Trending: Buzz Aldrin shares latest moonshot vision: No to NASA’s Gateway, but yes to China
(GeekWire Illustration)

Amazon, Uber, Google, Facebook, Microsoft and many other big companies have something in common. They’re all platforms, providing underlying technologies that others use to build or facilitate their own businesses. But there are different types of platforms, with different motivations, and understanding those differences is key for anyone looking to build, compete against, regulate or do business with a platform.

“Platforms are everywhere, and they’re part of the everyday fabric of life,” says Michael Cusumano, a professor at the MIT Sloan School of Management. And that’s why we need to think about them, because some of them become so powerful that they create room for abuse, and they need to be regulated. And we also need to understand, as consumers, what we’re buying.”

Cusumano is the author of 14 books about the business of technology. His latest is “The Business of Platforms: Strategy in the Age of Digital Competition, Innovation & Power,” with co-authors Annabelle Gawer and David Yoffie. The book examines the hidden forces behind some of today’s most powerful and influential companies. It’s an especially timely subject given the current scrutiny of many of those companies.

I’ve been following the work of Cusumano and his colleagues since reading the landmark “Microsoft Secrets” as a newspaper reporter starting out on the tech beat nearly two decades ago. I caught up with him recently for this episode of the GeekWire Podcast.

Listen above, or subscribe in your favorite podcast app, and continue reading for edited excerpts.

Todd Bishop: Why should people care about platforms? Why do they matter?

Michael Cusumano, Sloan Management Review Distinguished Professor at the MIT Sloan School of Management.

Michael Cusumano: “There’s probably nothing you do during the day that doesn’t bring you into contact with one of these platforms. And in many ways they have this great influence on the services we use, whether it’s an Uber car to get somewhere; or an Airbnb room where we stay; or buying something using your smartphone. Platforms are everywhere, and they’re part of the everyday fabric of life. And that’s why we need to think about them because some of them have become so powerful that they create room for abuse, and they need to be regulated. And we also need to understand, as consumers, what we’re using and buying.”

TB: You and your co-authors make a distinction between innovation platforms and transaction platforms. What’s the difference?

Cusumano: “There is a big difference. Innovation platforms are technological foundations that bring together users with outside creators of what we call complementary innovations. In other words, as in the case of a smartphone, those complements would be third-party software apps. Transaction platforms are also bringing together different market actors, but for the most part they’re bringing together buyers and sellers, or something resembling that. We think all platforms fall into one of these two categories.”

TB: An innovation platform is more powerful, it seems.

Cusumano: “Yes, in the sense that there’s really almost no limit to what could be built on top of an innovation platform. The other important point we make in the book is that the most valuable companies in the world, and in many ways the most powerful platforms, are really hybrid companies. In other words, they do both. Facebook started out as a transaction platform, essentially just bringing together people who wanted to communicate or share information with each other. We consider that a transaction, even though it may be priced at zero. But Facebook Platform opens up APIs and lets outside companies build specific applications and games, and connect websites that let users do all sorts of things on Facebook, some of which have not been well-controlled. So it’s really adding the innovation platform function to the transaction platform that has made Facebook and some other platforms extraordinarily powerful.”

TB: All four companies who appeared before the U.S. House Judiciary Committee recently are platforms: Facebook, Google, Amazon and Apple. What’s your take on the current inquiries into those companies, as viewed through the lens of platform economics and competition?

Cusumano: “We added a chapter in our book called the “Double-Edge Swords: Harness Platform Power, but Don’t Abuse It.” Most of the books written about platform companies, including my earlier work with coauthors, as well, really didn’t look at the potential abuse side.

“Platforms create quite a few problems for regulators because some of them have great market power. This could mean up to 90 or 95 percent market share. Now in those cases it’s a little bit clearer when platforms abuse these positions. But having a huge market share by itself is not illegal. A company has to abuse that position. We write about what we learned from the Microsoft antitrust case, where Microsoft did abuse its position with Windows, very clearly trying to crush or harm PC makers that wanted to load a competing browser from Netscape; or later on, Microsoft tying Windows to other products that it bundled into Windows, like the browser and a media player.

“It also gets difficult with some the transaction platforms because it’s not so clear what their markets are. Is something like Internet search a market? Or is online commerce a market? Amazon might move into groceries, scaring the hell out of every grocery chain in the world, as well as small mom-and-pop shops. But in aggregate it only has about 4% market share in groceries in the U.S. So it really doesn’t trigger antitrust. And even in terms of online commerce, it still has less than 50% of online commerce in the US, and globally it’s even less, including China.

“Platforms are built on data, and the data crosses markets. Amazon can understand user purchasing behavior in a lot of different markets where they have very little share. So one of the issues in the investigation of Amazon the possible abuse of information it has from Amazon Marketplace which it can use to offer its own products to compete with third parties who are selling on Amazon marketplace. And sometimes Amazon makes acquisitions using that information.

“Google has also been criticized and fined in Europe for giving preference to its own services versus competing services and in listing priorities for shopping items in search. There are all sorts of ways platforms can abuse their particular functionality and market power. We think they have to start more aggressively curating content or membership and regulating themselves, as well as working more with regulators in different governments. Otherwise there’s going to be even more backlash against these platforms.

TB: Microsoft is a company you’ve paid close attention to over the years, as we’ve been saying. But they’ve largely escaped antitrust scrutiny this time around, even though they were in many ways the OG of tech antitrust. Why do you think that is?

