Uber’s Greatest Trick Revealed

The greatest trick Uber ever pulled was to convince regulators that it is an open technology platform. That trick may no longer work if the California verdict that some Uber drivers are employees is upheld (I am assuming that the company will appeal the decision).

For disclosure, USV has two related investments with Hailo and Sidecar. Both companies have struggled and to no small degree because they did not pull the same trick. Hailo focuses primarily on the licensed market where rates are set by regulators and Sidecar offers a marketplace in which drivers can set their own rates.

What makes Uber so phenomenally successful is that it provides its service in the best possible way from a customer perspective: push a button and have a car show up. This is clearly the experience that the bulk of customers want and accounts for Uber’s impressive growth around the world.

But to make that experience possible, Uber had to operate not as a technology platform on which a market naturally occurs (the way it does on say eBay or Etsy) but rather as a transportation company that extremely pro-actively manages its supply of drivers. Uber has done that in many different ways, from paying drivers large guarantees to offering drivers leases for their cars and abstracting pricing and dispatch away from the end user experience (with Uber servers making pricing and dispatch decisions). Not surprisingly the times when the abstraction broke, eg during surge pricing, it led to the biggest objections by customers.

What the California decision reveals is that Uber has been trying to have it both ways: offer customers a transportation experience but be regulated as an open technology platform with independent drivers. One way that Uber could begin to address this issue would be to offer both a driver and a rider API that third parties can integrate without requiring Uber’s upfront permission. In particular a driver API would allow the emergence of apps that optimize on behalf of the driver and make it easy for drivers to participate in several networks simultaneously. Having APIs and letting more activity take place on Uber’s network in a less controlled fashion will make it harder to maintain take rate and to offer the same smooth customer experience but it would make Uber more genuinely a technology platform company. 

It will be exciting to see where this all goes and I very much hope that in the end we wind up with not one but several competing technology platforms that make transportation more efficient and accessible. With the right platforms underneath, many different experiences can co-exist in the market including some that feel more like a marketplace and others that provide a higher service level. The same set of questions and issues will have be settled in a variety of other on demand marketplaces, including last mile delivery.

PS Another way that this could eventually get resolved (but is a long time off) is by having a Universal Basic Income. A UBI that covers basic needs would give participants in on demand labor marketplaces a stronger walk away option and hence more bargaining power. It would allow us to transform the economy more into a loose networks of freelancers and away from traditional employment relationships – this is happening in any case but currently with a power imbalance too much in favor of the networks.

PPS Instead of Uber offering the APIs themselves if there was a right to an API key, or a right to be represented by a bot, drivers could “hack” the platform themselves. If you want to learn more about those ideas, watch my TEDx talk.