Tesla (TSLA -2.84%) is planning to launch its own robotaxi service in Austin this summer. The service will use a fleet of Tesla-owned vehicles employing the first version of its unsupervised Full Self-Driving (FSD) software.

However, Tesla won’t be the only autonomous vehicle ride-hailing service in Austin this summer. Uber (UBER -2.07%) riders may also find themselves inside one of Alphabet‘s (GOOG -2.07%) (GOOGL -1.92%) Waymo vehicles thanks to a partnership between the two companies. And Uber CEO Dara Khosrowshahi has a warning for Tesla and its plans to go it alone.

“Demand often is quite variable,” Khosrowshahi said at a recent conference. That’s why even a company that can afford to sink billions of dollars of capital into building an autonomous vehicle fleet like Alphabet has opted to partner with Uber in several cities. Uber believes it has already solved the biggest challenges of bringing an autonomous vehicle service to market.

Much of Tesla’s current stock value depends on getting autonomous driving and ride-sharing right, according to multiple Wall Street analysts. So, Tesla could have a make-or-break moment if it does finally launch autonomous vehicles this summer.

Image source: Tesla.

The biggest challenge in launching a ride-sharing service

It’s very easy for Tesla to put a fleet of vehicles in service and make an app for consumers to call for a ride in one of those vehicles. But there’s a lot more to successfully launching a ride-sharing service than that.

First of all, Tesla needs to drive demand for the service. Second, it needs to consider supply-side challenges like pricing, fleet management, and even routing the vehicles to make them broadly available to users in any given service area.

Fleet management may be the riskiest factor. Cars aren’t cheap, but if Tesla wants to meet the peak demand for its service, that means large portions of the fleet will be sitting idle during off-peak hours. Tesla will have to invest significant capital upfront with the hope that demand eventually grows enough to meet supply.

The other side of the challenge is if Tesla doesn’t put enough vehicles in service and demand exceeds supply. Users will get frustrated by wait times and switch to competing services to find a ride. That app would be, in all likelihood, Uber.

Meanwhile, Uber already has 171 million monthly active users on its platform. That far exceeds similar apps, and it’s growing faster than smaller companies, too. Second, Uber can ensure high utilization rates for a small autonomous vehicle fleet by supplementing those vehicles with human drivers and dynamically changing pricing to balance supply and demand.

Uber will also take care of routing, pricing, and customer support. Uber charges a fee for those services, but the value it provides is arguably worth the price.

Should Tesla investors worry?

There are a couple of considerations here for Tesla investors.

Khosrowshahi isn’t an unbiased commentator. It’s no surprise he thinks partnering with his company to launch autonomous rides in a new city makes more economic sense than going it alone.

If Tesla is willing to invest the necessary capital to put a fleet in service that can meet peak demand, it could pay off in the long run. Elon Musk is certainly good at driving demand for his products, and there’s no shortage of Tesla fans who can’t afford to buy a Tesla but would certainly pay a few bucks to take a ride in one every now and then.

But there’s significant risk in the approach, and perhaps more than many investors are taking into account considering the stock price. Shares currently trade for more than 100 times forward earnings estimates. That means there’s a high expectation that Tesla’s earnings will accelerate as its robotaxi service launches and scales.

Waymo is notably taking a much more conservative approach by slowly rolling out its ride-sharing service to new cities and testing different partnerships with Uber. In one city, it will sell the vehicles to Uber, which will be fully in charge of the fleet. In another, Waymo will own the fleet but put them in service through the Uber app. And in other cities, Waymo will offer rides directly through its own app (those cities are mostly those where it’s been testing vehicles for a long time). This approach will help it find the best economics for scaling its service.

Tesla’s approach is high-risk and high-reward, but that’s not new territory for Tesla shareholders.