How Digital Platforms Work

Digital platforms can disrupt entire industries, creating new winners and losers. They can influence the behavior of billions of consumers, and hundreds of thousands of developers and manufacturers.

Just look at the example of what the Internet platform did. In the book Machine, Platform, Crowd by MIT professors Andrew McAfee and Erik Brynjolfsson, the authors chronicle the disruptive impact of the Internet:

  • The print industry was decimated. In 2013, US print newspaper revenues declined by 70% over the previous decade. Online ads only contributed back $3.4 billion of the $40 billion in print sales.
  • The music industry tanked. Worldwide sales of recorded music declined by 45%, from $27 billion to $15 billion, between 1999 and 2014
  • Radio was crushed. US radio station revenue declined by 30%, from $20 billion in 2000 to $14 billion in 2010

We are all familiar with the disruption that the Internet caused (though seeing the stats above certainly put things in perspective for me). The question is — why? What properties of digital platforms make them so powerful? McAfee and Brynjolfsson do a superb job of answering our questions.

Free, perfect, and instant

Digital platforms harness the power of information goods — products made of bits rather than atoms. These information goods are easily accessible on networks like the Internet. Thus, as the authors point out, information goods are free, perfect, and instant.

  • Free: once you have digitized some information, content, or software, it’s essentially free to make a copy of it.
  • Perfect: each copy is just as good as the digital original.
  • Instant: networks allow immediate distribution of the digital information good to anywhere in the world.

As McAfee and Brynjolfsson point out:

“Platforms are online environments that take advantage of the economics of free, perfect, and instant. To be more precise, a platform can be defined as a digital environment characterized by near-zero marginal cost — of access, reproduction, and distribution.”

How platforms disrupt the incumbents

Platforms create very intense competition for legacy products, which just can’t compete with the near-zero marginal costs that platforms offer. The authors provided two examples of this competition: newspapers and the Internet; and SMS messaging and WhatsApp.

Newspapers and magazines struggled with the rise of the free, perfect, and instant Internet in three ways.

  1. Online classified services like Craiglist (free, perfect, instant) crushed newspaper’s classified revenues.
  2. Alternative content platforms fragmented and siphoned away their audiences. “A huge number of content platforms, spanning every medium, topic, industry, and contributor type, from professional journalists to freelancers to unpaid enthusiasts, emerged as alternatives to mainstream print media.” Content and news was now available in a free, perfect, and instant way.
  3. The growth of online advertising platforms (Google, Facebook, ad exchanges) provided much better targeting and ROI and reduced the transaction costs of creating and running ad campaigns.

This triple-threat was what caused the decline of print newspaper revenues by 70% between the mid-90s and 2013.

Another example that McAfee and Brynjolffson discussed was WhatsApp — the smartphone app that lets users send messages to each other via WiFi/data networks, rather than via SMS. WhatsApp provided a much less costly way for users, particularly in developing countries, to exchange messages with each other. As a result, users flocked to WhatsApp which provided free, perfect, instant messaging — and the global mobile carriers saw their SMS messaging revenue decline drastically.

Network effect

With the WhatsApp example, McAfee and Brynjolffson also pointed out another important property of digital platforms — the network effect that they often create. Even beyond lower cost, users began switching to WhatsApp because many of the people in their networks were already using WhatsApp.

“This is a clear example of what economists call a ‘network effect’: the fact that some goods, like WhatsApp, become more valuable to each user as more people use them.”

Network effects create “demand-side economies of scale” — meaning that bigger networks gain a significant advantage in acquiring new customers than smaller networks.

In WhatsApp’s case, the network effect grew out of an early architectural decision that the founders made. WhatsApp made the decision early on to not be interoperable with SMS, a decision which required users to install and use WhatsApp in order to send and receive messages from other WhatsApp users. As a result, you couldn’t simply go into your SMS inbox to receive WhatsApp messages — you had to go into the app itself. This created a powerful network effect for WhatsApp — the more users in your network who used WhatsApp, the more valuable it became for you and the more incentive you had to download it yourself. Hence, the decision not to be interoperable with SMS created the network effect for WhatsApp.

A touch of economics: demand and supply curves

To understand the true power of digital platforms, we need to dive into microeconomics a little bit. You are likely familiar with the concept of demand and supply curves.

The demand curve (Source: Machine, Platform, Crowd)

The microeconomics demand curve captures the basic concept that for most products, demand increases as price decreases.

The supply curve (Source: Machine, Platform, Crowd)

The microeconomics supply curve captures the concept that for most products, supply (willingness to produce) increases as price increases.

Supply and demand curve together (Source: Machine, Platform, Crowd)

When you draw the supply and demand curve for a product on the same chart, you can observe the market clearing price P* — the point at which the supply and demand curves intersect. This is the price that the market is willing to bear — at this price, producers will supply the quantity of products that consumers are willing to buy.

The price multiplied by quantity at the market clearing price is the total revenue going to suppliers. This is indicated by the dark shaded rectangle in the chart above.

