A Platform product is one that enables combinatorial creation and consumption of value by leveraging building blocks.

We are all familiar with a large number of platform products in the world such as Facebook, Tencent's WeChat, Google and Apple app stores, Amazon's Web Services (AWS), eBay, Craigslist and more.

Did these platform companies start out building platforms? Well, no.

It is my thesis that most platforms are not born but made - serendipitously, deliberately and often, both.

Let us explore this with a few examples.

The History of the Internet As a Platform

The Internet is the mother of all modern platforms. In the last 30 years, the internet has facilitated an unprecedented level of innovation and creation that is cumulatively more than the last millenia.

Computer networking (enabled by TCP/IP) had been around for several decades before the advent, and the proliferation, of the World Wide Web. The HTTP protocol and the first browser, Mosaic, led to very easy sharing of content and eventually a myriad of applications. Eventually, that led to widespread use of E-mail, messaging, web applications (now called desktop apps), and many other applications that would otherwise not be "unlocked" and nurtured by a single provider (Remember AOL?)

The creators of the TCP/IP or HTTP did not start out by building a platform. They built components that could be widely used. And used, they were!

Amazon's AWS as a "Company" Operating System.

In the commercial arena, Amazon's AWS is one of the most powerful platforms that has been invented. Even though Amazon.com started in the mid nineties but the first comercial incarnation of AWS wasn't launched until a decade later. As Amazon kept growing, they built their internal systems in a decoupled, reconsumable, API driven manner. That was for their own consumption. As they scaled their offerings, they realised that it was a set of services that would be widely usable by many aspiring and upstart companies.

If memory serves me right, they started with just a Compute (EC2) and a Storage (S3) service. That's it! It did NOT start out wanting to be Operating System for Building New Age companies. It found a couple of highly valuable services that they were using themselves. It is worth looking at the early history of AWS.

The Evolution of App Stores as a Platform Product

Apple and Google are two of the most innovative and powerful valuable in the world, and yet very different in their approaches. Apple tended to be closed, secretive and perfectionist. When the iPhone/iPads were launched, there wasn't an "open" app store, let alone an "open" operating system. Steve Jobs apparently took much longer than than his legendary predecessors Bill Gates and Andy Grove to figure out the power of "open" platforms.

Google, being open and collaborative, acquired Android in 2005 and open sourced it soon thereafter (arguably its best acquisition ever given the reported $50M price!). Indeed, Android led to an explosion of smartphones as OEMs around the world as phone handset manufacturers adopted it in droves. That further drove more application developers to build innovative new apps for smartphones.

Regardless of the paths both companies took to enabling application developers, they ended up creating a lot of value by enabling creation and distribution of apps. Some would even go as far as saying that despite the brilliance of the individual iPhone and iPad products, the real value capture and ongoing success for Apple happened as a result of the iTunes. These kind of peripheral components are called "complements" (aka enablers) and beautifully discussed in the Architecture of Platforms in HBS. There's a fantastic podcast by A16Z on Companies, Networks and Crowds that's worth listening to.

Most people know about the evolution of Facebook as a dominant platform that connects the world. However, Mark Zuckerberg discusses in this Masters of Scale podcast that they didn't intend to be a platform but evolved to be one!

Are there companies that inherently start out being "platforms"? Yes, there are. Often, many multi-sided marketplaces or networks behave like Platforms - for example, Stock Exchanges like NASDAQ/NYSE, Payment Networks like Visa/Mastercard or an e-commerce platform like eBay.

In my opinion, the big distinguishing characteristics between these kind of products and services vs. the former examples is as follows.

Can producers create new value that the "provider" of the platform could NOT have anticipated?

If so, the value creation will be far more multiplicative in nature.

What's common across all the dominant platform products?

• They all start out solving for and dominating a very small number of use cases (e.g. Facebook's network, Amazon's EC2/S3, etc.) Those use cases prove that there may be other consumers and producers that might benefit from those. Fast growing, high quality, organic demand is a good proof point that they have a system which might be worthwhile for others.
• They build on cleanly architected systems with open APIs and a usable data framework.
• They have a viable and sustainable business model for their core use cases.
• They are eat their own dogfood. E.g. Merchant.com was launched on a "primitive" version of AWS years before it was commercially available for anyone else.
• They have a close feedback loop, esp with producers since they are the ones that are going to create non-linear value creation.

Despite the romance of building a product platform, I would recommend early stage startups focus on solving repeatable use cases while building their prospective "platform-to-be" on open and a scalable data/API framework. The platform will be built over time!

About the Author -

(Amit Somani is a Managing Partner at Prime Venture Partners, an early stage Venture Capital firm based out of Bangalore, India. Prime VP invests in category creating, early stage companies founded by rock star teams. Prior, Amit has held leadership positions at Makemytrip, Google and IBM. He is also deeply engaged with the early stage startup ecosystem in India and actively volunteers with iSpirt, TiE and NASSCOM. He tweets at @amitsomani)

This article was originally published on Linkedin