Uber and the delicate business of creating a platform

[This post originally appeared in HBR blogs on 24th August 2012]

Over the last year, a new app has been changing the way people get limos in some major cities in the US. Uber — founded by Garrett Camp, Oscar Salazar, and Travis Kalanick, and launched in 2010 in San Francisco — allows people to get and pay for taxi rides easily. Here’s how it works: after you have signed up for an Uber account, you can launch the smart phone app and instantly find limos or town cars equipped with the service. One click hails them, and the nearest driver comes and picks you up. At the end of the journey, your driver charges your account. No standing in the rain trying to hail a cab. No grabbing for cash or fiddling with credit cards.

Uber has positioned itself for the time-sensitive rather than price-sensitive consumer. Its rates are higher than you would find for regular taxi, but the fast response drives value for consumers. On the technical side of things, Uber works well. Its challenge now is to get both customers and drivers signed up. But doing so involves a delicate balancing act, one faced by every startup that fancies itself a “platform.”

A tirade of a blog post by Bill Gallagher this week shows just how delicate the balancing act really is. Taxi rides have fees that vary with the length and time of the journey. People are used to that. But Uber, to encourage more taxis into a busy area, has implemented what economists call “dynamic pricing.” With dynamic pricing, when supply gets tight, prices rise. We see this all the time with hotels and airlines. Uber implemented it to reward those drivers who joined the service; drivers who service busy areas get paid more. In that way, dynamic pricing also helps to direct drivers’ attention to where they are needed most, which helps users. If you want to ensure that supply for Uber users isn’t tight when it is for everyone else, you pay suppliers more during busy times.

Put that way, it all sounded reasonable. But for consumers like Gallagher, it meant confusion. Gallagher, for instance, saw that his airport ride might cost him 1.5 times the usual amount, but there was a miscommunication with his Uber driver, who misquoted the price. Uber has tried to be responsive to Gallagher’s expression of ire, but one gets the sense that in this one case there was some fragility in the customer relationship.

If this were a one-time mishap, we could set it aside. But sticker shock has actually turned out to be a ongoing issue for the company. This was demonstrated most clearly in New York on New Year’s Eve, a night that should have been a triumph for the start-up. Because of high demand and the accompanying ever-higher prices, Uber’s supposedly simple and straightforward transaction became complex. Consumers saw that their ride home could be seven times the price of their ride into town. Consumers who were used to easy clicking missed the notice and became (reasonably) upset when the bill arrived. Either way, they weren’t happy. Basically, to satisfy one side of its market — the taxi drivers — Uber upset the other side — its customers.

One possible reaction to this issue is to conclude that dynamic pricing is a pipe-dream, fundamentally inconsistent with what consumers want. That, however, would be short-sighted. After all, if consumers want timely service, especially during peak times, surely they will have to pay for it. Uber doesn’t own the cars, so it must rely on the market to procure them. That means that when the competition heats up, prices will rise.

The problem is not that the consumers paid but how they paid. It was the surprise over the unexpectedly high price that disrupted the relationship. In light of this, Uber could offer consumers the option of an account with some price assurance. One extreme would be that the consumers’ prices were fixed but slightly higher than the average dynamic price. Uber could still pay taxis more when there was peak demand but the cost of that would be spread over consumers at all the times they grab cabs. Basically, Uber would offer their consumers insurance over a fluctuating cab price and then use the premiums to pay for the peak procurement. More to the point, this option would allow the company to shield consumers to the full brunt of their supply management challenges.

Such an approach would require a more delicate balancing act on Uber’s part. It cannot rely on consumers to self-mediate the peak times. But then again that is not its mission. To encourage intensity of use, which is where Uber makes its profit, it needs both consumers (who value simplicity and certainty) and taxi drivers (whose other option is to go it alone). That means it has to offer each side what it wants even if it results in a more problematic management and financial balancing act for the company itself. But that too will rely on having lots of customers with diverse preferences.

Uber will succeed or fail based its ability to generate a virtuous cycle — which is why platform establishment is easier to say than to do.

(For my part, I’ve used Uber several times and really like not only the convenience but the beautifully designed receipts.)