Why Uber Is Making Its Biggest Bet Yet on Carpooling in China

Travis Kalanick stands in front of China. Well, a map of China.

China Stringer Network/Reuters

Uber has never been shy about making big bets, and so, in true Uber form, this week it made another huge one. It’s called UberCommute and is described on the company’s blog as “carpooling at the press of a button.” UberCommute is designed to pair drivers with riders traveling along similar routes. The basic vision is that a driver heading to work, or running an errand, or otherwise traveling in a given direction will log into the app and pick up a passenger heading the same way. In a first for a to-be global product, Uber has chosen to develop and pilot UberCommute not in the United States but in Chengdu, China, which it calls its biggest city.

Uber launched in China in February 2014 before setting up shop that August in Chengdu, the main urban area of which has a population of 4 million to 5 million, depending on whose data you’re citing. This June, Uber CEO Travis Kalanick disclosed in a letter to investors that all of Uber’s Chinese cities—but in particular Chengdu—were growing at a truly fantastical rate. After its first nine months on the market, Kalanick wrote, Chengdu alone was doing 479 times as many trips as New York City had at the same age. Yes, 479.

Combine those Chengdu population estimates with that towering growth figure and you get a city with a ton of liquidity, which for a product like UberCommute is a big deal. Most of Uber’s services depend on what the tech and venture capital communities lovingly refer to as “network effects,” i.e. the principle that the more people use something, the more valuable it becomes. This is presumably one explanation for why Uber has been so aggressive in its expansion—as the number of people riding with and driving for Uber increases, so does the strength and stability of its platform, in a sort of perpetual supply-and-demand feedback loop.

Anyway, getting back to Chengdu. Since UberCommute is being conceived of first and foremost as an “entirely new product for drivers,” the market’s liquidity is especially important. The crucial inversion of UberCommute is that instead of supporting riders who are going somewhere and need a driver, it’s designed for drivers who already have a destination in mind and are willing to bring along a passenger. On the surface it might sound very much the same, but in practical terms it means that the drivers on UberCommute are going to have a much lower tolerance for going out of their way to pick someone up than the typical Uber driver looking for a fare. What exactly that tolerance is, Uber doesn’t know yet. What the company does know is that for UberCommute to work, it will need a ton of drivers and a ton of riders, so that drivers who may not want to deviate very far from their paths are in fact able to find a passenger. Hence why liquidity is key.

Uber currently seems to have two strategies for ginning up sufficient interest in UberCommute. The first is playing up in promotional materials how UberCommute promises to improve the well-known congestion problems in Chinese cities, providing a “real alternative to a world that looks like a parking lot and moves like a traffic jam.” The second is making UberCommute ultra-ultra-cheap. I say ultra-ultra because People’s Uber, the UberX-like platform that Uber offers in China, is already heavily subsidized by Uber (on an almost unfathomable scale) and therefore ultra-cheap. UberCommute will be even cheaper—and so for Uber, which currently isn’t taking a commission on the service, in the short run even more expensive. But in the long run, if UberCommute catches on, it could also be ultra-ultra-successful.