This interview breaks down the history and future of social media companies. Turner Novak says the key isn't fast growth but quality growth—Facebook built a moat by using real identities and user-provided data for ads. He's bullish on TikTok because it's mobile-first and removes the 'follow' feature, using algorithms to recommend videos, making it hard to copy. He also highlights Pinduoduo as the first mobile-only e-commerce app, using group buying for free customer acquisition. Snapchat's 'Snap Minis' could become a US WeChat, but needs better payment and delivery systems.
Turner Novak (partner at Gelt VC) discussed the history, current state, and future of consumer social companies on the program. Core views include: social networks are the fastest-growing companies in history, but quality growth is more important than speed; network effects are key, and attention mu
Turner Novak (Partner at Gelt VC) systematically reviewed the historical evolution, core analytical framework, and future trends of consumer social companies in an interview. Core judgment: social networks are the fastest-growing company type in history, but quality growth is more important than speed; TikTok's rise stems from its parent company ByteDance's product logic designed entirely for mobile, and simply copying TikTok is futile.
Turner Novak argues that the core competitive advantage of social networks lies not in growth speed, but in growth quality—especially the moat created by real identity and friend relationships.
Facebook was once the fifth fastest-growing social network, yet it ultimately became the biggest winner. Novak points out that the key was Facebook's construction of a user-generated content (UGC) system based on real identity. UGC essentially lets users produce content for free, thereby generating high profit margins (advertising equals revenue). More importantly, Facebook's product allows users to "actively create their own advertising profiles"—when updating their personal information, users fill in 10 favorite movies, 10 favorite musicians, and favorite activities, data that later becomes the foundation for ad targeting.
"Facebook users are not only creating UGC; they are also creating their own advertising profiles." — Turner Novak
Comparison Data: Reddit vs. Facebook
| Metric | ||
|---|---|---|
| User Identity | Anonymous/Pseudonymous | Real Identity |
| Ad Targeting Capability | Weak (does not know who the user is) | Strong (precise personal profile) |
| US User ARPU | Not disclosed, but far lower than Facebook | Approximately $186/year (over 12 months) |
Evaluation Framework for New Social Networks: Novak proposes three key markers—① UGC content; ② Real identity; ③ Helping ordinary people express themselves and connect with others. Successful social networks typically capture an emerging behavior as it arises, build a new friend network around that behavior, and ensure that behavior is difficult for existing giants to replicate.
Novak argues that TikTok is the first scaled broadcast network built entirely for mobile, with its core innovation being the removal of the traditional "follower graph" from social networks, replacing it with algorithm-driven content recommendations.
Historical context: The "feed" on Facebook and Instagram was originally designed for desktop (Instagram's feed code was copied directly from Facebook, modified in just six weeks). Snapchat was the first true mobile-first social network; its "Stories" feature enabled full-screen mobile video, but users still had to open the feed first and then tap into a story. TikTok eliminates all intermediate steps—opening the app brings you directly to a full-screen video.
Mechanism breakdown:
Data support: ByteDance found that during 2018–2019, among the top 100 mobile game advertisers in China, 68 allocated more than half of their marketing budgets to ByteDance's platforms (including TikTok's predecessor).
Falsification condition: Novak believes that if Instagram truly wants to compete with TikTok, it must "completely remove the concept of the follower graph," but he judges that Instagram will not do so—because the follower graph is its core DNA.
Novak argues that Pinduoduo is the first e-commerce company built at scale entirely for mobile, and its model could be replicated in the U.S., but only with the support of payments and logistics infrastructure.
Pinduoduo’s model:
Growth data:
U.S. opportunity: Novak believes the U.S. needs two conditions: ① a programmable payment system (with fees below 2%-3%); ② a programmable last-mile logistics API (e.g., Postmates, DoorDash’s APIs). These conditions are not yet mature, but Snapchat’s “Snap Minis” (similar to WeChat mini-programs) could serve as a breakthrough—Snapchat has approximately 84 million daily active users in the U.S., close to Instagram’s North American user base, but with a huge gap in advertising revenue, indicating a “pricing gap.”
Novak notes that Zynn, a TikTok-like product under Kuaishou, briefly topped the App Store chart through cash incentives and content replication, but this exposes the customer acquisition cost problem for social companies without a social graph.
