Uber CEO Predicts Profitability in Three Years. Don’t Count On It.

But comments on data sharing could be positive

Uber CEO Dara Khosrowshahi attended the annual elite conference in Davos, Switzerland earlier this week and made a bold claim about when he expects Uber to achieve profitability. In an interview with Bloomberg, he said he expects the company to be in the black before 2022 and that the core ride-hailing business should be profitable within three years. However, there are a lot of reasons to be skeptical about that timeline.

Uber’s losses for the 2017 fiscal year are on track to hit $5 billion, a significant jump from its $2.8 billion loss in 2016. The company’s operations are being subsidized by venture capitalists who hope that by offering low fares for long enough that Uber will achieve a dominant market position that will then allow it to raise prices and finally earn enough to cover its costs. But given that Uber’s cash is running low, it needs to show progress in the very near future while it’s still facing competition from other ride-hailing companies and other modes of transportation.

The company’s other means of achieving profitability is to automate its drivers, and Khosrowshahi seems to recognize this. He used to be more cautious about self-driving cars, but now expects Uber to have them on the road within eighteen months in a very limited capacity. The limitations are necessary because it appears that Uber will use level-4 vehicles, similar to Waymo, meaning the system will not be able to navigate every possible driving scenario and will retain its steering wheel. He gave Phoenix, Arizona as an example of where they will likely operate — a city with clear weather, plenty of suburban areas, and where Waymo recently launched a very limited driverless taxi service — acknowledging that the vehicles will only work in five percent of use cases in the beginning, but will be able to handle more situations the longer they’re on the road.

The big question is whether Khosrowshahi’s timeline checks out. Could Uber potentially have a level-4 service in eighteen months? It is possible. Waymo is already running one, yet it’s important to remember that most companies working on autonomous vehicles are pushing their timelines back to at least 2021 and promising only semi-autonomous features at that time. In addition, a recent assessment of the state of driverless tech put GM, Waymo, Ford, Daimler-Bosch, and Volkswagen in the lead, while Uber, Tesla, and Apple lagged far behind.

Source: Navigant Research

Given that companies seem to have misjudged their driverless vehicle timelines and Uber’s tech is among the worst — Lyft, by contrast, has partnered with Ford — the company may be able to begin limited use of self-driving cars in eighteen months, but it seems highly unlikely that they’ll reach the necessary scale in three years to eliminate enough drivers to reduce costs to the degree necessary to achieve profitability. This is compounded by the fact that Khosrowshahi is trying to give Uber a more positive image by claiming to work more closely with authorities and implementing policies designed to treat its contractors a little better — though nothing like employees — such as a new initiative in Washington to contribute to a benefits plan for gig workers, which labor groups are opposing. It should go without saying that these moves will cost the company more than Kalanick’s move-fast-and-break-things style.

And if you need anymore reason to question Khosrowshahi’s timeline, remember that he’s also promising flying cars in 5–7 years — as if we haven’t been waiting for those, as a society, for the better part of a century.

Flying car ad from 1958. Source: Smithsonian

While his timeline to profitability may be dubious, Khosrowshahi did make a potentially positive statement that’s worthy of note. Uber, like most on-demand platforms, has been known not to cooperate with local governments, but as part of his effort to improve the company’s image, Khosrowshahi said that Uber will share its data with governments to try to improve congestion. This could be good, depending on how it’s shared. Uber already provides some data to governments, but tends to require them to keep it secret, which runs counter to norms of government transparency.

If Uber wants to operate on public roads, it should be forced to give up its ridership data with no conditions attached, but it remains to be seen whether that’s part of Khosrowshahi’s new spirit of cooperation. However, that may not be an issue for much longer, as a three-year path to profitability still seems like wishful thinking with the current state of autonomous tech.