Every business wants to be a platform, because that is where the big money is to be made. But what are platforms? The literature contains a confusing variety of examples. Search engines, social networks, online marketplaces, operating systems, browsers, the statistical package R, real estate brokers, shopping malls, nightclubs, color TV’s, auto body frames, credit cards, labor unions and political parties have all been called platforms.

Conceptual confusion makes it difficult to understand where we are, makes it hard to make decisions, and makes decisions difficult to communicate. In this blog I review a few platform definitions and then provide a unified view by giving, and motivating, my own. I conclude with some guiding strategic questions about platforms that follow from this.

Platform definitions

Tiwana (Platform Ecosystems, 2014) defines a platform as the extensible codebase of a software-based system that provides core functionality shared by apps that interoperate with it, and the interfaces through which they interoperate. An example is Apple’s iOS, on top of which apps are developed by third parties.

Srnicek (Platform Capitalism, 2017) has a different view and defines a platform as a digital infrastructure that enables to or more groups to interact. The Facebook and Uber apps are examples.

Rogers (The Digital Transformation Playbook, 2016) gives a variation of this definition, focusing on business instead of software. A platform according to Rogers is a business, not a software infrastructure, that creates value by facilitating direct interactions between two or more distinct types of customers. Parker, Van Alstyne and Choudary (Platform Revolution, 2016) and Reillier & Reillier (Platform Strategy, 2017) give similar definitions.

McAfee & Brynjolfsson (Machine|Platform|Crowd, 2018) view platforms as software, not as a business, and define it from an economic point of view as a digital environment characterized by near-zero marginal cost of access, reproduction, and distribution. This definition emphasizes the economic properties of digital platforms. Online marketplaces are platforms in this sense, shopping malls are not.

Cusumano, Gawer & Yoffie (The Business of Platforms, 2019) are agnostic about what kind of entity platforms are –business or software– but say what platforms do: Platforms connect individuals and organizations for a common purpose or to share a common resource. Credit cards and political parties are both examples of platforms according to this definition.

These examples and definitions give a feeling of similarity. But they also confuse us because they define different concepts. And we would like to have more than a feeling, we would like to understand what these definitions are aiming at. Let us start with the common core.

The common core

Baldwin & Woodard provide the common core of the platform concept: a platform is a set of components of a bigger system. It is the set of core components with low variety. The other components of the bigger system are called complements of the platform and have high variety.

For example, a car body can be the core component of a product family that contains many variations around a core. The car body is a platform for the product family and the remaining components are its complements. A political party can collect a variety of people around a core set of ideals about society. The party is a platform on which people, its complements, can work towards these ideals. An online marketplace is a core component in an ecosystem of buyers and sellers. The marketplace is the platform, the buyers and sellers are the complements.

Platforms support ecosystems

In the definitions above, platforms support a particular kind of system, that I will call a business ecosystem. Definitions of the concept of business ecosystem are even fuzzier than platform definitions. Here I offer the following simple definition:

A business ecosystem is a system of economic actors that depend on each other for their economic well-being and survival.

With economic actors I mean legal or natural persons responsible for their own well-being and survival. They can be companies or consumers, but also government organizations or non-profits, in as far as these are responsible to acquire a budget in order to achieve their goals.

This gives us the following definition of platforms:

A platform is a common infrastructure used by members of an ecosystem.

For example, travelers may use AirBnB as well as Booking.com as platforms to book lodging. Owners use these platforms for advertising their property. AirBnB and Booking.com in turn use Android and iOS as lower-level platforms to offer their services to smartphone users. This points to an important architectural property of platforms, discussed next.

Platforms define layers in ecosystems

Layering is the most basic structure of any system. In software engineering we speak of an n-tier or multitier architecture when functions are separated in layers of abstraction. For example, in a three-tier architecture, presentation, domain functionality and data storage are separated in three layers, where each higher layer can use services provided by lower layers. In some layered architectures, a higher layer can only use services of the adjacent lower layer. In others, the only constraint is that lower layers must not use services from higher layers.

Here is an example of a few layers in the Uber software. Each layer is a platform for the next to build on.

Ecosystems consist of business actors, which may be platform owners, developers, users, etc. The software layers define corresponding layers in the ride hailing ecosystem as illustrated in the following diagram. Some layers contain several platform owners, that compete with each other.

Two kinds of platforms

Since platforms support ecosystems, we can divide the functionality of platforms into two types:

  1. Functions to support value-adding activities of ecosystem members, and
  2. Functions to support coordination among ecosystem members.

Why these two? Because each economic actor needs to perform at least one value-adding activity in order to be economically sustainable, and because an eco-system consists of multiple business actors by definition, these actors need to communicate with each other. A platform can decide to support one or both kinds of needs. Most definitions above refer to the coordination function of platforms.

Coordination platformssupport coordination among business actors. Typical functions provided by these platforms are searching, matchmaking, communicating, performing a transaction, payment, and evaluating a transaction. Examples of coordination platforms are Uber, Lyft, LinkedIn, Google search, Facebook, WhatsApp, Amazon marketplace, and many more.

Value-adding platforms provide tools and components on which others can build value-added services. The functionality is usually domain specific. Examples are game platforms such as Xbox and PlayStation, cloud platforms such as Amazon Web Services, Microsoft Azure, Google Cloud, and Alibaba cloud, IoT platforms such as GE’s Predix and Siemens’ MindSphere, and operating systems such as Android and iOS. Value-adding platforms often provide software development kits for free.

The definitions of Srnicek and McAfee & Brynjolfsson listed above refer to value-adding digital platforms. The definition of Cusumano, Gawer & Joffie includes both coordination and value-adding platforms. They refer to the two types of platform as transaction and innovation platforms, respectively.

So what?

With these insights, what can we do now that we could not do before? We can derive a few guiding questions.

If you want to develop a platform, you need to know what ecosystem the platform is for. Why do economic actors want to be a member of an ecosystem? To survive and thrive. Why would they need your platform to achieve this? Do they need coordination functionality? How can you support them in their value-adding activities? Platform thinking is ecosystem thinking.