My AI-Generated Influencer Made Me $100K
19-year-old app founder Raphael Kramer explains how his app FaceKit generated over $100,000 in revenue in just two months using AI-generated influencers on TikTok. He built 40+ fake but realistic AI personas that posted over 1,300 videos across platforms, garnering 100 million views organically. The strategy eliminated traditional influencer costs while driving app downloads through before-and-after transformation content.
Summary
The video features Rafael Kramer, a 19-year-old serial app founder who has built over 50 apps since age 13, discussing his latest app FaceKit and the unconventional AI-generated influencer strategy behind its rapid growth. FaceKit is a 'looks maxing' app that uses Apple's TrueDepth camera technology (the same used in Face ID) to perform 3D facial scans and analyze various facial ratios and measurements. The app operates on a subscription model with weekly, monthly, and yearly pricing options.
Raphael's growth strategy centers entirely on AI-generated influencer personas posted across TikTok and Instagram. He and his team created approximately 40 accounts featuring realistic AI avatars — complete with consistent profile pictures and varied content — that post before-and-after transformation videos promoting FaceKit. In two months, these accounts published over 1,300 videos, accumulated over 100 million views, and nearly 10 million likes, all without spending money on traditional influencer partnerships or paid ads. Monthly clipper costs were approximately $2,000 against $35,000 in net proceeds after Apple's platform fee, indicating extremely high profit margins.
Raphael describes his ideation process as largely coming from 'doom scrolling' TikTok, which is how he discovered the AI influencer page 'Logan Reed' that inspired FaceKit. He reached out to its creators, collaborated on the app concept, and then scaled the same content strategy across dozens of accounts. The team uses a tool called Nano Banana to generate consistent AI avatars and OpenAI's Sora for one-off viral video concepts, such as a fake professor telling a student their facial ratios are unattractive — absurd content designed to go viral while still funneling viewers to the app.
On the ethics of deceptive AI content, Raphael argues that the method is simply a more efficient form of marketing, and that as long as the underlying app isn't harmful — which he believes FaceKit is not, given it promotes skincare, hydration, and fitness — the approach is justifiable. He acknowledges the app capitalizes on insecurities but frames this as standard marketing practice. He also acknowledges the looks maxing space is 'questionable' but defends FaceKit as science-based.
Raphael's broader advice for aspiring app founders includes identifying niches with clear, visual before-and-after transformations, building an MVP with just an onboarding flow and paywall as quickly as possible, and falling in love with the entrepreneurial process itself rather than fixating on outcomes.
Key Insights
- Raphael claims that 100% of FaceKit's $100,000+ revenue in two months came from organic AI-generated influencer content across 40 TikTok accounts, with over 1,300 videos posted and 100 million views accumulated — all without paying traditional influencers or running paid ads.
- Raphael argues that the core unlock for this strategy was reaching a point where AI-generated human avatars became indistinguishable from real people — he discovered this when he encountered the 'Logan Reed' TikTok page and genuinely could not tell it was AI, despite considering himself someone who could usually spot it.
- Raphael describes keeping clipper costs to approximately $2,000 per month against $35,000 in net proceeds after Apple's fee, demonstrating that replacing human influencers with AI avatars produced exceptionally high profit margins for the business.
- Raphael's team uses a coordinated group chat system where any team member who identifies a high-performing TikTok format immediately shares it so all 40+ accounts can replicate it simultaneously, treating virality as a repeatable, distributed process rather than a one-off event.
- On the ethics of AI-generated deceptive content, Raphael frames it purely as an efficiency gain in marketing, arguing that as long as the app being promoted is not harmful, using AI to generate fake personas is no different from hiring real content creators — and that capitalizing on insecurities is simply what all marketing does.
Topics
Full transcript available for MurmurCast members
Sign Up to Access