The most valuable publicly listed companies in the world today—Apple, Microsoft, Amazon, Alphabet-Google, Facebook, Tencent, and Alibaba—share a common trait. They are platform businesses—that is, they bring together different market actors in order to distribute or exchange products, services, or information.
Alibaba and Amazon link retailers with customers. Google and Facebook connect advertisers to consumers. Microsoft, Apple, and Tencent (WeChat) own operating systems that bring together application developers and users.Uber, Lyft, WeWork, Airbnb, and other sharing-economy startups are competing to become the next blockbuster platforms. Some will succeed. However, for many, the bigger their platforms get, the more money they’ll lose.
True platform businesses are powered by network effects. The more people and organizations that use them, the more valuable they become, which encourages others to join, creating a self-reinforcing positive feedback loop. Amazon and Alibaba need not search for providers of goods for their e-commerce platforms; sellers and buyers come to them.
My colleagues and I recently compared the biggest 43 publicly listed platform companies to 100 of the largest firms in the same businesses over a 20-year period. They all had comparable annual revenues of about $4.5 billion. But the platforms generated nearly twice the operating profits, had market values more than twice as high, and were growing about twice as fast. The platform businesses achieved these sales with half the employees. But we also counted 209 companies that competed directly with these 43 companies and subsequently failed, suggesting that failure is far more common than success.
Take Uber, for instance. It has raised prices and cut staff, yet annual operating deficits of several billion dollars are expected to continue. Lyft loses less money because it’s smaller, but the ride-sharing platform faces the same problems.
Why do companies with such high demand lose so much money?
First, creating a platform in a “bad” business (an industry with low profit margins) does not make it a “good” (i.e., money-making) business. Platforms that sell digital goods, such as software, music, or advertisements, have close to zero marginal costs. Transporting people and goods, on the other hand, is expensive. Uber and Lyft spend a lot of money to find drivers, and they must keep prices artificially low to attract riders.
Second, platforms should deliver a product or service without the cost of having the same number of employees or large capital investments as conventional businesses. Ride-sharing platforms work best if the driver side is heavily populated so that customers can always get rides. However, both companies have had difficulty maintaining large pools of drivers. Prior to Uber’s IPO, drivers quit at a rate of 12.5% per month and the company was paying about $650 to hire each new one. With 3 million drivers, this equates to a cost of $250 million per month! Uber and Lyft would like to get rid of drivers altogether by using autonomous vehicles. But owning those would require a massive capital expense.
WeWork loses a lot of money too because of its flawed business model. WeWork is a one-sided platform: it leases office space and then resells it on short-term arrangements. It’s not matching owners and renters and collecting transaction fees. Another problem is that WeWork must take on the full financial risk of the real-estate leases, which total about $50 billion.
Airbnb has a better business model because it links two sets of actors and it can charge both of them. It doesn’t pay people to list rooms or to rent real estate, and it allows renters to charge market prices. (Uber and Lyft pay drivers even when they don’t have fares; WeWork pays for real estate even when it doesn’t have renters.) Still, Airbnb faces competition from other sites, like HomeStay, and from traditional hotels entering the apartment-sharing market. As result, Airbnb must spend heavily on advertising.
Why do investors back platforms that consistently lose billions of dollars? They’re likely hoping for a “winner take all or most” outcome. Once competitors disappear, then the platforms can raise prices and reduce subsidies. But Uber and Lyft already control the U.S. market, and neither is profitable. WeWork’s strategy has been to dominate the market for short-term leases and then gradually raise prices. However, the company has only about 2% of the office real estate markets in New York and San Francisco, and it can never control enough of these big markets to raise prices significantly.
To be sure, even these troubled platforms could succeed as smaller, more focused businesses. Uber and Lyft make money in a few cities where taxis and other transport options are limited and expensive. WeWork might survive if it can cut employee headcount, sell off its leases, and retain properties it can rent at a profit. Meanwhile, Airbnb must control advertising expenses and emphasize its differentiation from hotels, which don’t have the same combination of global and local coverage and variety of accommodations.
Successful platforms can be extraordinarily efficient and valuable. But these businesses are not so easy to create. And regardless, platforms still have to be better than the conventional competition. If platforms grow mainly through “subsidies” rather than through network effects, they will probably always lose money – and lose more money the bigger they get.
Michael A. Cusumano is a professor at the MIT Sloan School of Management and co-author of The Business of Platforms: Strategy in the Age of Digital Competition, Innovation, and Power (Harper, 2019).