3 Takeaways - Why Platform Companies (Facebook, Amazon, Airbnb and Uber) Are the Opposite of Traditional Companies, Why They Become So Enormous, and How They Can Be Effectively Regulated: Marshall Van Alstyne (repost) (#79)

Episode Date: February 8, 2022

Marshall Van Alstyne shares why platform companies dominate traditional businesses and why 7 of the 10 largest companies in the world are platform companies. Learn how they outcompete traditional companies while employing just a tiny fraction of the number of people, how they are completely different from companies of the past, and why platforms beat products all the time. Boston University professor Marshall Van Alstyne is the co-author of the international bestseller Platform Revolution and he is one of the world's experts on network business models. This podcast is available on all major podcast streaming platforms. Did you enjoy this episode? Consider leaving a review on Apple Podcasts. Receive updates on upcoming guests and more in our weekly e-mail newsletter. Subscribe today at www.3takeaways.com. 

Welcome to the Three Takeaways podcast, which features short, memorable conversations with the world's best thinkers, business leaders, writers, politicians, scientists, and other newsmakers. Each episode ends with the three key takeaways that person has learned over their lives and their careers. And now your host and board member of schools at Harvard, Princeton, and Columbia, Lynn Thoman. Hi, everyone. It's Lynn Thoman. Welcome to Three Takeaways. Today, I'm excited to be here with Marshall Van Alstyne. He's going to tell us how platforms, which underlie the success of many of today's most powerful and disruptive companies, companies like Apple, Google, Airbnb, Uber, and Amazon, are going to produce even greater changes in everyone's daily

lives in the years to come. Marshall is one of the world's leading experts on platforms. He's won numerous awards for his work. He's the author of Platform Revolution, which is an international bestseller. He teaches at BU and MIT and is a consultant to both global firms and startups. As Marshall says, quote, no matter who you are or what you do for a living, it's highly likely the platforms have already changed your life and are poised to produce even greater changes in the years to come in ways that few people fully understand, unquote.

Let's find out what those changes are. Hi, Marshall, and thanks so much for being here today. Thanks for inviting me. It's a pleasure to be here. How important are platforms? It's hard in some ways to understate important. One of the arguments that we like to make is that we believe that we're going through a transformation in the internet era that is as significant as the transformation that took place in the industrial era. There is a big difference, however. In the industrial era, it was all driven by supply-side economies of scale, high fixed costs and low marginal costs. If this is the case for electricity, for example, the railroad, first transshipment from New York to Chicago, huge expenses. You have

to buy rights away or build power plants or lay cable or lay track. But your second line of electricity, second transship friendship is extremely inexpensive. Hence, we get these giant monopolies. Not what's happening in the internet. Here, they're all governed by network effects, also called demand economies of scale. Here, users are creating value for other users, which ones that happening is you attract new users,

which has some great value, which attracts users, which has some great value. So Amazon merchants attract the consumers, which attract merchants on there. So you get these positive feedbacks on the demand side. So it's on the revenue side rather than the cost side, and we're seeing this gigantic monopoly. Give you the evidence, seven of the top 10 most valuable firms in the world today are platform. It used to be energy and banking were the dominant organizations in terms of market capitalization. Today, it's these technology platform companies, all driven by network effects.

So we think there are really deep and profound changes that are taking place in the internet companies and the economy today. Taking a step back, what is a platform? Well, we should always be careful with terminology. So we like to define a platform as three things. One of them is open architecture with rules of governance to facilitate interaction. The open architecture is what lets third parties come in and build. So it's the Android developers, it's the merchants on the stores, it's third parties posting on Facebook. The rules of governance are how they create value and how to share the value they get and how you

resolve conflict on that. The third place is the interactions. That's what actually creates the value, whether it's a ride, a post, a tweet, an app, a stay. It's third parties creating the value through their interactions on top. The platform is, in effect, designed to facilitate these exchanges or these interactions where third parties are offering this value. So again, a platform is an open architecture with rules of governance to facilitate interaction. Taking it to a very basic level, in some ways, it's like a farmer's market where different farmers come and bring their produce to sell. Are platforms new? That's a fair question. You should always think of these in some ways as

almost a spectrum. So if you think of a flea market or a farmer's market, one thing that might be missing from that is a true governance model. In the farmer's market, if you don't like the tomatoes, well, you got to haggle with the farmer and try to get your money back. But on Amazon or eBay, you have someone to appeal to. There's the governance system on the ecosystem. For example, I recently had the experience that I had tried to purchase a bicycle for my son for his birthday and honestly got cheated. There was a bad merchant on there that never shipped the bike.

