How GlowUp grows

Picked a niche by looking at the biggest TikTok hashtags (makeup was one): check how many videos are posted and how many views they get. Read comment sections on makeup videos to find what people were asking for, then made content around the idea (getting a makeup look and the products you need) before building. Only after the content went viral did he build a quick MVP. His earlier ~10 apps were built first and marketed after, and got no views or users.

In the app: 'If you invite three friends, you'll get a free makeup look.' Viewers downloaded the app, saw they could share a code, then came back to the TikTok, rewatched it and posted their code (or asked for codes) in the comments. Non-payers kept returning to engage with the post, which boosted it in the algorithm. Combined with an anchor download link added at ~500K views.

Subscription-only: $9.99/week or $39.99/year for unlimited AI makeup looks; no other revenue stream. Paywall via Superwall.

Use a new iPhone with a US SIM card and a VPN set to the US 24/7 (if outside the US). Create the TikTok account, then spend 15-30 minutes a day watching, liking and commenting on videos in your niche for about 3 days so the algorithm treats you as a normal user, not a bot, and you learn the niche's trends.

When a video starts getting good views (a few thousand up to 100K), add an anchor link: a 'download now' button on the video that opens the App Store page, so viewers don't have to go to the bio. Adding it under ~1,000 views kills the video. On the 48M-view video it was added at ~500K views.

Replicate a trend that works well in the niche and put the app idea in it. Post once a day at first and wait for engagement; once there is traction after a couple of days, post as much as possible, 4-6 times a day. Choose a trend that is easy to reproduce, faceless if you can't be on camera, so it can be repeated many times across accounts.

Repeat the warm-up and posting steps on new accounts: 2-3 accounts per phone, buy another phone for more. At peak they ran about 7 accounts posting 8-12 times a day.

Onboarding is deliberately long so it feels personalized: questions about sensitive skin, wrinkles, preferred makeup look and skin tone, all used to generate the result. After the questions and before showing the user's AI makeup result, the app asks them to leave a rating and review. Then it shows the result, the products to buy and the routine.