What You Need to Know Before Starting a Platform Business

There’s a great deal of enthusiasm about platform strategies these days. Entrepreneurs pitch their startups as the next Uber, the next Facebook, or the next Airbnb, while executives in established companies are retooling their strategies around platforms to drive growth and compete digitally.

But creating a successful platform business is not easy, as economists Richard Schmalensee and David S. Evans will tell you. Schmalensee and Evans have studied multisided platforms — in other words, businesses that create value by connecting two or more sets of participants via a physical or virtual platform — for more than a decade. Their most recent book for executives, Matchmakers: The New Economics of Multisided Platforms, was published by Harvard Business Review Press in 2016.

Schmalensee, a former dean and professor emeritus of management and of economics at the MIT Sloan School of Management, and Evans, chairman of the consulting firm Global Economics Group and a lecturer at the University of Chicago Law School, don’t sugarcoat the difficulties associated with pursuing a successful platform strategy. As they wrote in Matchmakers, “A multisided platform is one of the toughest businesses to get right. The entrepreneur has to solve a tough puzzle and use counterintuitive strategies to make a go of it.”

MIT Sloan Management Review editorial director Martha E. Mangelsdorf spoke with Schmalensee and Evans to learn more about just what it takes to get a platform strategy right. What follows is an edited and condensed version of their conversation.

In your most recent book, you point out that what you call the “matchmaking” business model — businesses that create value by connecting different groups — have been around for a long time, and have included everything from a trading area in ancient Athens to credit card companies. Is it safe to say that digital technologies make these kinds of businesses more common — and more important?

Schmalensee: Yes, the “matchmaker” model of making it easier for two sides to connect and create value — and then capturing some of that value — goes back at least to the ancient Greeks. But it’s the new digital technologies that have really turbocharged this business model. It’s easier to connect now.

Evans: Think about Uber. As a business, it could not have existed in 2008.

Schmalensee: Right. Without a lot of smartphone penetration, Uber’s business model wouldn’t work.

Evans: What has created massive opportunities for these platform businesses is the fact that, in effect, the internet is pervasively available in physical space. Between mobile phones that people carry around, and then the internet of things connected to wireless networks, you basically have plugged the physical world into the internet. More and more points globally are connected. That provides an immense opportunity to develop platform businesses that connect different groups.

So we’re in an era that, thanks to digital technologies, is highly conducive to platform businesses — and to their existing on a greater scale than in the past. Yet one of the points you make in Matchmakers is that the multisided platform business model — the matchmaker model — is not a very easy one to succeed with. Why is that?

Schmalensee: Our editor said what he liked about our book Matchmakers was that we were both curmudgeons. Our perspective is that there’s too much hype about this business model. It’s really easy to look at companies like Uber and say, “Look at these wildly successful businesses!” but forget all the people who tried to do similar things and failed.

In particular, there are two big difficulties associated with starting a matchmaker business. First, you have to solve a real problem for both groups you seek to connect. You have to be enabling something that wasn’t there before — and it’s got to be a sufficiently big thing that the outcome is good for Group A, and it’s good for Group B, and there’s a little of the value created left over for you.

The second difficulty is that you have to get started somehow. It’s easy to imagine matchmaker business models that work when you have 10 million people on one side of the platform and 10 million businesses or people on the other side. That’s easy, but getting there — figuring out how to get off the ground and attract enough of both groups to make it interesting to both — is not. You’re giving one side of the platform access to the other side for the purpose of creating value — and if there’s nobody on the other side, you don’t have a product. We refer to it as the “chicken-and-egg” problem.

Let’s talk about the first issue: solving a big enough problem. Can you give an example that illustrates that?

Schmalensee: Sure. The mobile payment system M-Pesa solved a big problem in Kenya: transferring money from family members who work in the city to the family members left behind in the village.

Could you make that system work in the U.S. with feature phones and convenience stores? No, because we don’t have a big problem transferring money in this country.

Ask yourself: What’s the economic friction that your platform will reduce? What problem are you solving? It doesn’t have to be something that people have been trying to do for years. YouTube is an interesting example. They thought video sharing might be neat, but nobody had done it before. It turned out they were right. So the platform could involve something completely novel that technology enables, but that’s a harder gamble than figuring out how to solve a big problem that already exists.

Let’s talk about the second problem you mentioned: the “chicken-and-egg” issue.

Schmalensee: Once you get lots of people participating on a platform, it can be really attractive to new participants, because there are a lot of people already there who they want to deal with. But if you can’t get momentum, you don’t ever have an attractive product, because you’re not selling access to an attractive group of potential partners.

Think of the restaurant-booking service OpenTable. If there aren’t enough restaurants on the system, then it’s not of much interest to consumers, and if there aren’t enough consumers, it’s not of much interest to restaurants to sign up. So in its early days, OpenTable had to solve a hard problem: How do you get both restaurants and consumers?