Amazon not only refunded the money, they found that there were a large number of other complaints, and they booted the merchant off the platform. That same kind of intervention tends not to happen in a flea market in the same way. So you have someone that actually tries to make the rules work for both parties. And what makes platforms different now from the old style platforms like the flea markets or farmers markets? What else besides governance? So there's a really subtle but truly distinctive feature, which is the internalization of a network effect or an externality. A really good platform tries to take your interaction with me and then try to improve the other interactions on the platform. To give you a really simple example, your search on Google will help to make my search better, or my movie-watching behavior or my purchases on Amazon might then make others' purchases better.

What they've done is they've learned from one interaction to make the next one better. You won't see that in a typical standard farmer's marketplace. You won't see that in a typical shopping mall or something of that sort, where you still get producers and consumers interacting in location. The governance mechanism harnesses those positive externalities for the benefit of the ecosystem and tries to make the ecosystem better. And this is, again, if Google can then spread the value of those interactions across millions or billions of other searches, or Facebook can spread the value of those interactions across millions or billions of other searches, or Facebook can spread the value of those interactions across millions or billions of other users, you get this cascade of value spread across numerous interactions. It increases the value of the ecosystem as a whole as each one is learning from another.

Give you another really simple example. Suppose that a child runs in front of a Tesla vehicle and the machine learning has to learn to avoid recognizing the child in the ball. That can be spread across all the other vehicles and all the other driving. It's not just learning for a single vehicle. It's being spread across the entire ecosystem, which is this positive externality, which increases the value of the whole. And that's new and different. To me, that's so exciting, this idea of sort of a hive mind where one car can make all Tesla cars smarter or one AI doctor make all AI doctors smarter. Absolutely. But this also has to be designed.

It doesn't happen by accident. You need to design the observant. You need to gather the data on each end. You need to gather the data of what's happening and use that to learn to make recommendations and learn to improve the other interactions on the system. To think of it differently in the governance model, what you're doing is managing positive interactions and mitigating bad interactions. So you're improving the quality of good ones and discarding the bad ones and pushing them

off the platform. How are the new platform companies different? How do they compare with traditional companies? We like to characterize the platform companies as what I call inverted firms, meaning that lots of the value creation is happening outside the firm as opposed to inside the firm. In a traditional firm, take Dell or Coca-Cola, what you do is you assemble the ingredients, so sugar and water and tons of branding, and you deliver it from your supply chain to your destination, which is your end

customers. The firm takes responsibility for all the value add through the portions of the supply chain. The interesting thing about platforms, again, is the hint on network effects, users creating value. You can't scale network effects inside the firm as easily as outside the firm. Simple reason is there are more people and resources outside the firm. So you have to scale outside. What that means is you shift value creation from inside to outside. So third parties create a lot of the value. You and I create the webpages that are searched by Google. You and I create the posts, the value of Facebook. Third parties create the value on Uber, create the value on Airbnb. Merchants create the extra sales on top of Amazon. If you take the ratio of market capitalization to employees for platform company relative to traditional company,

that ratio is about an order of magnitude higher. Why? It's because the external ecosystem is doing the production of creating the value as opposed to the internal system. That's why that market ratio is so different. So again, there's a fundamental difference that we call them inverted firms for the reason that a lot of the value creation is happening outside rather than inside. And so that's the reason that these platform companies have no real assets. Uber doesn't own any cars. Facebook doesn't create content. And Airbnb doesn't own any real estate? That is exactly the reason. In effect, these platforms become orchestrators of third-party value creation. Now, there are hybrids.

So, for example, Apple makes things in addition to having its own ecosystem. Samsung has its own ecosystem. Google makes things and then also has its own ecosystem. So, it's both. But it is the case that the platform firms do an immense amount of orchestration of third-party value creation, which again is why there's a governance model or set of rules for participation and value creation that it has to orchestrate. So how does that change the business strategy of a platform company versus a traditional company? Business strategies are fundamentally different. The product strategy is fundamentally different than a platform ecosystem strategy.

Products have features. Platforms have communities. Often what you're trying to do is create features or cost differentiation or product differentiation in a product strategy that you're not having to worry about in the platform strategy. Often in a platform strategy, you want people you don't know to bring you ideas you don't have. Birds, all the kinds of sales of new product in place on Amazon are things that Amazon didn't necessarily think about. You are orchestrating and managing an external community that's creating immense value. Another huge difference from traditional strategy, you're trying to erect barriers to entry.

