What should an app post on TikTok?
By gr0w Research · Updated · From 185 founder interviews · 5 min read
Short answer
Post short videos that open with a hook (a scenario, a problem or a reaction) and then show the app doing something interesting. In an analysis of 185 founder interviews, 40 of the 64 TikTok-organic founders who said how they started posted content themselves, and the ones who reported conversion say videos that only get views, without the product on screen, don't turn into downloads. [1] [2]
- 40 of 64
- TikTok-organic founders who said how they started began by posting content themselves (63%)
- 72
- case studies in the library with TikTok organic as a growth driver
- 75M views
- StudyFetch's lecture-hall demo, which brought hundreds of thousands of sign-ups, per its marketing lead
- $20 a video
- Halo AI's paid trial rate for UGC creators, all posting one fixed format
What formats work for awareness?
For awareness, founders used novelty formats, slideshows and face-to-camera memes. These get views but convert less than demos, so most treat them as top of funnel.
- Slideshows: Glam Up's first faceless post, a before/after makeup slideshow with a screenshot of the app, got about 10M views, according to Nicole. [3] Jack Friks copied a viral hobby-slideshow format and posted it about 400 times. [4] Pingo said its slideshows could reach a million views but burned out quickly. [2]
- Novelty hooks: Coconote's PDF-to-Brainrot lead magnet got 100–200M views but drove only about $25K. [1]
- View-heavy, low-intent content: one relationship-themed TikTok got 150K views and only about 100–200 downloads. [4]
A machine-learning researcher found that scratch-map videos with a 'post-breakup activity' hook got millions of views and almost no downloads. Changing the text to 'procrastinating for exams' produced App Store jumps. [5] [5]
What does a good demo video look like?
A good demo video shows the core feature working, in a way that is easy to follow on a phone screen.
StudyFetch's marketing lead said his launch-day TikTok, filmed by pointing a camera at his computer screen, got about 3M views overnight. [6] A later lecture-hall video ('I thought finals were supposed to be hard') panned to a laptop recording the lecture while notes generated. It reached 75M views and hundreds of thousands of sign-ups. [6]
Cal AI required influencers to show three screens in every video: the camera, the loading screen and the food detail page. [7] Coconote built creator-only screens that look better on camera than the real UI. [1]
Some founders build features to be filmable. StudyFetch designed notes to stream onto the screen with hands off the keyboard. [6] Pingo kept an open-ended custom scenario mode, which creators used to find its viral format. [2]
How do you post problem/solution videos?
Frame the product as the fix for a specific problem the viewer has, then demo the fix.
Coconote's Zack said videos showing the product solving a student problem drove $100K ARR in 45 days. He preferred a 10M-view problem-framed video to a 40M-view novelty one. [1] [1]
Copia's hook had a deadline and a promised result: 'If you're seeing this before May 7th, repeat these affirmations for 7 days.' It had phone-screen demo overlays and hit 10M views. A separate 6.6M-view video skipped the hook and demo, which the founder estimated cost 20–30% of potential downloads. [8] [8]
Stella's Sarah Pearl said videos that mention the app without showing the use case convert worse. A 300K-view demo converted very well. [9]
Evo Marketing's Dawson Gibbs said a simple formula is a hook with a scenario, then the demo. [10]
Do reaction videos work for apps?
Reaction videos work when a face or a reaction is the hook and a demo follows straight away.
- Face reaction into demo: Prayer Lock's Mao Baron copied a competitor's girl-reaction video with a demo, filmed by his girlfriend. After about 40 posts one hook worked, giving about 10K views a day and his first $1K a month. [11] [11]
- Green screen or stitch: a creator reacts to an expert's or founder's video. Evo Marketing described it as low effort with borrowed credibility. [12] Wrestle AI's founder said stitch and reaction videos did very well as ads. [13]
- Reaction to the app itself: Pingo's 52M-view video showed the AI roasting a user until they cried on camera. Creators found it through the custom scenario feature. [2]
The Evo Marketing founder said a 5-second long-text video gets many views but converts badly. A face reaction that cuts into a demo converts better. [12]
How should the app show up in the video?
The app should be the main character: on screen early and doing something, but not announced like an ad.
- Early: Cal AI's Alfie Dickens said the app should appear within the first 15 seconds. Its best converters were day-in-the-life vlogs where a creator scanned food without naming the app. [14]
- Name on screen, not in the script: Cal AI often doesn't say the app's name, just shows it on screen. [15] Halo AI uses a large loading screen with its name and logo. [16]
- Soft CTA: Payout's Casper put 'search it on the App Store' in the comments and cut videos shorter, because CTAs reduce retention. [17] StudyFetch instead ended some videos with a large URL for 4–5 seconds. [6]
- Not at the end only: StudyFetch said videos where the product appears only in the last 5 seconds of unrelated content don't convert. [6]
How do you find formats and how often should you post?
