How to market an app on TikTok

You shipped an app. Nobody knows it exists. Paid ads cost more than the app earns, and the App Store gives you nothing until people already search for you.

TikTok is where indie apps get their first real users, because it's the one place where a brand-new account can out-reach an established one. But the way most builders use it — demo videos, feature announcements, "check out my app" — is the way that doesn't work.

Why app demos don't travel

The feed doesn't distribute products. It distributes usefulness and feeling. A demo of your interface is interesting to you and to nobody else, because the viewer hasn't yet got a problem your app solves — they've got a thumb and eight seconds.

The posts that travel are the ones that would be worth watching even if the app didn't exist. Your app appears at the end, as the thing that makes the advice easier, not as the reason the post exists.

That inverts the usual plan. Instead of "how do I show my app?", the question is "what does my user already want to know, and what can I tell them that's actually true?"

The content that works for apps

Tips in your app's subject. A habit tracker posts about habits, not about habit trackers. A budgeting app posts about the specific ways people lose money. You have this knowledge — it's why you built the thing.

Mistakes. People will watch an entire post about what they're doing wrong and skip the same information framed as what to do right.

Before and after. Only if it's real. A fabricated transformation is the fastest way to lose the audience you just found.

The thing you learned building it. Builders underrate this. The research you did to make the app is content nobody else in your niche has.

Slideshows are the cheapest format for all of this. No filming, no editing, no being on camera — a hook, four or five tips, and photos. You can make a week's worth in the time one video takes, and volume is what this needs.

Mentioning your app without killing the post

Getting this wrong is the most common mistake. Two rules.

Put it at the end, on its own slide. The last slide names the app, shows what it looks like, and says what it does in one line. Everything before it stands alone. If someone swipes away at slide four they still got value, and the platform still saw a good post.

Say it like a person recommending something. "the app i use for outfit scores" works. "Download NAME today!" does not. You're not writing a call to action, you're finishing a thought.

Some of your posts shouldn't mention the app at all. A feed that's 100% promotion reads as an ad account. Aim for something like three in four posts ending with the app and one in four just being useful.

Posting like it's a job

The single strongest predictor of whether this works is how many posts you make. Not how good any one of them is.

Three to five posts a day, every day, for at least a month, is the shape of a serious attempt. Most posts will do nothing. One in twenty will do ten times the others, and that one pays for the month. You cannot pick which one in advance — anyone who tells you they can is selling something.

This is also why the format matters. If a post takes you an hour, you'll do five a week and quit in three weeks. If it takes two minutes, you'll do five a day and still be there in month three, which is where the results are.

Accounts: one or several

Start with one account in one niche, so the algorithm learns who to show you to. Once it's working, a second account in a nearby niche is the cheapest way to double reach — the same app, a different audience, different tips.

Don't run five accounts with the same posts. Duplicate content across accounts gets suppressed, and you'll learn nothing from the comparison.

Knowing what actually worked

The App Store will tell you installs went up. It won't tell you which post did it.

Use a separate install link per account, so downloads are attributable. It's the difference between "TikTok is working" and "the fitness account is working and the food account isn't", and only the second one tells you what to do next.

Watch three things and ignore the rest:

• Views, as a measure of whether the hook stopped anyone.
• Completion, as a measure of whether the post deserved the stop.
• Installs per account, as the only number that pays rent.

Likes and follower count are the numbers that feel best and mean least. A post with 400 views and nine installs beats a post with 40,000 views and none.

A first month that's actually doable

• Week 1 — one account, twenty posts, five hook shapes. Learn what your niche responds to.
• Week 2 — keep the two hook shapes that held attention, drop the rest, keep posting daily.
• Week 3 — add an install link per account and start reading the numbers.
• Week 4 — if one niche is clearly ahead, open a second account next to it.

Nothing here is clever. It works because almost nobody does it for four weeks straight.

Fyp AI exists to make step one take minutes instead of evenings: paste your App Store link, get slideshows written from your listing and your real reviews, edit what you want, send them to your phone and post. See the plans.