The shaded triangle above the revenue rectangle is the consumer surplus. This is the value that is going to consumers who were willing to pay higher than the market-clearing price P*, but were only charged the market-clearing price for the product. So these consumers were essentially able to keep this money in their pocket.

The triangle to the right of the revenue rectangle captures the unserved consumers. These consumers were willing to pay some price >0 for the product, but weren’t willing to pay the market-clearing price P*.

Complements shift demand curves

Now we’ll bring another concept into our discussion — that of complements. There are many different examples of complements — hamburger meat and buns, cars and gasoline, bottle and bottle caps. As McAfee and Brynjolffson write:

“In general, complements are pairs of goods with the following property: when the price of good A goes down, the demand curve of good B shifts out (meaning that demand goes up)… this means that a higher quantity of good B will be in demand, even though its price doesn’t change at all.”
As the price of hamburger meat goes down… (Source: Machine, Platform, Crowd)
…the demand for hamburger buns goes up (Source: Machine, Platform, Crowd)

How does all this discussion of demand curves, supply curves, and complements relate to platforms?

Many platforms have apps that are built on top of them — think of examples like Windows (licensed software); the Internet (websites); iPhone / Android (mobile apps). These apps are complements to the platform. And the availability of these apps increase the demand for the platform in two ways.

First, a large variety of apps increases demand for the platform because different consumers have different preferences for what apps matter to them.

“The ‘killer app’ varies across potential iPhone customers. Some want games, some want business tools, some want to stream music while others want to make music, some want to use social media, some want to use their phones as small scientific instruments, and so on. The best way to discover these preferences, let alone to satisfy them, is to turn the App Store into something closer to an open marketplace than a store with a single owner.”

Second, apps increase demand for the platform because apps are complements. Think of each app having its own demand curve, and for the platform to have its own demand curve as well. As we saw from the chart above, the demand curve for complements interact with each other. If apps and the platform are complements, then when the price for one of them decreases, demand for the other increases.

Many of the apps on the iPhone are free, because the developers plan to monetize with a freemium model, with ads, or in-app purchases. A good example, the authors point out, is the Angry Birds app, which is free.

Since Angry Birds is priced at zero, there is a tremendous amount of consumer surplus created. Some users would have been willing to pay for using Angry Birds, and now they’re getting it for free. And since Angry Birds is a complement to the iPhone platform, as its price goes to zero, it has the effect of pushing the demand curve for the iPhone out.

As the authors write:

“So, the existence of free apps like Shazam and Angry Birds has two effects: it generates consumer surplus (which is great because you always want your customers to feel like they’re getting a bargain), and it nudges the iPhone’s demand curve outward, which is exactly what Apple wants — more people who are willing to pay the iPhone’s price.
“Each of these apps probably shifts the phone’s demand curve outward only a small amount; after all, how many more people are going to be willing to pay $ 599 (the original 2007 iPhone price) for an iPhone just because they can play a game on it? But the effects of these complements are cumulative. Each free app adds one dollop of consumer surplus to the total bundle offered by the iPhone, and also pushes its demand curve that much farther out in the direction desired by Apple.
“An expensive phone doesn’t become that much more attractive a purchase because there’s one desirable and free app available for it. But what about when there are hundreds of thousands of free apps, some large subset of which are going to be desirable to almost any imaginable customer? Such a huge collection of apps would yield a lot of consumer surplus, and would push the demand curve out a long, long way. Expectations play a role as well: all those apps and app developers give consumers more confidence that their phone will continue to be valuable, or perhaps become even more valuable, as time passes and new apps are introduced.”

So the presence of all of these apps have a significant cumulative effect on the iPhone’s demand curve, pushing it out. There is tremendous amount of consumer surplus created by all of these free apps. Furthermore, there is also much more demand for the iPhone as its demand curve gets pushed out, even if the iPhone’s price remains the same.

Summing it up

We all know that digital platforms are disruptive to industries (Windows, Internet, iPhone) — but many of us have wondered why, and how?

Andrew McAfee and Erik Brynjolffson have done a fantastic job of explaining why digital platforms are so disruptive.

  1. Free, perfect, and instant. Digital platforms take advantage of the fact that networked information goods are free, perfect, and instant. These properties make it difficult for incumbent players to compete in a market. Examples include the impact of the Internet on the newspaper industry, or the impact of WhatsApp on mobile carrier SMS networks.
  2. Network effects. Digital platforms can produce network effects, where the value for each user in the network increases with every additional user who joins. These “demand-side economies of scale” mean that larger networks have a significant advantage in acquiring customers than smaller networks.
  3. Complements. Most digital platforms have applications that are built on top of them. These apps are complements to the underlying platform. As the number and variety of free apps grow, it increases the demand of the underlying platform in two ways. First, the variety of apps increase the chance of a user finding his/her “killer app” on the platform, which would make the platform more valuable to the user overall. Second, each free app nudges the demand curve of the platform out a bit. The cumulative effect of all these apps mean greater consumer surplus, as well as greater demand for the platform at the same price.

For anyone building a platform or working with an existing platform, I highly recommend reading Machine, Platform, Crowd. I have a much deeper appreciation for the power of platforms, and I hope that you will too.