Most founders copied formats that already worked in their niche, then posted many variations.
- Copy and reskin: NaturalWrite's founders copied viral niche videos one-to-one. Their copy got 1.5M views against about 150K for the original, and an 'essay due in five minutes' video hit 4M. [18] Stronger reskinned formats from other verticals. [19]
- Expect a drop: Clover's Matt said a clone does about 20% of the original's views at median, so 5–10 variations match the original. [20]
- Volume: Runafy said about 1 in 10 videos went above 500K. [21] Halo AI used one fixed format, paid $20 a video in trials, and the founder said views follow a power law. [16] [16]
- Frequency caveat: Runafy saw views fall to about 5 on TikTok after raising its posting volume, then moved to Instagram. [21]
The library doesn't show a single best posting schedule.
| Goal | Format | Example | Reported result |
|---|---|---|---|
| Awareness | Before/after slideshow | Glam Up [3] | ~10M views on first post |
| Awareness | Novelty lead magnet | Coconote PDF to Brainrot [1] | 100–200M views, ~$25K |
| Demo | Screen recording of a feature | StudyFetch lecture hall [6] | 75M views, hundreds of thousands of sign-ups |
| Problem/solution | Deadline hook plus phone demo | Copia [8] | 10M views |
| Reaction | Face reaction into demo | Prayer Lock [11] | ~10K views/day, first $1K/month |
| Reaction | AI roasts the user | Pingo [2] | 52M views |
| Reaction | Green screen or stitch | Evo Marketing [12] | Low effort; borrowed credibility |
Frequently asked questions
Should an app say its name in TikTok videos?
Not necessarily. Cal AI videos often show the app name on screen without saying it. Payout's Casper put 'search it on the App Store' in the comments, since CTAs reduce retention. [15] [17]
Do videos with millions of views bring downloads?
Not always. Coconote's 100–200M-view novelty video drove only about $25K, and a 150K-view relationship video drove about 100–200 downloads. Founders say views need the product on screen to convert. [1] [4]
How many videos until one works?
It varies. Prayer Lock's founder found a working hook after about 40 posts. Runafy said about 1 in 10 videos went above 500K views. [11] [21]
Sources
Founder interviews on YouTube; numbers are as the founders said them. How this research is done.
- [1]COPY this ai app's $6.7m/yr marketing strategy · Coconote (Superwall)
- [2]How I built a $500k/mo ai app (so you can just copy me) · Pingo (Superwall)
- [3]she makes $400k/mo from ai apps (while in college) · Glam Up, Sprout (formerly Prep AI) (Superwall)
- [4]He Went From Struggling to Make $1 From His App to Turning Down $1M · Curiosity Quench, PostBridge, Doof, Lovely (Superwall)
- [5]The NEW Short Form Strategy Dominating For App Growth (Superwall)
- [6]The EXACT System to Go From 0-$10M+ With Your App · StudyFetch (Superwall)
- [7]I speedran a new app from 0 to $2m/year in 6 months (here's how) · Sway AI, Cal AI (Superwall)
- [8]Copy This Growth Strategy, It'll Blow Up Your App · Copia (Superwall)
- [9]She Built This $340K/Month App in 60 Days · Stella (Starter Story)
- [10]How I Built the #1 App on The App Store (Twice) · Lobby, Bro (Superwall)
- [11]I Built a $20K/Month Mobile App: Here’s My Entire Playbook · Prayer Lock (Starter Story)
- [12]The 1 Billion Views Content Strategy Every App Should Steal in 2026 · Evo Marketing (Superwall)
- [13]Meet the ex-wrestler making $41k/mo with AI apps · Wrestle AI, Brawler, AbMax (Superwall)
- [14]Meet The Marketing Genius Behind Cal AI · Cal AI (Superwall)
- [15]The Secrets of Consumer Club: Insights from Joseph Choi - Superwall Podcast · Consumer Club (Superwall)
- [16]0 to $300k/mo in 45 days with my ai app (just copy me) · Halo AI (Superwall)
- [17]The Stupid Simple Content Strategy That's Making Apps Go Viral · Payout (Superwall)
- [18]Our simple app made $100K in 90 days · NaturalWrite (Starter Story)
- [19]My App Makes $600k/Year Using This Easy Growth Strategy (copy me) · Stronger (Superwall)
- [20]The NEW App Content Strategy Dominating in 2026 · Clover, Echo (Superwall)
- [21]I speedran an ai app from 0 to $100k exit in 26 days · Runafy (Superwall)
- [22]The First App I Ever Built Makes $25K/Month · Tone Adapt (Starter Story)