
What is a Platform?
This post is inspired by Ryan Sarver’s identically titled post. Ryan has been a huge influence to how I think about platforms and I’m excited to riff off of his considerable expertise. Both of us owe much of our thinking to the book Invisible Engines.
“Platforms” are the cronuts of the tech industry. Every start-up claims to offer one, but the real thing is actually quite rare. This bugs me because all the different “platform” definitions out there dilute people’s understandings of what platforms are and why they’re valuable. It also bugs me because I already spend a lot of time explaining platforms to people at cocktail parties who made the mistake two hours ago of asking me what exactly it is I do as Foursquare’s “Platform Lead.”
When I talk about “platforms,” I’m talking about systems connecting two or more groups together in ways benefitting both sides. It’s a pretty standard definition and you’ll hear the same one from my colleagues at Facebook, Twitter, and Apple.
For example, shopping malls are a retail platform. Shoppers come to a mall to connect with retailers and retailers open up shop to connect with shoppers. Malls are valuable to shoppers because they can park their car once and get all their Christmas shopping done in an afternoon. Malls are valuable to retailers because they don’t need to set up parking spots, clean the bathrooms, drive foot traffic, etc.
A great example of a developer platform is the iPhone. Developers write iOS apps because it’s the easiest way to run their service on people’s phones and they know that Apple will drive customers their way. Customers buy iPhones because of all the amazing apps developers write. In fact, the iOS developer platform is so strong Apple literally sells phones based on it. Remember Apple’s “There’s An App For That” advertising slogan? Nothing about features or specifications of the iPhone itself, just the promise of all the apps you could ever want.
So then what isn’t a platform? I separate most things people call “platforms” into three buckets: Services, Networks, and (legit) Platforms.
Services provide value to their users. Your laundromat is a service. GMail is a service. With services, the exchange of value is generally straightforward — I do this for you, you pay me money.
Networks are services that get better as the number of users increase. Your laundromat doesn’t get better if your neighbor starts using them. But a potluck gets better with every new guest. LinkedIn gets better the more users there are. Because the user is contributing value back, networks often don’t charge for the basic services offered, but instead charge for high-value, less frequent actions (e.g. recruiting on LinkedIn).
Platforms are networks connecting different groups together. eHarmony is a dating platform. EBay is a retail platform. Unlike a generic network, participant type matters. Each new straight man on eHarmony isn’t useful to other straight men, but a draw for straight women. Each new seller on EBay makes it a more valuable site for buyers. Platforms generally make money by skimming off some of the value they’re creating for one or more of the platform participants (e.g. rent or transaction fees).
For example, Netflix has historically been a content service. Consumers paid Netflix, and Netflix offered them content licensed from creators (studios, networks). New users on Netflix didn’t benefit other users, and content providers didn’t really care about the size of Netflix’s consumer base, except to negotiate higher licensing fees.
As a service, a product has to churn out features or lower prices to stay ahead of competitors (Hulu Plus, Amazon Prime, Cable TV). These competitors can copy features and lower prices as well, since there’s nothing inherently stopping consumers from switching.
As Netflix experiments with original content, it is evolving from a service into a platform — attracting content creators specifically to Netflix to reach Netflix customers and attracting customers to Netflix for Netflix original content they can’t find elsewhere.
Because Netflix has the biggest audience (or at least perceived as such), they can attract the best content, attracting even larger audience, attracting more top-tier content. As a platform, Netflix wins as consumers pick them based on the size and quality of their content network, not based on price or features. They also win as creators want Netflix to carry their content (to reach users) rather than trying to extract maximal $$ in licensing fees.
Every company wants to be a platform because successful platforms are hard to compete against — no new feature or giveaway will tempt users away. There’s a tendency these days to call any site with an API a “platform,” but an API is merely indicative of a service. To be a platform, an API needs to be powering connections between different groups. Similarly, the true value of a network or platform stems not from its feature-set but from its user-base. A sophisticated mall in the middle of the wilderness is technically a platform, but a worthless one.