Network Effects — Part II (Why Uber Fights)
A few weeks ago I shared Fred Wilson’s “dentist office story parable“, making the case for using network effects to create a defensible competitive advantage.
This following post by Ben Thompson in Stratechery is a fantastic in-depth analysis of how a unique network effect dynamics is manifesting itself in the ride-sharing market:
Here’s a summary of his analysis:
Ride-sharing in a single city:
- There are little to no network effects within each side of the market. The winner-takes-all dynamic is created when the two sides of the market interact:Company X has the majority of riders (more demand) –> Drivers will increasingly serve Company X’s customers (more supply) –> Company X’s service level improves (more supply liquidity) –> Even more riders to choose Company X.Company Y has fewer riders (less demand) –> Drivers faced with increasing cost to serve Company Y’s riders (more idle time, switching costs) –> Drivers will service other company’s riders instead (less supply) –> Company Y’s service level decreases (less supply liquidity) –> Even fewer riders choose Company Y.Just like in the CPG industry, given that purchases are habitual, prices are similar and the products are not highly differentiated brand allegiance is high, since the value of product comparison at the moment of transaction is fairly low. The only differentiating factor then becomes supply liquidity which gives the larger competitor a big advantage
Ride-sharing in multiple cities:
- This is a fairly commoditized market where first movers advantage is real, given the brand loyalty discussed above“Winning” other cities helps tilt the scale in your favor when trying to win a new city, through global/national brand awareness and (frequent business) travelers that are already loyal to your brand.
Tipping points: Major expansion of the market potential can be caused by:
- Getting and keeping supply liquidity at a high point when ride-sharing starts becoming the default choice for other transportation needs
- Transporting not just people
- I think network effects do exist within each side of the market, but perhaps not according to the formal definition that conforms to Metcalfe’s Law: in places with dense social networks (like cities) there’s definitely a viral component to building brand loyalty and both companies are incentivising existing riders to get new riders to start using the service. Furthermore, more recently both companies started experimenting with products that benefit from more conventional network effects (Uber Pool / Lyft Line).
- I share Ben’s concerns about Uber’s perceived lack of ethics especially in light of the monopolistic dynamics in the market that it’s in. Such dynamics is a textbook example of market failure, and in my mind the only real and valid reason why some form of regulation is required here. So far, it seems like most regulatory efforts have been centered around preventing ride-sharing services from disrupting existing incumbents, which feels like the wrong regulatory objective to optimize for.
- Some, but not all, of the patterns discussed above, exist in other “sharing economy” marketplaces as well, and can potentially lead to similar outcomes and risks.