| raw | raw/roger-chen-number-1-app-twice.md |
|---|---|
| url |
TL;DR
Roger Chen built two consumer-social apps to #1: Lobby (500k DAU, #1 in 5 countries via product-led viral growth) and Bro (AI companion that floats over your texts, #1 US off TikTok). His method: prototype in ProtoPie before building, use ads as validation not scaling, and engineer network density. A product-led-growth + content-market-fit masterclass.
Key lessons
- Prototype before you build — fake the app in ProtoPie (clickable mockup), carry it around, use it yourself, then test the mockup on TikTok. Closes the dreadful 2-week–3-month “in your head, no feedback” gap; Bro’s first 7.9M-view video was a mockup.
- Ads as validation, not scaling — a few thousand dollars of TikTok ads gets instant feedback on (1) a working format and (2) whether people want the product — far faster than negotiating creators for weeks. Scale with organic/creators (cheaper); validate with ads (reverse of the usual advice).
- Hook + demo with a real person — “people love people”; a face stops the scroll, an app interface gets scrolled past. Mass-produce + a few outlier hits; the winners are unpredictable (“humbled every day”).
- Network density beats content marketing for social — Lobby’s breakout was Israel (60% of users got 5 friends day-one vs 20% elsewhere → day-90 retention 25–30%). Growth came from product features that drive invites: the “group picture” (watermarked boomerang shared to stories), exit-prompt photos, 1-minute “roll-call” calls. Needs relationship density and temporal density (à la BeReal).
- Use AI for context, not content — an AI “third friend”/hype-man that keeps conversations alive, augmenting (not replacing) real social.
- Founder-market fit — only build social if you love social; passion sustains the endless detail-tweaking.
Tools
ProtoPie (prototyping), TikTok ads (validation), Superwall.