Why Uber has to be first to market with self-driving cars

The race to get the first network of self-driving cars on the road is off to a lukewarm start.

Uber recently launched a limited test of its self-driving cars in Pittsburgh as part of a partnership with Volvo. Singapore-based self-driving startup nuTonomy launched its own limited pilot a few days before that and struck a partnership with the Southeast Asian ride-hail company Grab.

It’s all exciting — how can robot cars at your beck and call not be? — but it doesn’t yet mean much.

The true test of the viability of the self-driving technology will come down to which company will launch a fully operational network of self-driving cars, and when.

While it would be advantageous for any of the companies attempting to develop a shared network of self-driving cars to be first or near-first to market, it’s a near necessity for Uber.

Valued at a lofty $70 billion, Uber has a lot more at stake than its competitors. As Uber CEO Travis Kalanick said in an interview with Bloomberg, developing self-driving cars is “basically existential for us.”

That’s because the second another ride-hail company or automaker launches its own network of self-driving cars, the market share Uber has poured time and money into dominating is at risk.

Without that market share, Uber may not be able to provide the returns on investments the company’s portfolio of backers expected. And without those returns and that scale, it’s unlikely Uber can uphold its valuation.

The key to operating a ride-hail service efficiently is having enough drivers to meet rider demand. If a competitor — let’s use Lyft in this example — rolls out 100 self-driving cars in a city, a decent portion of the market share Uber spent countless time and money attempting to capture is now up for grabs.

A self-driving car could easily perform 100 rides a day (around four rides an hour) while Uber drivers typically perform somewhere between one and two rides an hour, depending on the city, and can only drive up to 12 hours a day. At most, Uber drivers would typically be doing 24 rides a day.

That means 100 Uber drivers would only perform at most 2,400 rides a day compared with a self-driving network that could turn 10,000 rides a day. That’s without accounting for the fact that Uber can’t tell its drivers how long they have to drive. In fact, Uber takes pride in how flexible driving is. As of 2015, 52 percent of Uber drivers were part-timers, according to a study the company commissioned.

With Uber’s valuation, holding on to that market share is crucial to its survival. Especially since Uber has lost $1.2 billion pursuing its ambitious goals in just the first half of 2016, according to reports.

While the company’s decision to pull out of China certainly helps its bottom line, it’s important to find new revenue streams as the company continues to face off against strong regional players like Grab and Ola. But being able to generate profit from existing revenue streams becomes harder in the face of subsidy wars.

Not to mention, Uber can’t just keep raising money if the company wants to preserve the value of its investors’ equity.

Driverless cars are one of the company’s primary ways out of that predicament — the holy grail, if you will. Replacing human drivers with robot cars dramatically increases the company’s profitability while reducing costs, and not just because the cars can perform more rides than human drivers.

Today, Uber pays drivers 65 percent to 80 percent of each fare, so for every dollar a driver brings in, Uber only takes home 20 to 35 cents. Eventually, when drivers are replaced by robot cars, Uber could capture close to 100 percent of the fare.

But if, instead, another company sees that dramatic increase in profitability and drop in operating costs, it may and likely will hurt Uber’s IPO position and its valuation.

A lot has to happen before self-driving cars are truly a reality — from regulations to public support — and showing that Uber can not only successfully operate a self-driving pilot but an actual money-making self-driving network is important to its ultimate success.

That’s not to say that it’s not important for a competitor to go to market with a self-driving network quickly. Lyft, for example, has only raised more than $2 billion and was most recently valued at $5.5 billion, so losing any more marketshare to Uber is not ideal. But with the backing of a player like General Motors — which will likely front the costs of production and car ownership in a self-driving network — and a lower valuation, Lyft has more leeway to bide its time, particularly if GM ends up acquiring the company.

At the end of the day, it may have been easier for Uber to continue to operate as it does at the valuation its investors have assigned it had self-driving cars not become near-reality. But the looming threat of self-driving cars is becoming increasingly real, and it’s important for Uber to be the first to the finish line.

This article originally appeared on Recode.net.