paid ads can get your first 100 app users, but at this stage they are a learning tool, not a growth engine. run one capped test on one platform with three to five creatives, and judge it on cost per retained user, not cost per install. often the honest answer is “not yet”, and that is worth knowing.
this is the paid only playbook. the other channels, most of them free, are in how to get your first 100 app users.
When paid ads make sense this early
paid ads make sense this early only when you need an answer faster than organic can give it: which promise makes strangers tap, which audience sticks, and roughly what a retained user costs. they make sense when the app already keeps some people coming back and you can see who returns. without that, you are paying to learn nothing.
good reasons to run a small test now:
- you have two or three possible positioning angles and cannot choose
- organic posts get views but you cannot tell which message drives installs
- you want a first, rough cost per retained user before raising money or planning a budget
- you need a handful of real strangers, not friends, to test onboarding
bad reasons: “we launched and nothing happened”, “a competitor runs ads”, or “we have budget left this month”.
Why the answer is often “not yet”
the answer is often “not yet” because paid traffic magnifies whatever the app already is. if users leave after one session, ads make that happen at a higher price. early installs from paid are also expensive compared with what a brand new user is worth, so spend buys very little growth before the product is ready.
in my experience, consumer AI apps struggle once blended CAC rises above roughly $0.50, and install costs of $3 to $15 kill an app whose early users are worth cents. blended CAC means total spend divided by all new users from every channel. the fuller cost picture is in how to lower CAC for a consumer AI app.
a quick readiness check, all should be true:
- a stranger can understand the app in the first three screenshots
- some organic users still open the app a week later
- you can see day 1 and day 7 activity in your analytics
- onboarding asks “where did you hear about us?”
- you can lose the whole test budget without pain
Pick one platform for the test
pick one platform, the one closest to how your users already find apps like yours. Meta and TikTok suit apps that are easy to show in a short video. Apple Ads suits apps people already search for by category on the App Store. running two platforms at once on a tiny budget splits the data until neither result means anything.
| platform | best for | what it teaches you |
|---|---|---|
| Meta (Instagram, Facebook) | visual apps, broad consumer audiences | which message and which audience responds |
| TikTok | apps you can show in 15 seconds, younger users | which hook stops the scroll |
| Apple Ads (formerly Apple Search Ads) | apps people search for by category | which search terms bring users who stay |
on Apple Ads, the Basic tier charges per install and runs itself, with a budget of up to $10,000 a month per app in the US, according to Apple’s Basic vs Advanced comparison. Basic hides which searches worked, so for learning, Advanced teaches more. setup is in Apple Ads for apps.
Set a capped test budget
set the budget as a hard cap before launch: a total you can lose completely, spread over a fixed window, with equal spend per creative. write the cap in the ad account, not just in your head. the numbers below are illustrative, to show the maths, not a benchmark for your app.
an illustrative test:
| line | illustrative number |
|---|---|
| total cap | $300 |
| window | 10 days |
| daily budget | $30 |
| creatives | 4, roughly equal spend |
| installs if CPI were $3 | about 100 |
| still active on day 7 if 20% retain | about 20 |
| cost per retained user | $300 / 20 = $15 |
that last line is the one to remember. a “cheap” $3 install became a $15 user who stayed. your real numbers will be different; the point is to find them.
Make 3 to 5 creatives, each a different idea
make three to five creatives, and make each one a different reason to download, not five edits of the same video. a small test can only answer a few questions, so ask different ones. phone footage of a real person using the app is enough. polish is not the variable you are testing at this stage.
a brief you can copy for each creative:
angle: the one reason to download (for example “stop forgetting what you studied”)
hook, first 2 seconds: a line a real user would say out loud
proof: the app doing the thing, on screen, within 5 seconds
ending: what they get in the first minute after install
take hooks from your reviews, support messages and Reddit threads, where users describe the problem in their own words. how to structure a proper testing loop later is in app creative testing.
What to measure: cost per retained user, not CPI
measure cost per retained user: spend divided by the users from that test who are still active on day 7. CPI, cost per install, only tells you the ad got a tap and a download. it says nothing about whether the person stayed, and the cheapest installs are often the worst users. judge every creative on the same retained number.
| metric | what it tells you | use it for |
|---|---|---|
| click through rate | the hook stopped the scroll | comparing hooks only |
| CPI | cost to get a download | a warning light, not a verdict |
| day 1 return rate | onboarding worked | spotting a broken first session |
| cost per day 7 user | what a real user costs | the decision |
ad dashboards and your own analytics rarely match exactly, so cross check the platform’s numbers with your analytics and the “where did you hear about us?” answers for the test window.
Stop rules to write before you launch
stop rules are conditions written before any money is spent that end the test or cut a creative, so emotion never keeps a loser running. write them in a note, share it with your cofounder, and follow them even when you like an ad. a test without stop rules quietly becomes a budget that never ends.
stop rules i use as a template:
- the cap is hit. the test ends, no top ups.
- a creative spends its share with zero day 1 returns. cut it.
- nothing gets a day 7 user after half the budget. pause everything and look at onboarding, not the ads.
- one creative beats the others clearly on cost per day 7 user. keep it, cut the rest, and let the remaining budget confirm it.
- no clear winner at the end. the answer is “not yet”. go back to organic channels and the product.
What to do with the result
what you do next depends on the answer. a clear winner gives you a message to reuse everywhere, in store screenshots, organic posts and your next test. no winner tells you the product or positioning needs work before paid can help. either way, write down what you learned before spending another dollar.
- clear winner: reuse its hook in your first screenshot and organic content, then run a second capped test with new variants of that angle
- cheap installs, no retention: fix onboarding and the first session, then retest
- nothing worked: stay organic for now, and come back when people who find the app already stay
a test that ends in “not yet” is not wasted money. it is the cheapest way to avoid spending ten times as much later.
Unlock the full playbook. Free, takes one minute.
tell me a little about your app and the rest of "How to get your first 100 app users with a small paid ads budget" opens right here, plus every other playbook on this site. i read every answer myself and reply with ideas for your app. no pitch.
Frequently asked questions
Is it worth running ads for a new app?
only as a small, capped learning test once people who find the app already come back. early ads are best at answering questions: which message makes strangers tap, which audience stays, what a retained user really costs. if nobody returns after day one, ads just pay to show that faster. fix retention first, then test.
How much should I spend on ads for my first users?
spend an amount you can lose completely without hurting the business, and cap it before launch. treat it as the price of a lesson, not a growth budget. as an illustration, a few hundred dollars over ten days on one platform is enough to compare a handful of creatives, if you judge them on retained users.
Which is better for a new app, Meta, TikTok or Apple Ads?
pick the one closest to how your users already behave. if they search the App Store for your category, Apple Ads catches that intent. if the app is easy to show in a short video, TikTok or Meta will tell you faster which message works. run one platform at a time so the results mean something.