The boundaries of your industry, the standard economic model is to erect barriers where you control the assets your competitors don't have, how you minimize their market power. Here, you want permissionless entry, as much production consumption as possible. So Facebook wants as many contributors as possible. Google wants as many searches and webpages as possible. Amazon wants as many merchants on its ecosystem as possible. Apple wants as many developers on its ecosystem as possible. So it's permissionless entry of production and consumption as distinct from barriers to entry for competitors. Again, completely different way of thinking about it. So platforms must have enormous advantage if, as you mentioned, seven of the 10 highest market cap companies in

the world are platform companies. What are the advantages of platforms? There are multiple advantages. One of them, again, is this idea that people you don't know will bring you ideas you don't have. So you tend to be much more innovative. A second thing is if you design network effects correctly, you get positive feedback. You said users create value for users, which attracts users, which creates value, which attracts users, which puts you on an exponential growth trajectory relative to a platform. In some sense, products depreciate from use. Platforms appreciate through use.

That's a big difference. Also, they tend to be more resilient in the face of adversity. If you look at what happened with the rise of COVID, the adaptation in most cases is much higher in the platform community. As a matter of fact, the Financial Times say that in traditional energy banking industrial sector, a lot of firms lost market value and most of the technology and platform firms gained in market value. And as you shifted more things online and orchestration properties helped. It is the case, for example, that Uber took a real hit on its drive business, but it grew in its food delivery business. Interestingly enough, Marriott and Airbnb

both took a hit on their day business. But fascinatingly, Airbnb had its IPO during COVID, and it is now worth more than Marriott and Expedia combined. That's quite extraordinary how that's happened. But then it wasn't carrying these assets, so it didn't have the same infrastructure costs as Marriott did or these others did. So, the anticipated growth will be higher when the economy resumes. So, it's quite interesting. Resilience would be a third benefit alongside the innovativeness and the exponential growth that you often see. And the market capitalization, you harness and orchestrate assets as opposed to having to own them yourself. And why are platform companies able to scale so quickly? Many of them start with very inferior products or services, and yet they seem to disrupt industries fairly quickly. Not incurring the marginal costs of production that can scale as fast as they can add partners.

So if you're not having to buy cars and pay for gas when Uber, it's basically how fast can you get a driver to sign up? If you're not having to buy and research real estate if you're Airbnb, it's how fast can you get the partners to sign up. If you're not having to invent the apps yourself, it's how fast can third parties experiment and bring you new apps. So again, if you're not incurring the marginal cost of production, you are able to add value as fast

as third parties can get the feedback going. Now there is a launch and scan question. When you're just getting out, you may not have that. So again, Amazon was around for a long time before it actually managed to get the feedback going. Facebook didn't monetize for a very long time until it managed to pass MySpace and Friendster. It's not easy to get the network effects going, but once you do, it's very, very easy to scale. And where are platforms? Are they only in certain fields or industries? portion of information as value, the easier it is to make a platform. It's easier it is to spread that across more and more interaction because information is easily shared. Hence, software

and news and information industries transform very quickly. Another is how modular is the value proposition? So again, a ride, a tweet, a stay in a very modular Boeing aircraft. It's so complex. Even Boeing has a hard time getting 737 MAX off the ground. It's too complex. And so you wouldn't crowdsource an aircraft. And a third property, of course, is this riskiness, how risky is third-party experimentation. If it's not very risky, then it's easy to invert the firm, if you will. Or if it's risky, then you don't do so. Interestingly, healthcare is a high information business, but it's more risky, so it's slower to transform. So that's a nice illustration of an example where it will transform, but it's after the software and news industries.

You're orchestrating third-party value creation. So those are ways to look at it. But again, at its heart, how easy is it to invert the front? So most platform companies are focused on distinctive industries and markets. So Airbnb, Uber, Facebook, Amazon, YouTube, eBay, Wikipedia, the iPhone, Twitter, and Instagram. What do you see as coming next? What other transformations are coming on the horizon? What other areas?

Well, let me politely challenge that assumption. So I would actually say that platforms start niche, but one of the things that they do is they then rapidly expand into adjacent markets. In the examples that you gave, for example, Airbnb starts with stays, but now it's moving into experiences. If you take a look at Amazon, goodness, it starts with books, but now it's in cloud services, it's in movies, it's in healthcare, it's in groceries. If you look at Amazon, goodness, it starts with books, but now it's in cloud services, it's in movies, it's in healthcare, it's in groceries. If you look at Google, it starts in search, but now it's in mapping and it's in self-driving cars and it's in wearables.

