The Past, Present, and Future of Consumer Social Companies
What this covers
Patrick talks with Turner Novak about how consumer social products have changed across three eras. Turner starts with Facebook, whose ad business worked because users listed their own movies, musicians and activities, and whose margins grew because members supplied the content. He then covers Instagram, which grew on sharing to Twitter and Facebook. After the acquisition, about 75 percent of its new installs reportedly came from Facebook's feed, an example of Distribution Piggybacking. Facebook also copied its newsfeed code into Instagram in about six weeks, and Instagram's design stayed desktop-derived.
The middle of the conversation is about TikTok and its Chinese relatives. Turner describes how TikTok tests each video on about 10 people and widens distribution only if shares, rewatches and follows hold up. He traces its audio memes to Musical.ly. His most contested claims are that TikTok is really a media company and that Zynn and Kuaishou may be next-generation TV networks worth $10 to $50 billion, not $500 billion. Zynn's story is a cash-for-viewing app backed by Kuaishou's financing. Pinduoduo is the commerce counterpart: it began with group buying of fruit, has no real search, and runs a feed that is about 30 percent ads. The talk also covers WeChat mini-programs and China's programmable payments, which Turner contrasts with the 2 to 3 percent fees in the US.
Turner closes with advice for founders. He doubts Instagram will rip out its follower graph, calls "TikTok for X" clones weak, and says Venmo's social strategy did not carry over to Cash App. He also proposes a fantasy angel portfolio as a way to build an investing track record.
Novak argues that durable advantage in consumer social comes from user-generated content, real identity, and friend networks, but that mobile-first entrants like TikTok win by replacing the social graph with algorithmic discovery and by pushing social products toward media and commerce.
- UGC and real identity give social networks high, durable margins because users supply content and ad targeting data for free
- Mobile-first products designed around full-screen video and algorithmic discovery outperform desktop-era feeds and follower graphs
- Social products are moving into commerce and media, where friend graphs are no longer the only source of retention or defensibility
Frictionless, addictive algorithms let social platforms learn users intimately and compete with commerce.
- Social companies learn users' preferences far faster than Amazon's physical feed idea, so a social product that knows users intimately could become a competitor to Amazon's commerce business.
“these social companies are doing that at far faster feedback loops. Like they're learning way faster than a sort of physical feed ever could”
- On the user side, TikTok has almost no onboarding friction: a user downloads the app and is watching videos within seconds, with no need to follow anyone or build a profile, and the algorithm is so effective that it becomes addictive.
“there's literally no friction like you download it, and two seconds later you're watching videos”
- On June 30, 2020, India banned most Chinese apps, a development that adds a geopolitical dimension to the competitive battles between global platforms.
“the majority of these Chinese apps were banned in India”