency. Run a few of these small, watch the numbers, kill what stays silent, and double down where real users actually arrive. Which channels to test first and how to read the results is its own post: the best marketing channels for startups. The pattern that wins is never "picked the right channel on day one." It is "found the right channel by week six because the tests were cheap and honest."

One measurement note, because it changes what "working" means: count the users who come back, not the ones who arrive. Raw signups are the vanity layer; an ad test that produces forty signups who never open the app again has actually failed, and twelve signups who are still around on day seven is a real signal. Early on, define one action that means someone genuinely used the product (planned a week of meals, ran the tool twice) and judge every channel by how many of those people it delivers. Channels differ far more on quality than on volume, and the quality number is the one that predicts whether the app lives.

What you can hand off, and what you cannot

All of the above is work, and the reason most shipped apps die is not that the founder could not do the work; it is that the founder was also fixing bugs, answering support, and having a life. The research, the drafting, the posting, the measuring, the follow-ups — that is hours every day of execution, and it is exactly the kind of long-horizon loop that AI agents have become able to run.

That is the line Revnu is built on. You can already tell an AI agent "build me this app" and watch it happen; Revnu exists so "now get me users" is a sentence too, whatever kind of business you built, B2C or B2B. It runs the loop from this post — joins the channels, drafts the content and the outreach, runs the cheap experiments, kills the silent ones, doubles down where users respond — as one growth employee with one memory, not a stack of channel tools. What it never takes from you: judgment. Every post, email, and ad waits for your approval before it goes out, and what your product is and stands for stays yours. You are the founder; it is the growth hire.

Where this leaves you

The app is built, so build the other machine. This week: find three rooms where your users already gather, start real conversations, and invite ten people personally. This month: collect the exact words users say about the problem, pick the two channels whose economics fit your app, and run one small honest test in each. Then follow the evidence without sentiment — kill what is silent, feed what responds, and let the message that wins in one place run everywhere. Do that yourself, or hire it done (human or agent) — but treat distribution as the second product you are building, because for every app shipped after this year, it is the one that decides.