How do platforms behave in digital ecosystems?

October 18, 2023 Digital

Emerging from enhanced connectivity between sensors, software, clouds, cars, phones and other devices, the Internet of Things (IoT) will vastly extend the reach of modern technologies. Understanding complex, rapidly expanding IoT ecosystems is essential if governments are to keep pace with the digital revolution. At the frontier of TSE Digital Center’s exploration of platform economics, TSE’s Doh-Shin Jeon, Yassine Lefouili and Yaxin Li joined up with Timothy Simcoe (Boston University) to investigate this issue.

How does your research attempt to break new ground?

In the early 2000s, when tech giants like Google, Alibaba and Facebook were only in their infancy, several pioneering articles (including those of Jean Tirole and Jean-Charles Rochet) on two-sided markets began to uncover the hidden costs and benefits of the internet economy. They examined the strategies of an intermediary “platform” firm with two distinct user groups that provide each other with network benefits. For instance, credit cards are effective only if enough customers have the card and enough merchants accept it, so credit-card companies must balance the demands of both “sides” when they set fees. For reasons of tractability and exposition, economists often continue to focus on two-sided markets. In practice, Big Tech platforms increasingly serve multi-sided markets that can be described as ecosystems. 
Existing studies have also focused on two types of pricing: monopoly and competition. With the proliferation of platform business models, however, intermediaries can find themselves in complementary rather than competitive relationships. In the licensing market for 5G Standard Essential Patents (SEPs), for example, patent owners such as Ericsson, Nokia, Qualcomm, Samsung or Huawei supply complementary inputs to various IoT devices. The services of these complementors add value to mutual customers.
We develop a tractable model of complementary multi-sided platforms, and use it to study a number of questions: How does a side’s position within its ecosystem influence pricing and demand? What are the equilibrium prices charged by complementary platforms that serve overlapping user groups? How does the presence of a complementary intermediary influence decisions to extract value from, or subsidize, a particular side of the platform?

How do IoT patents help to inform your analysis?

Patent holders have traditionally licensed two sides of the cell-phone network: handsets and base stations. To the extent that handset users value greater coverage (i.e. more base stations) and carrier investments reflect the size of the user base, licensors face a two-sided pricing problem. The emergence of IoT, where connected products include not just phones and networks, but also watches, appliances, eyeglasses, and many other goods, converts this into a multi-sided ecosystem in which increased consumption of one device can raise demand for other devices. Network externalities in demand can also arise because users of one application generate data that improves the quality of another device. For example, data from search engines can improve the quality of maps and shopping sites, and vice versa.
We use the example of IoT patent licensing to motivate our model, in which each platform supplies an essential input such as 5G standard essential patents to every device. For a monopolist whose patent portfolio covers all IoT devices, our model yields a simple characterization of the optimal pricing. The platform could be a single firm that controls downstream device prices through vertical integration (e.g. iPhone and iWatch) and essential inputs (e.g. iOS and the AppStore). Our framework can also be used to analyze the scenario of multiple patent holders in a pool that licenses essential patents to the producers of IoT devices.

What are your key findings?

Our results reveal that the prices that a monopoly platform charges for its interconnected products reflect the usual tradeoff between internalizing externalities and extracting value. Prices depend on each device’s Katz-Bonacich (K-B) centrality, which measures the influence exerted by a particular device in a network.

All else equal, a monopolist sells more of a device when it is better connected. This helps to explain why large platforms subsidize products like search, navigation, and core “Smart Home” devices (e.g. Echo/Alexa and Nest/Google Assistant). These central devices generate data that can be leveraged across many applications, and interact with many other devices.
The relevant network used to compute the K-B centrality differs for a monopolist, social planner, or a group of complementary platforms. We also find that a monopolist that internalizes network effects may produce more welfare and consumer benefits than zero prices. This result has interesting implications for patent licensing of IoT and other platform technologies.

We then use our model to study how pricing changes when complementary platforms serve overlapping user groups. Our key insight here is that adding complementors leads any single platform to place more weight on externality internalization relative to value extraction in its pricing decisions. This can expand the range of outcomes for the total price charged to any single side/device, reversing the classic intuition that complementary monopolists charge a higher combined price than an integrated monopoly. 
Finally, we study a partial merger between complementary monopolies on just one side of a two-sided platform. For example, a platform might sell all its cellular infrastructure patents to a second platform to focus on handset producers. This produces a novel tradeoff: merging firms can eliminate the problem of double marginalization – or “royalty stacking” – that pushes up their baseline prices; but at the cost of increasing incentives for value extraction on the other side. Partial mergers can improve social welfare but are more likely to be harmful when demand externalities are very asymmetric.

What are the implications for future research?

Our key simplifying assumption of linear demand can be relaxed but we have not considered a more general specification of network effects. Another extension to our research would be to analyze competition between digital ecosystems. Our framework represents a first step in that direction, as called for by various competition authorities and commentators. In particular, future research might study the link between pricing and network centrality when platforms compete in a multi-sided ecosystem.


Ecosystems and Complementary Platforms’ and other publications by Doh-Shin Jeon, Yassine Lefouili and Yaxin Li are available to read on the TSE website.

Article published in TSE Reflect, October 2023