Network Effects
"A product becomes more valuable as more people use it"
Origin: Robert Metcalfe (Metcalfe's Law) and economics literature
The Key Insight
Network effects are positive feedback loops, and biology shows that all positive feedback loops eventually encounter negative feedback. The question isn't whether you have network effects - it's whether you can sustain the positive loop and prevent the negative one.
What People Think
Network effects create winner-take-all markets. Once you have enough users, the network itself becomes the moat - no one can catch up because your product is more valuable simply due to scale.
The Deeper Truth
Network effects are a specific case of positive feedback loops, which biology calls 'autocatalysis.' The same dynamics drive ecological 'tipping points,' epidemic spread, and the formation of biological networks. But positive feedback isn't inherently stable - it can flip. Networks can unwind as fast as they built (see: Facebook among teens, Twitter/X after management changes).
Biological Parallel
Mycorrhizal networks in forests show biological network effects - trees connected to larger fungal networks have better access to nutrients and information (chemical signals about pests). But these networks also show how network effects have limits: beyond a certain density, the network becomes fragile, and disturbances can cascade through connections. The same network that amplifies benefits can amplify damage.
Business Application
Network effects are real but often overstated. Key distinctions: (1) direct network effects (phone network) vs. indirect (marketplace platform), (2) local network effects (who you know matters more than total users) vs. global, (3) same-side vs. cross-side effects in platforms. Many 'network effect' businesses actually have weak effects that don't prevent competition. The question isn't 'do I have network effects?' but 'how strong are they, and can they flip?'
When It Breaks Down
Network effects reverse when: (1) the network becomes too noisy/crowded (social media algorithmic feeds), (2) a subnetwork becomes large enough to sustain itself elsewhere, (3) multi-homing is easy (users can be on multiple networks), or (4) negative network effects emerge (spam, low-quality content, trust erosion). MySpace had massive network effects - until it didn't.