So they move into adjacent markets relatively quickly by finding other interactions. If what you want to look at is the transformation for overlapping communities and what interactions you could add to an existing set of interactions in ways that add value. Once that happens, you can mark a platform move into an adjacent market to see what kinds of value you can create in that adjacent market. So you should expect the boundaries of platform markets to move and shift all the time. Returning to your early question about differences in product and platform strategy, the boundaries for product markets are much clearer than they are for platform marketplaces because you tend to move and march rapidly into market adjacencies and envelop those adjacent markets. This is one of the reasons you get such gigantic companies. If you look in China, Alibaba and Tencent are in all kinds of different industries, perhaps even more diverse than Google and Amazon are today. So it's quite remarkable how

diverse those industry interconnections are. And is that because they are largely information-based businesses and the marginal cost of growth is so low. So almost everything is profitable. It's a huge asset. And it's also the case that information in one industry may give you leverage into an adjacent industry. To use an example from Alibaba, when Alibaba entered banking, it had the transactions data of merchants on its sales platform, which then allowed it to underwrite loans at a lower cost than banks. The banks would have to send a human in to audit books, but Alibaba had already seen the transactions. So they knew how successful they were and could use that additional data into the loan origination estimates. Then what happens after you've given out the loan? Well,

the interactions are conducted on the platform itself. Banks don't have that, so they can't take a share of the revenues from the subsequent transactions. So it creates advantages to move into the adjacent markets that the banks didn't have. Hence, they've now launched Ant Financial, which was, until it was nixed by the Chinese government, about to be the largest financial IPO in history. So it's quite remarkable how these things can actually get moved in adjacent markets. And it is, again, it's possible to use the data from one to leverage into another, use the transactions in one to leverage into another. What other fields do you think are ripe for transformations as platforms? We'll go through that checklist we gave you on what industries are likely to transform. So if you think of proportionate information value, if you think of modularity, if you think of risklessness, healthcare is absolutely going

to transform. Interestingly, finance is another one that's going to transform. Unlike news and software, it's had more regulatory protection, so it's riskier in some sense. Those are going to be overcome and modularized. So finance, especially in open banking coming out of Europe, that's also going to transform. Fascinatingly, automotive and mobility is absolutely going to be overcome and modularized with finance, especially in open banking coming out of Europe. That's also going to transform. Fascinatingly, automotive and mobility is absolutely going to transform. More machine interaction with more machine. You are advising the EU and other governments on antitrust. I talked to Eric Schmidt, the former CEO of Google, on a recent episode, and he said platforms are different to regulate and regulators have to be careful of unintended consequences. And he gave an example, hypothetically, that Senator Warren had suggested of splitting the Apple App Store away from the iPhone. And Eric talked about what

would happen and how platforms are different from an antitrust perspective to regulate. What are you advising the EU and governments on regarding antitrust perspective to regulate. What are you advising the EU and governments on regarding antitrust of platforms? Great question. First, let me concur that platforms are different, and they probably do need some differences in regulation. So let me give you more two examples and then try to give you some of the solutions. So one or two of the examples are, one, usually there's a test of predation if you're pressing below marginal cost or above cost. Above cost, and that's evidence that you have market power. Below, perhaps that's evidence

of predation. Well, guess what? Most of these platforms often have free services. So how does that compare? Well, the reason for this goes back to the two-sided pricing I told the original one. So free pricing is profit maximizing, even in the absence of competition, but it's devastating the competition. Another one, often in order to raise prices, you'll be restricting supply. Google isn't restricting your searches. Facebook isn't restricting your posts. Amazon isn't restricting your purchases. Not what's happening.

In effect, what they're doing is they're learning from that in order to maximize the value of other interactions. So what they're doing is creating value using third-party interactions. So again, it's a completely different business model. Again, market boundaries are extremely difficult to define. Is Google in search or mapping or self-driving cars or healthcare or movies? The boundaries of these get very, very challenging. So the traditional antitrust has some, the traditional tests of predation and things don't particularly work very well. You know, illustration. Here's why some of the core proposals of Senator Warren and some of the others of splitting Apple, breaking Facebook and different companies. I can give you grammar school algebra to show you why that's a bad idea.

All the platforms, again, are governed by network effects. Imagine the following. Imagine you split Apple into companies or you split Facebook into three companies. Facebook, WhatsApp, and Instagram. You've destroyed value. In effect, users can't connect with other users or you can't spread them across more interactions. So it's a bad design decision. Those interventions don't work. That's not to say that they should not be held accountable. They should be held accountable. There's no question that there is abuse of market dominance. They are umpire and player in their own ballgame, calling the shots again, governance model. They're entering markets for their own

partners with data that the partners themselves don't have because they're able to observe what's happening. So there's an information asymmetry there. So there are a couple of different solutions that I think we need to work with. One of these is not the horizontal separation. We're splitting companies up across different marketplaces. One is the separation of the value on top from the infrastructure on the bottom. Imagine, for example, that you separate the social network underneath from the advertising on top. But now you were to let Google and Amazon and Facebook all compete for advertising on your social network. And you get to choose who gets to advertise you.