Cusumano: “First I think they have changed their behavior … The second element is, most monopoly discussions today are centered around mobile devices and mobile services and Microsoft has kind of missed the boat there. And that’s another complicated story, which we do go through in the book — what happened to Windows Phone? So Microsoft is not being scrutinized because most of the attention in the last few years has been on the newer platforms that largely rely on mobile services, smartphone devices and Internet services sold via smartphones. So we see Facebook, we see Amazon, we see Google being the main objects of scrutiny or regulation at this point.”

TB: What makes Uber different as a platform company?

Cusumano: “One of the things that makes Uber different is it’s a bad platform. A platform should be solving some type of market imperfection, such as cases where there’s not enough supply to meet a particular demand. All platforms, transaction or innovation, bring different market sides together. They have to solve the chicken-or-egg problem and get the momentum going. One way to solve that is to subsidize one side or the other. The problem with Uber is that it pays out too much in subsidies and is in too many markets. There are some markets where there’s not enough transportation, not enough taxi cabs, for example. But to be in every market in the world, to replicate the world market for taxi cabs, it’s not necessarily a good market. With a global Internet platform, instead of just losing thousands of dollars, you can lose billions of dollars, and essentially, that’s what Uber’s doing at the moment.

“Uber is subsidizing both of its market sides. It’s keeping the price for riders below the market price, which is the taxi-cab price. It’s also paying bonuses to riders. For example, I just got an email the other day for 3£ off my next ride in London. Uber also subsidizes the drivers’ side by making set payments to attract drivers. So, rather than just pay commissions on each ride, it also pays drivers a separate fee, such as $2000 for the first 300 rides, and that’s become enormously costly. The other problem with Uber, and this was brought out by a consulting firm who looked at their pre-IPO documents: about 13% of Uber drivers in the last couple of years have been quitting each month. And it costs Uber $650 to replace the driver. Now, with 4 million drivers, they may be up to about $4 billion in annual driver replacement costs, unless they have reduced the driver churn rate. With all these costs, we expect Uber to lose between $3 and $4 billion this year.

“So, Uber has been a terrible platform. Taxi rides is not a good business, and it never will be a good business. The only way for them to make money is actually to just scale back and go into markets where there’s enough demand so that they can raise prices and not subsidize both sides of their marketplace, the riders and the drivers. They also have to reduce turnover of drivers.”

TB: Is scaling back the only option to make it a good platform?

Cusumano: “The other thing Uber has been been doing is to try to find niches where they could charge a little more or operate more like a true transaction platform. Like restaurant food delivery, Uber Eats, but again, delivering food is also not a particularly good business. And there’s also a lot of alternatives for restaurants. I noticed recently Amazon had started in that business and then they quit. It’s not a good business. They are also trying to match truckers with demand, and that might work.

“I would warn investors, even if Uber makes these new businesses successful, what kind of profit margins do you think it can generate? They’re still going to be tiny margins and they’re still going to be a pittance compared to what Microsoft or Google or Facebook can generate from selling digital ads or selling software. So Uber is a bad business; it never was a good business.

“By contrast, Airbnb is also a transaction platform, but a very different business model. Airbnb always makes money because it doesn’t have to subsidize anybody. It just provides access to assets that people already have and which they may decide to rent if there’s demand in their market. So Airbnb is a great example of a profitable transaction platform and we should see more of those. That’s where I think entrepreneurs and investors should focus — finding platforms with the kind of platform business model that doesn’t require massive subsidies. And even though Uber and Airbnb are both transaction platforms, the demand-supply imbalance they’re targeting operates very different.

TB: You and your colleagues take a data-driven approach in the book, looking at the revenue and profits of the different types of platform companies. What surprised you that would be a key takeaway from the book?

Cusumano: “Most people talk about these platforms, but I was surprised to learn that no one had actually collected any real data on multiple platforms over multiple years. And so this was the first example of trying to create a database of at least publicly listed platforms and look carefully at their performance. The biggest publicly listed platforms turned out to be, on average, very successful businesses.

“Compared to a control sample of firms in the same industries, platforms generated roughly the same amount of revenues, about $4.5 billion on average. But platforms did it with half the number of employees, and they were almost twice as profitable and also much more valuable — several times more valuable. The whole excitement over platforms is in many ways justified. If you can make it to this magical level of being a multi-billion dollar public company, you probably have a very lucrative business. We were surprised at how financially successful our sample of platforms were.

“But the other surprise was the large number of failures we identified — over 200 that competed with these 43 companies. And we wonder why investors keep throwing money at platforms that are bad businesses. Uber is the poster child for the observation that ‘platformizing’ a bad business doesn’t make it a good business. When we look at the most successful and profitable platforms, they are inherently good businesses. None of them are bad businesses. That’s an important point to think about. So don’t jump onto what we call the “platform-mania” bandwagon, but really think about what you are you investing in. What is the underlying business, and is it fundamentally good or bad? Those are some key takeaways from the book.”

Listen to the full conversation in the audio player above, or subscribe to GeekWire in your favorite podcast app. The Business of Platforms: Strategy in the Age of Digital Competition, Innovation, and Power, by Michael Cusumano, Annabelle Gawer and David Yoffie, is published by HarperCollins.

Like what you're reading? Subscribe to GeekWire's free newsletters to catch every headline

Comments

Job Listings on GeekWork

Artificial Intelligence Technical Expert (Senior Level)United States Patent and Trademark Office
Digital Marketing ManagerSeattle Southside Regional Tourism Authority (RTA)
Senior UI DesignerALLEN INSTITUTE FOR ARTIFICIAL INTELLIGENCE (AI2)
Deep Learning Engineer // Co-FounderALLEN INSTITUTE FOR ARTIFICIAL INTELLIGENCE (AI2)
Find more jobs on GeekWork. Employers, post a job here.