Startup general interest

Most platforms start as applications

By March 10, 2010 5 Comments

Creating a platform business is the dream of many entrepreneurs and their VCs – and it is easy to understand why.  Successful platforms have huge scale and customer lock in and, to an extent at least, if you own a platform you can sit back and watch the dollars roll in as other companies build their businesses on top of yours.

So I’ve been thinking about how successful platforms come to be, and it seems to me that most of them start off as applications.

I’m writing this post today having just read about Facebook’s latest step towards being a platform business imminent launch of a share location feature which will work both on site and via an API.  As we all know they started as a profile surfing application business and then communication service via messaging and news feeds.  Then when they had sufficient volume of users to be attractive to third parties they became a platform as well, first with their applications API, then with Facebook Connect and now with location.

The Google story has similarities.  They started as a web search application and once they had a sufficient volume of traffic to attract third parties they became an advertising platform.  This began with Adwords on their own site, and then when they had a sufficient volume of advertisers to attract third party publishers they added the Adsense product.  It probably isn’t an exaggeration to say that their other products since have been funded from the profits generated by these two.

Two more quick examples.  Twitter always had platform ambitions but it was the success of the microblogging service which convinced developers to build Twitter clients and other apps based on the service.  And then on the business side – Salesforce started as a SaaS CRM application and used their success and scale to launch the platform.

There are counter examples, of which Ning is the best I can think of right now, but these are riskier ventures requiring much more investment before proof of value can be generated.  I suspect the application first route to platform development is a new one enabled by the internet.  On previous platforms aggregating huge volumes of users and cross selling them to new services were both much harder.

UPDATE: in more of the same Google are now leveraging the traffic on Google apps, gmail and google calendar to launch Google Apps Marketplace.

Reblog this post [with Zemanta]