Platforms face unique constraints that make them very sensitive to friction — this limits their monetization options to just 3 out of the 6 models used by other types of networks — Platforms share many characteristics with marketplaces. They create value by connecting demand with supply — specifically, they connect users with app developers. So it is natural to assume that platforms have the same monetization options available to them as marketplaces do. …

Data networks can use five out of six possible monetization models, depending on the way they acquire data from users and the way that data is consumed — Data networks are unique within the world of network effects. Most network types create value by allowing participants to interact with each other in some way. Data networks, however, do not connect participants directly. Instead, they crowdsource data from participants to improve the product for all of them. This has…

Marketplace monetization depends on two factors — Transaction complexity and the asymmetry in value to the demand and supply sides — Like other types of startups built on network effects, marketplaces create value by connecting participants. Specifically, they connect demand with supply to enable transactions. This gives them an obvious way to monetize — take a cut of every transaction. However, this cannot be blindly applied to all marketplaces as there…

Interaction networks can use one or more of six possible monetization models — the choice between them depends on the structure of their network — Startups succeed by uncovering a unique insight to create value for their users. This value creation is only sustainable if they can find a way to capture some of it themselves, i.e. monetize. This is just as true for startups built on network effects. However, they are more complex to…

Generative AI has its first platform, but what kind of platform? The answer depends on how ChatGPT solves the “cognitive load” problem. — My last essay about the potential of network effects in generative AI (on the Speedinvest blog) attempted to seperate AI hype from reality. Despite breathless discussions of a “platform shift”, foundational models like GPT-4 and Stable Diffusion did not have the characteristics of a platform…

Get answers in less than 1 minute — When I started writing Breadcrumb.vc in 2020, my goal was to create a set of mental models that founders, operators and investors could use to evaluate the potential of network effects. Of late, I’ve been trying to productize these frameworks to make them more accessible. My first attempt at this…

Tokens can be an effective way to bootstrap networks that need passive user participation, but they can be counterproductive for those that need active participation — Crypto tokens introduce financial incentives to technology products. The startup and web3 ecosystem are still working out the implications of this, including how and why this can be useful. During this period of experimentation, one theory appears to have gained some steam — tokens are a way to incentivize early…

Instagram is the new Facebook: Increasing UX complexity and a shift from social interactions to broadcasting makes Instagram vulnerable to a new wave of “unbundlers” — Facebook is among a handful of truly generational internet companies. Its journey to over $1T in value started with a simple idea: a social network based on actual identities and real-world social connections. …

There is a long list of metrics that attempt to measure network effects — putting them in context is the best way to identify the most relevant KPIs for each situation — Metrics (or KPIs) are among the most well-researched topics within the startup ecosystem. This is just as true for those built on network effects. There are great resources available from A16Z, Point Nine Capital, and Speedinvest Pirates just to name a few. The goal of this post is not to…