OpenTable figured out the trick was to offer software to restaurants that would enable them to manage tables and reservations and also get online. And OpenTable’s first idea was to start working with restaurant chains across the country to get a lot of restaurants onto its service. But the fact that there are lots of restaurants in Omaha on OpenTable is not of much interest to me if I live in Boston. I want there to be a dense set of restaurants where I live.

So OpenTable backed off its nationwide strategy for a while, focused on San Francisco and then Chicago, and signed up a bunch of leading restaurants in those markets. Then the company could go out and get consumers to sign up because it had a product. But the trick was to go first for restaurants and then to consumers and to give restaurants a reason independent of the existence of consumers to sign up with OpenTable — in this case, table-management software. That’s one way around the chicken-and-egg problem: You sell them something that is of value to one side of the platform and gives them access.

What are some of the other ways to address the chicken-and-egg problem?

Evans: Sometimes, like OpenTable, it’s a ratcheting or zig-zag effect where you get more restaurants, then you get more diners, and you keep moving the numbers up that way. There are other cases where you really need to go out and sign up — in effect, anchor tenants for one side of the platform — in order to really increase the number of customers on the other side. Take game consoles, where you need to sign the game developers up a year or two in advance in order to make sure games are available for your platform when you release it. In the game console case, it’s essential that you figure out a way to sign up developers, so that console customers have something to interact with.

Another solution to the chicken-and-egg problem is to make your own chickens. We all think of the iPhone as a multisided platform with an app store, developers, and users. But when Apple introduced the iPhone in June 2007, there was no app store. It wasn’t open to developers at all.

Steve Jobs and Apple did two things. One is that they created a bunch of applications themselves, so out of the box, the iPhone was a great experience for the consumer. And then they went out and did specific deals with a small handful of developers. The iPhone in June 2007 wasn’t what we would call a matchmaker or a platform at all. It wasn’t until about a year later that Apple opened it up and made it that way.

These different strategies — gradually ratcheting up the numbers on both sides of the platform through a zig-zag effect, signing up key participants such as anchor tenants, and creating your own chickens — aren’t mutually exclusive.

And do most successful platforms end up employing a range of them?

Schmalensee: I think they try all of the above, unless there is some obvious best approach.

Evans: In some cases, it is possible to go out and just develop one side of the platform, and then you can be pretty confident you can then sign up the other side. That’s how ad-supported platforms work: You come up with a bunch of great content. You get a bunch of people to come look at the content. Once you have enough eyeballs looking at the content, then you go sell ads.

You mention eyeballs. One of the points you make in your book Matchmakers is that, in the early days of e-commerce, one of the fallacies was that if you just got enough eyeballs, then you would necessarily win. Can you talk a little bit about the problems with that strategy?

Evans: Sometimes that strategy will work, but sometimes it won’t. The fallacy is in believing it’s just a numbers game. It’s not. It’s a game of coming up with the right number of participants on each side that actually want to do business with or interact with the participants on the other side.

If you go back to the OpenTable example, if you had a choice between having 50,000 restaurants and 100,000 consumers spread all across the U.S. or having 1,000 restaurants and 10,000 consumers in San Francisco, you’d probably rather have the latter. In the former case, you’re not going to have enough restaurants and consumers that actually want to do business with each other because the groups are just spread over too large an area. That was the flaw in the OpenTable strategy initially: They tried to go national when what they needed to do was to go local first.

Another point you make in Matchmakers is that pricing — how you allocate the pie — is more complicated in multisided platform businesses than if you’re simply dealing with one set of customers. Can you talk a little bit about that?

Schmalensee: One of the first things that strikes you when you study multisided platform businesses is that very often, one of the sides rides for free. For example, OpenTable gives me bonus points for using its service. I don’t pay a thing, and it’s a valuable service. And you can go on down a list of matchmaker businesses and see that very often, one set of participants goes free or doesn’t pay enough to cover its share of costs.

Sometimes it’s obvious which side should be subsidized. Sometimes it’s not — and it’s not always the case that one side goes free. The balance may shift over time as the market changes. So getting the pricing balance right initially doesn’t mean you get it right forever. You have to keep thinking about it. If you’re going to do some completely novel platform business, figuring out how to price it for both sides may be hard; you may have to experiment, and you may have to change pricing strategies.

Businesses that aren’t platforms don’t have the same kind of balancing act. You may have one group of customers over here, and another group of customers over there, and you price to each group, but in a normal business, those customers don’t interact. In a platform business, the interaction is key — and pricing to make sure you get that interaction right requires a balancing act.

Evans: If you see a successful platform business, the pricing solution looks obvious — after the fact. But it isn’t. If you’re actually starting one of these businesses, it’s not obvious at all. These entrepreneurs typically struggle, and it’s the ones who latch on to the right model who become successful. But it’s very, very complicated to figure out the right things to do to launch these businesses — both in terms of how to price them and how to get them off the ground. With these businesses, there are more dimensions of things you need to get right — and more uncertainty.

What are some of the other common mistakes that people make when launching platform businesses?

