Uber's core technology is the marketplace: software that predicts demand, matches a request with a nearby driver or courier, routes them, prices the trip and settles payment, across more than 15,000 cities. It reports three segments built on it.

• Mobility: rides in cars, taxis and on two wheels, and, through partners such as Waymo, robotaxis. A rider sees the price up front, a driver accepts, the app does the rest. Mobility also includes advertising during trips and financial partnerships.
• Delivery: Uber Eats, for meals and, increasingly, groceries, alcohol, convenience and retail, plus Uber Direct, which delivers for merchants' own channels, and advertising for merchants.
• Freight: a digital marketplace that matches shippers' loads with carriers, plus managed transportation and logistics software from the Transplace acquisition.

Mobility and Delivery share drivers, consumers and technology, and that is the design. A driver can carry a rider at rush hour and a meal at dinner, so the network stays busy; a consumer with one account and payment method can ride and order in the same app. Uber's filing says a majority of first-time Delivery customers were new to Uber, and that people who use both spend several times as much as people who use one. Uber One, a membership with delivery perks and ride credits, is the glue.

Drivers and couriers are independent contractors who switch on when they want; Uber owns no cars and no kitchens. That keeps its capital spending small, and makes its relationship with drivers (their pay, their classification, their safety) the central question of regulation in nearly every country it serves.

Consumers pick Uber for coverage and reliability: a car comes quickly in most cities and airports, the same app works in other countries, and food and groceries are one tab away. It competes with Lyft, Bolt, Didi and taxis in rides, DoorDash, Instacart and many local apps in delivery, and C.H. Robinson and other brokers in freight.

For a rider or an eater, switching costs almost nothing: Lyft or DoorDash is a download away, and many people keep several apps and compare prices. What keeps them is that Uber is usually the fastest and most available option, and that Uber One makes the second service cheaper once they pay for the first. Saved places, payment methods, a business profile for expenses, trip history and the habit of opening one app add a little more.

Drivers and couriers use several apps too, often running Uber and Lyft or Uber Eats and DoorDash at once. They stay where the requests are, and requests follow drivers: more drivers mean shorter waits, which bring more riders, which bring more drivers. That loop is why the largest network in a city tends to stay the largest, and why Uber left China and Southeast Asia, where rivals had more of it.

Merchants stay because Uber Eats brings orders they would not otherwise get, though most also list on DoorDash and others. Freight shippers move between brokers easily, which is part of why that business is harder.

The long-run alternative may not be another app at all: robotaxi operators such as Waymo and Tesla running their own services. Uber's answer is to put their cars on its network.

Revenue was $52.0B in 2025, up 18.3%, after 18.0% in 2024. Because in most markets Uber records only its fee rather than the full fare or order, revenue is a fraction of what consumers spend on the platform.

The business has turned around. Uber lost money from operations every year on file until 2023: an operating loss of $4.86B in 2020, 43.7% of revenue, and still $1.83B in 2022. Operating income was $1.11B in 2023, $2.80B in 2024 and $5.57B in 2025, a 10.7% operating margin, as trips grew faster than costs. Net income of $10.05B in 2025 was lifted by a one-time tax benefit and moved by changes in the value of the investments Uber carries in companies such as Didi and Aurora, so operating income is the cleaner read.

Owning no cars keeps capital needs small: capital spending was $336M in 2025, under 1% of revenue. Operating cash flow was $10.10B and free cash flow $9.76B, 18.8% of revenue. Equity pay was $1.83B, 3.5% of revenue, low for a technology company, so free cash flow after equity pay was $7.94B, 15.3% of revenue. Uber ended 2025 with $7.63B of cash and short-term investments against $10.52B of debt, net debt of $2.89B.

The largest open question is autonomy. Uber sold its own self-driving unit in 2020 and now partners instead: Waymo's robotaxis run on Uber in Phoenix, Austin and Atlanta, and Uber has deals with other developers in the US and abroad, including an agreement to buy, or have fleet operators buy, Lucid vehicles fitted with Nuro's driving system. Uber's position is that robotaxi makers need a network that already has the riders, the demand forecasting, the pricing and the fleet operations, and that a mixed fleet of human drivers and robotaxis covers demand peaks better than robotaxis alone. The risk is the reverse: operators such as Waymo and Tesla running their own apps and keeping the riders.

Elsewhere, the direction is breadth on one account. Delivery is pushing beyond restaurants into grocery and retail, Uber One into more countries, and advertising into more of the apps and trips. Uber Direct turns the courier network into a service for merchants' own channels. Freight is waiting out a weak market while selling more software and managed services.

With the business at a 10.7% operating margin in 2025, the next phase is less about whether the marketplace can earn money and more about whether autonomy changes who holds the rider.

• Autonomy could cut Uber out. If robotaxi makers such as Waymo and Tesla scale their own apps, they would hold the riders and set the prices, and Uber's driver network, its main advantage, would matter less.
• Drivers' status and pay. Courts and lawmakers in many countries keep revisiting whether drivers are independent contractors, and rules on minimum pay and benefits raise costs market by market.
• Competition that never stops. Lyft in the US, Bolt and Didi abroad, and DoorDash in US delivery compete on price, and switching costs nothing for consumers or drivers.
• Regulation and trust. Safety incidents, data privacy, antitrust, airport access and the FTC's case over Uber One sign-ups and cancellations each bring costs and limits.
• Freight's cycle. Brokerage is low-margin and its volumes swing with the freight market, which has been weak.
• Prices. Riders and eaters already find Uber expensive, and the fees that pay drivers and couriers leave little room to raise prices further.