What I mean when I say ‘Platform’

I do lots of government work. I’ve done lots of digital healthcare and private sector work too, but recently I’ve been very immersed in central government.

When you work in digital in central government, you find that terms like ‘service’ or ‘product’ or ‘platform’ factor heavily in your day-to-day conversation. You can also take for granted that others understand what you mean, and are working from the same definition.

I was talking to a colleague recently about what I call ‘Platform Product Management’ (see my previous post). Very early into the conversation, I found myself unable to answer a very simple question:

What do you mean by ‘Platform’?

What a good question.

I’ve worked on digital platforms for government, health and private sector clients, and each had their own unique sector-specific challenges. When I went to answer with my government-focused definition, I realised it wouldn’t work. I didn’t have an answer.

So here (weeks later) is my answer.

Platforms in a government context

In government, the answer is easy (mostly). There’s even a term for the shift towards platforms in government, referred to as Government-as-a-platform:

Government as a Platform is a common core infrastructure of shared digital systems, technology and processes on which it’s easy to build brilliant, user-centric government services

And there are plenty of examples, that are well documented, with clear value propositions:

• GOV.UK Notify — enabling public servants and government services to send messages to users
• GOV.UK Pay — enabling government services to process payments from users
• GOV.UK PaaS — Enabling government services to be deployed, hosted and monitored in the cloud

So basically — a common tool, component or API that other government services can use. But some government platforms can be used for more than just government services. Data platforms, and APIs in particular, are valuable tools for non-government organisations to build products around (see data.gov.uk and the DVLA API Developer Portal).

Richard Pope notes this in this wonderful post, where he offers a definition of Government-as-a-platform that covers non-government use:

“Reorganizing the work of government around a network of shared APIs and components, open-standards and canonical datasets, so that civil servants, businesses and others can deliver radically better services to the public, more safely, efficiently and accountably.”

Upon reading the above statement, you can understand exactly what government platforms are. But you might also see that the same definition clearly won’t work outside a government context

Thinking about platforms more broadly

Let’s start by looking at some examples that come to mind when thinking of platforms:

• Heroku — A cloud platform enabling users to build, deploy, scale and monitor their applications.
• Twilio — An API that enables users to send and receive text messages from their own applications.
• Paypal Commerce — Components that enable users to take payments from in their own websites and apps.

These examples all have two things in common:

• they provide a foundation for users to build their own applications upon, rather than providing the application itself
• the platform solves a common problem, effectively commoditising the solution

So, there’s a distinction between platforms and products (see Aha.io’s “What is a product?”), and a key attribute of platforms is how they underpin others, summed up well by Adrian Bridgwater in “What’s The Difference Between A Software Product And A Platform?”

the industry considers a platform to be anything that you can build upon

But there are some issues to be aware of.

Issue #1: Content platforms don’t fit

Web 2.0 and the explosion of social media shifted the way people use the Internet. Rather than consuming content, users provide the content — in an ecosystem where other users can consume it.

The same goes for marketplaces like the App Store — developers create an app, submit it to the store and other users can consume it. The value of the store itself is in the curation, quality standards and delivery of those apps.

These content platforms don’t fit into the definition of being “anything that you can build upon” — people aren’t building on top of Youtube or the App Store. But they are platforms… aren’t they?

Well, there isn’t an easy answer. Platform-hunt conducted an analysis and categorisation of 160+ digital platforms, with the results of their categorisation published onto this Trello board

When I think of platforms I think of “infrastructure platforms” or “utility platforms” but not necessarily “content crowdsourcing platforms” like Medium, TripAdvisor or StackOverflow. Nevertheless, I think this categorisation is very useful — it provides a framework for understanding where your platform fits. Unfortunately, it doesn’t help us in coming up with a definition to cover all platforms.

For now, I tend to exclude content platforms when I talk about platforms.

Issue #2: Products can be both products and platforms

When trying to understand the difference between a product and a platform, it can be confusing when you try to categorise things like Facebook, Spotify and Salesforce. All of those examples are both products and platforms:

For example, Facebook is a product for most users, they log in and have a need met, then log off. But Facebook is also a platform for developers to launch Facebook-connected apps.

The NHS App is another great example — users receive value by opening the app, booking an appointment with a GP, then closing the app. But it’s also a platform that surfaces other health services, benefitting them by having solved common problems already— offering biometric single sign-on and notifications as part of the platform.

You can try to disaggregate the examples to help make sense of them — e.g. Facebook is the product and ‘Facebook Platform’ or ‘Facebook for Developers’ is the platform.

But it might simply be easier to accept that some digital brands represent both products and platforms, and separating them into boxes might not be that helpful.

Issue #3: ‘Platforms’ with only one consumer can’t really be called platforms

I think it’s fair to say that platforms are about common or shared solutions to common or shared problems. If you build a digital payments component that is only used by one government service, it’s not really a platform — whereas a payments component used by hundreds of services across government would be a platform.

There’s an interesting point of time when developing and rolling out a new platform— the point where you move from one consumer to two. You might have designed a platform to work for hundreds of consumers, and built the word “Platform” into the name of your offering, but that merely reflects your aspiration rather than the reality. When you only have one consumer it’s not really a platform.

The transition from having one consumer to two consumers is a litmus test for the platform’s design. I’ve seen platforms struggle at this stage as they’ve bent over backwards to tailor their offering to the first consumer, and when the second consumer comes along, half of it isn’t relevant or suitable.

This is why as a consumer of platforms, I’ll always ask how many other consumers are using it in a production capacity. It’s also why I think the word “common” is really important in the definition of a platform.

My preferred definition

Taking into account everything in this post, I’ve put together a definition that I can use the next time I’m asked ‘What do you mean by ‘Platform’?”

Where a digital product is any technology that provides direct value by serving a user need,

Platforms are a common infrastructure of digital systems, technology and processes upon which other digital products can be more easily built

I know this definition doesn’t cater for content platforms like the App Store or Youtube, but I think that’s okay.

That’ll do for now.