Uber and Lyft are spending big money to ramp up their electric scooter businesses even as they face rising scrutiny about their lack of profits.

They’ve hired big teams, bought thousands of scooters, and are continuing to invest in creating their own scooter designs. But despite disappointing recent initial public offerings and super-sized losses, the two companies are resisting pressure by investors to cut costs. In fact, they’re continuing to push into new cities with their scooters, betting that the costly initiatives will eventually pay off down the road.

The topic is likely to come up on Thursday during Uber’s first earnings report as a public company. Lyft, which reported earnings on May 7, said it would continue to expand its scooter business.

Both companies, along with startups Lime and Bird, are hoping to dominate what many expect to become a big scooter rental market. Overall revenue for scooter rentals in the U.S., Canada, Latin America, and Europe is expected to grow 21% by 2023 to $5 billion, according to research firm Morningstar.

Today, Uber and Lyft collect negligible amounts of revenue from scooters. It’s so small that Uber doesn’t disclose it while Lyft has said that it’s immaterial, financial speak for next to nothing.

Both companies declined to comment to Fortune about their scooter businesses.

The scooter rental business is proving to be a difficult one for all the companies in the space. Scooters often break down or are vandalized. They also require charging and regular maintenance. Meanwhile, companies must navigate a hodgepodge of city regulations to get their scooters onto sidewalks.

Tom White, an analyst at D.A. Davidson, said scooters make sense for residents in cities where owning a car is too expensive or where public transportation is limited. But turning that into a profitable business is another matter considering the high cost of managing thousands of scooters, he said

But Kersten Heineke, leader of the McKinsey Center for Future Mobility in Europe, which researches autonomous, connected, electric, and shared vehicles, argued that today’s money-losing reality will change in the future. By sometime next year, after two warm-weather seasons of scooter wars, he expects the industry to consolidate to about a handful of strong players.

“When we get to the third season, which is next year in the U.S., we’ll begin to see which business models are profitable and surviving,” Heineke said. “It will be a market of five-to-seven players per region or country.”

The keys to success are scooters that can withstand a beating and strong partnerships with public transportation agencies, transportation companies, and cities, Heineke said. Streamlining operations, including maintenance and managing fleets of scooters economically, is also critical.

Uber, which had 91 million users for its ride hailing and delivery service last year, and Lyft, which had 31 million during the same period, have a leg up considering their huge numbers of existing customers. Plus, both companies have existing partnerships with transit agencies for providing services like free or discounted car rides for people traveling to transit centers that could help them market their scooters.

When it comes to designing scooters, Uber and Lyft are investing a lot of money. Like their competitors, they’re trying to create designs that can withstand harsh weather, vandalism, and inevitable wear and tear of the roads.

Neither Uber nor Lyft disclose how much they spend on scooters. But they do report broader categories of expenses that include scooters.

Lyft said it invested $68.7 million on “property, equipment, and scooters” in 2018 plus an additional $25.4 million in the first quarter this year. Meanwhile, Uber, which operates its scooters under the brand Jump, said it spent $558 million in 2018 on the “purchase of property and equipment.”

Scooter-only competitors like Lime, in which Uber has invested, and Bird, also have been upgrading their scooters. Lime is rolling out its fourth iteration, while Bird is on its third.

Both say the increased durability of their newest generations of scooters have extend their lives well beyond the time it takes to recoup their costs. An average $400 scooter could break even on costs, which include city permits and maintenance, in less than four months, according to McKinsey.

Bird’s latest model has an expected life span of one year, an improvement over its previous model, which survived 10 months on average. Both models deliver enough revenue to cover the cost of the scooters, said Bird CEO and founder Travis VanderZanden.

Similarly, Lime said its newest generation of scooters lasts five times longer than its first model. It did not specify exactly how long that is.

“We’re already profitable in some cites,” said Ted Tobiason, Lime’s chief financial officer. Still, he acknowledged that the scooter rental business is tough and that skepticism about it is natural.

“People who say that it’s very hard to make money in scooters, and it’s a very challenging business are right,” Tobiason said. “If all this business was about was buying a bunch of scooters and having an app, it wouldn’t be that interesting.”

A big part of success depends on one thing, Tobiason said: data. Companies that excel at using information they collect about where and when their scooters are used to make their businesses more efficient have a big advantage.

Many early investors in scooter companies argue that scooters are no different from any other kind of early-stage, high-growth startups. Initially, they spend a lot of money to get customers. But at a certain point, the financial side catches up and the companies no longer bleed cash.

Miles Clements, partner at venture capital firm Accel, which funded Bird, said he’s excited to be investing in “an enormous category with unmet consumer demand.” But the winners of the category will have to master the complications of creating and managing fleets of scooters—all while keeping operational costs down.

“If you can accomplish those things, the market is enormous,” he said.

~source