Schmalensee: Well, one that we talk about is getting the governance structure wrong. Again, it depends on the platform, but a lot of these businesses involve interactions that need to be governed in ways that are platform-specific. Take the social networks. The actress Lindsay Lohan at one point got thrown off of Facebook for posting under an assumed name — a violation of Facebook’s rules. OpenTable will throw you off the platform if you don’t show up for dinner reservations, and if you want to sell through the Amazon Marketplace, there are a whole set of rules you have to comply with. Securities exchanges have rules governing market makers, liquidity providers, traders, and liquidity takers.

If you get the rules wrong, even if your pricing is right, you can get bad behavior that drives people away from the platform.

Interesting. On a different topic, Peter Weill and Stephanie Woerner of the MIT Center for Information Systems Research have written in MIT Sloan Management Review about how some established companies are creating digital ecosystems that may include their competitors. It seems to me that that’s a different kind of platform problem — when you have an established business and you’re thinking about making it into a platform that connects not just you to your customers, but also maybe involves other companies that may be competitors of yours. Is that a question you’ve looked at, or does that just get into even more complicated economic questions?

Schmalensee: I think it is more complicated, but some of the same principles hold. The question is, how are you adding value, and how are you capturing value? If you’re going to become a platform, what’s the proposition you offer to folks on one side, and what’s the proposition on the other side? What problem are you solving?

It may be a problem for all participants or a problem for the end customer, but you have to be solving a significant problem. You have to give value to all sides. You’ve got to solve the chicken-and-egg question. You’ve got to get the incentives right. So, the basic economics are the same. The context may just be much more complicated.

Who’s done that successfully? Maybe Amazon with Marketplace?

Evans: It’s worth noting that Amazon tried to take on eBay early on with a Marketplace-like initiative and failed miserably. Initially, Amazon was not able to crack the chicken-and-egg problem, and it took them several tries before they were able to get Marketplace off the ground. It was initially an abysmal failure; they had trouble attracting merchants.

Dick described us earlier as curmudgeons. That’s probably an unfair term, but we’re realists; we’re thoughtful skeptics. So one thing I’d like to point out is that lots of companies shouldn’t become platforms. There are probably lots of areas of commerce where it probably doesn’t make sense to have a platform. Companies need to be mindful of not getting caught up in the hype — the idea that since Uber and so many other companies are doing one of these platform things that therefore that’s the right solution for you.

It may be that you’re in a part of commerce that is just not suitable for a platform business. Or it may be that you’re better off joining someone else’s platform, like a retailer joining Amazon Marketplace, rather than trying to create your own platform. Because you may not be successful at creating your own platform, whereas you may be very successful as a participant on someone else’s platform.

Good point. You’ve described the increased potential for platform businesses that exists because of the ability to connect groups, wherever they are, more seamlessly through internet-enabled devices. At the same time, you’re making it very clear that a platform business model is absolutely not a slam dunk. When considering a plan for a platform business, how can an executive gauge whether it represents a good opportunity?

Evans: One way to figure out whether there’s an opportunity is to see if there is an existing platform business that’s operated by traditional methods where, using modern technology, it is possible to provide the intermediation in a much more efficient and scalable way. It’s identifying platforms that already exist, but where there’s a lot of economic friction and inefficiency. And a great example of that is Uber.

The taxi industry was a two-sided platform. Taxi companies were basically intermediaries that connected taxi drivers and riders through dispatch systems. It was just a very inefficient platform, and Uber came along and figured out a way to operate much more efficiently within a city, but then also in a way that could be scaled geographically.

Schmalensee: I would emphasize the economic friction part of that point more than the technology, because a common mistake made by platform entrepreneurs is to think, “Oh, gee whiz, I could do a platform and use the internet and it would be wonderful.” But their platform turns out not be enough of an improvement over existing solutions. You have to ask: Does the proposed platform actually solve a real problem — as opposed to being just “gee-whizzy”?

And the second question to ask is: How are you going to get this thing off the ground? What’s the launch strategy? What’s the strategy for solving the chicken-and-egg problem? How do you get to this wonderful place where you’re making a lot of money because you’ve got the half the world on one side of your platform and half the world on the other side? What’s the plan?

Not every successful platform business had a perfect plan at the start, but they sure had to think about it — and often readjust on the fly when the first idea didn’t work.

Evans: Dick and I, I think, are resistant to making things sound too simple. But we have a lot of experience now at studying these platforms at an early stage and tracking them and finding out which ones succeed and which ones don’t. And while there’s all sorts of complicated stuff we can talk about, in my experience, the mistake that these businesses make time after time after time is not focusing enough on the point that Dick raised: Making sure the friction they eliminate — the problem they solve for users on both sides of the platform — is really significant. Many would-be platforms fail because they’re trying to use technology to solve something that might be a friction, but just really isn’t a big enough problem.

Schmalensee: That said, the combination of the matchmaker platform business model with all of today’s digital technologies means people are likely to come up with some new matchmaking businesses where, in five years, you’re going to say, “Oh, what a great idea! I wish I’d thought of that.” We don’t know what those new platform business ideas will be yet, but there will be some.