She Built This $300K/Month App in 60 Days

Works 'backwards': build distribution first (1.2M Instagram, 2.8M TikTok followers in the manifestation niche), then pick the product. Once a video pops off, ask what product could be naturally inserted into that video to help its virality and solve a viewer problem, and read comments for 'I wish this existed' / 'I'm struggling with this'. Stella came from her own habit of listening to generic visualizations and wanting custom ones, which AI now makes possible.

Affirmation video: she recites affirmations ('My mind is worth millions... I get paid to exist'), a second person asks 'How can I remember all these?', and she answers 'You can actually get custom affirmations on your lock screen every single day using my app, Stella.' Core approach: find the most viral content in your niche, study the psychology, and 'recreate it to infinity' once a video hits; she has remade the same viral content for years and it still converts.

Affirmation videos open with an instruction like 'Repeat these affirmations for 7 days' or 'Listen three times'. Because viewers need to come back, they save the video; saves raise engagement and push distribution. General rule: make content that naturally earns shares or saves.

On Instagram, uses ManyChat: 'Comment download and I'll send you your custom affirmations.' The auto-DM sends them to download the app, and also says 'Hey, if you review the app, let me know so I can thank you personally' to encourage App Store reviews.

After launching Stella she started a brand-new TikTok account, and separately runs another account she 'spam posts' on, beyond her 2.8M main account. Claims she can take any new account to ~50K followers in 3-4 months because virality is a skill, not an existing audience.

Compared videos that only mention Stella with videos that demo the use case; demos convert better even with fewer views. Balances pure-viral videos with specific app-demo videos that convert.

No coding background; used Claude Code, Claude and Google AI Studio. MVP took about 2 days, the full app 2-3 months. Treats speed as the main ingredient: the sooner you test an MVP, the sooner you have a product that performs.