Hi, I’m Pavlo. Thanks for subscribing to my newsletter, where I interview people doing cool things in tech—entrepreneurs, product managers, and VCs.
At the beginning of our first meeting with Sameer, I somehow spilled an entire glass of orange juice on my pants and had to sit like that for the next hour. It felt too embarrassing to admit that my pants were soaked, so I just decided to carry on with the meeting, hoping they would dry. They kind of didn’t.
Sameer is a network effects expert. Not only does he teach network effects at Reforge, but he also invests in network-effect-powered projects on behalf of Speedinvest.
Go ahead and follow his blog if you’d like to learn more about network effects.
I was lucky enough to learn quite a bit from him!
Here are my takeaways.
What Are Network Effects?
The Basics: Network effects happen when every new user makes a product better for everyone else. Simple, right? But here’s the twist: it only works if the product creates a unique, structured way for users to interact. If it’s not unique or structured, you’re just competing with someone else’s network.
Great example of uniqueness: Airbnb’s supply side and listing structure, which have more parameters than Booking.com.
The Four Types of Network Effects
Data Network Effects: More users = more data = more value. Think Waze or TripAdvisor, where your input helps everyone.
Interaction Networks: Users exchanging info directly, like WhatsApp or GitHub.
Marketplaces: Buyers and sellers making deals, like Airbnb or Amazon.
Platforms: Developers building apps for users, like iOS or Shopify.
Common Misunderstandings
The Zoom Myth: Zoom feels viral, but it’s not a real network effect. Adding users doesn’t actually make Zoom better for other users—it’s just easy to share.
Asset-Heavy Marketplaces: If a company owns its inventory (like Opendoor or Gorillas), it’s not a marketplace—it’s just a fancy store.
How to Scale and Measure Network Effects
Scaling Hacks:
Fix annoying friction points. Amazon, for instance, uses AI to write product descriptions—because who has time for that?
Borrow early audiences, like Airbnb posting on Craigslist.
Long-term growth? Deliver such a good experience that people can’t stop talking about it.
Metrics That Matter:
For social networks: How often are users inviting or engaging with others?
For marketplaces: Look at transaction frequency, search-to-fill rates, and how much available supply gets used.
Challenges in the AI Space
AI’s Weak Spots: Generative AI is cool and all, but most applications don’t create strong network effects or lock-in (switching costs).
What’s Next: There’s potential in AI-powered interactions, but it’s still early days. Something like “AI 2.0” that blends AI and multiplayer interactions could be the real game-changer—but we’re not there yet.
Takeaways for Founders
Be Obsessed with the Problem: Building networks is no walk in the park. It’s going to test your patience, your confidence, and probably your sleep schedule. If you don’t deeply care about the problem you’re solving, it’s easy to give up when things get tough.
Defensibility is the Name of the Game: If your product doesn’t have a solid moat—whether it’s network effects, economies of scale, or switching costs—you’re basically inviting competition to crash your party. Think of defensibility as your product’s secret sauce that no one else can replicate.
Metrics Matter, But the Right Ones: Forget about vanity metrics like “growth at all costs.” Early on, it’s all about figuring out if users are really engaging with your product and if it’s actually solving their problems. For a marketplace, for example, are users coming back? Are transactions happening? Those are the signs you’re onto something good.
Tune in!
Ask
I hope you enjoyed the podcast! Who else would you like me to record with? Feel free to share any suggestions in the chat.
Share this post