Zynn's tactics:
The company behind it: Zynn is backed by Kuaishou, ByteDance's main competitor in China, valued at approximately $30 billion, which raised $3 billion in December 2019 (led by Sequoia and Tencent). Kuaishou's revenue model differs from ByteDance: over half of its revenue comes from live streaming (similar to TV shopping), rather than advertising.
Analysis: Novak believes that media companies without a social graph are essentially "building the next generation of television networks" — they need to constantly spend money to acquire content, and are unlikely to become Facebook-level companies ($500 billion+), but can become $10 billion to $50 billion level companies (like Viacom, Disney, Netflix). The true moat still comes from user retention driven by "friends creating content."
Novak pays close attention to Snapchat's "Snap Minis" strategy, believing it could launch an American version of the WeChat ecosystem.
Snapchat's Assets:
Snap Minis: Similar to WeChat mini-programs, running lightweight third-party applications within Snapchat. Novak envisions that U.S. developers could quickly build social e-commerce companies like Pinduoduo on Snap Minis, but this requires mature payment and logistics APIs.
| Target | Guest Attitude | Key Data |
|---|---|---|
| TikTok / ByteDance | Positive (core innovation is mobile-first + no social graph) | #1 in global downloads in 2019; 68 of China's top 100 mobile game advertisers allocated over 50% of their budgets to ByteDance |
| Snapchat | Neutral-to-Positive (bullish on Snap Minis potential) | 84 million daily active users in North America; analysts value Instagram at $200B–$400B (implying a pricing gap for Snap) |
| Pinduoduo | Positive (leading social commerce model) | Revenue of $4.1B in its 5th year; China's second-largest e-commerce platform; 100 million WeChat Mini Program users |
| Zynn | Risk Warning (no social graph, high customer acquisition costs) | Topped App Store charts; users earn $1–$1.50/hour; $20 for inviting a friend |
| Kuaishou | Neutral (focus on differentiation from ByteDance) | Valuation of $30B; raised $3B in December 2019; over 50% of revenue from live streaming |
| Facebook / Instagram | Neutral-to-Negative (feed ad model facing ceiling) | U.S. user growth has stalled (1%–2%/quarter); U.S. ARPU of $186/year |
| Risk Warning (anonymous identity limits ad revenue) | No specific data disclosed, but ad targeting capability is weak |
1. “Quality growth matters more than speed for social networks” (Turner Novak) — Facebook was once only the fifth-fastest-growing social network, but built the deepest moat through real identities and user-created advertising profiles.
2. “TikTok is the first scaled social product without a social graph — it is essentially a media company” (Turner Novak) — TikTok’s core product is algorithmic recommendations, not follower relationships. Users can get a personalized experience without “following” anyone, completely eliminating the initial friction.
3. “For Instagram to truly compete with TikTok, it must remove the following graph — but that goes against its DNA and it won’t do it” (Turner Novak) — Existing social giants’ counterattack against TikTok is constrained by their own product logic. TikTok’s moat lies in “doing something that incumbents cannot do because of legacy baggage.”
4. “Pinduoduo is the first e-commerce company designed entirely for mobile — no search function; users open the app and see algorithmically recommended products” (Turner Novak) — This aligns with TikTok’s “search-less” logic. Advertising revenue is the main source (commission rate ~0.6%), and social virality (group buying) achieves zero-cost customer acquisition.
5. “The Zynn/Kuaishou model proves that a media company without a social graph is just ‘the next-generation TV network’ — it can become a $10–$50 billion company, but not a Facebook-level $500 billion company” (Turner Novak) — The real moat comes from friend-driven UGC retention, not cash-incentivized purchased content.
6. “Snapchat’s Snap Minis could be the starting point of a US version of WeChat’s ecosystem” (Turner Novak) — Snapchat’s North American DAU is close to Instagram’s, but the advertising revenue gap is enormous; Bitmoji is already integrated into Samsung phones, with cross-platform distribution capabilities.
7. “Facebook’s advertising product is ‘one of the best products ever’ — 8 million businesses run automated ads with near-zero marginal cost, but this also becomes a drag on its transformation” (Turner Novak) — Facebook’s ad revenue is a cash machine, but US user growth has stagnated for nearly a decade, and the product experience is deteriorating (“users feel ‘too many ads, friends are gone’”).
8. “To evaluate a new social network, the key is whether it captures an ‘emerging behavior that incumbents cannot replicate due to DNA issues’” (Turner Novak) — TikTok’s “no social graph + mobile-first” is a textbook case; Pinduoduo’s “no search + social group buying” is another.