What that means is they're going to have to share more value with you to let you choose them. And so that's going to create the competition on top that might actually help. So it's a vertical separation as one element of that. Another one, what we should do is also require, if you were going to enter a marketplace where one of your partners is already resident, what we need to do is declare what are the terms for partner participation. Then your own vertically integrated service is going to have to offer those exact same terms, so there's equal footing. So if you're going to use data to enter that market, you're going to have to give them exactly that same data. It's not that we don't want the platform

to enter. Matter of fact, under any other circumstance, you want competitive entry. It creates benefit for consumers. What we don't want is unfair entry where it's only the platform that gets to do it. The rules of entry are the same for everybody in terms of access to customers and in terms of access to data. And the platform itself is going to abide by those same rules in addition to some of the others. To give you another one, one of the recommendations has been data portability. The premise is to restore market power of the users in their data and create some competition with it. Unfortunately, that doesn't work because that creates islands of secrecy. We can't create network effects and individuals can't create the value for themselves. You can't make my Google

search better using your search data. I can't make your recommendations better based on my movie. It's just too small. You need an orchestrator to do that. What we're proposing is rather than portability, you gain in situ data access rights. What this means is you can then allow third parties to access your data where it's resident. So rather than pulling it out of the platform, where it goes stale and where you can't act on it anymore, you can grant third party access with your permission to manage it on your behalf. What that means is you then create competition inside the platform itself. So a model for this already exists in Europe where they introduced open banking, where Citibank could manage your Santander account or Amazon could manage your Citibank account. This means they compete with one another to create banking benefits on your behalf. We can do exactly the same thing here.

This also then solves this problem of actionability of the data. It's not stale anymore. And third parties get competition resident inside the platform. So I think these are several different mechanisms for creating superior competition and harnessing the value of network effects. Most of the traditional interventions destroy the value of network effects. So we need to respect how value is created, but we create new forms of competition. We allow third parties to also create those network effects on top. And we have another way. And C2 access is a good mechanism to do that. Such an interesting and different approach.

It would have huge implications in health care. We're now a variety of health care providers, doctors, hospitals, and others have essentially ownership of people's healthcare data. Imagine that you could give Amazon permission to manage that data and then try to get you more cost-effective medicines. You'd be able to act on that in your own benefit and reduce healthcare costs. Everyone wins except those who have fat margins

in the current system. And finally, Marshall, what are your three takeaways that you would like to leave the audience with today? One, I do believe that we are in an economic change as profound today in the internet era, as profound as in the industrial era a century ago, that really we are seeing the rise of gigantic firms today, analogous to the rise of gigantic firms in steel and oil and banking a century ago, but it's the opposite region. Today, it's network effects and demand economies of scale when a century ago, it was supply economies of scale. This means lots of different things,

but the real issue is that we're getting gigantic firms leading to the antitrust issues that we see today. The second thing that I would say is this leads us to a managerial set of principles. You cannot scale network effects inside the firm as easily as outside the firm. So platforms are inverted firms. They create value outside. If you scale network effects, you flip the point when you have to shift to creating value outside because there are more users, more resources, more partners outside. What this means is all of the things we teach in management schools, whether it's strategy, resources, supply chain management, marketing, all have to take on an external counterpart where you're also creating value using third parties,

not just creating it yourself. So all of these traditional management practices, all the antitrust practices, need to change to recognize how value is created outside the firm, as you think from inside the firm. The third thing which we haven't quite touched on, again, is going back in time, I would argue that we probably need, in addition to some things we have got so far, is a magna carta of citizens' rights on platforms. You and I don't get to vote in the rules of Page, Bezos, and Zuckerberg at as high time that we did. And I think that will happen as we start to actually get antitrust legislation in place to deal with antitrust and misinformation and citizen voting.

I'm going to argue for a Magna Carta of Citizens' Rights and Platforms, and that will happen. So I think those three things would be the takeaways. Marshall, this has been great. Thank you so much. My pleasure. If you enjoyed today's episode, you can listen or subscribe for free on Apple Podcasts or wherever My pleasure. 3takeaways.com is with the number 3. 3 is not spelled out. For all social media and podcast links, go to 3takeaways.com.