More platforms than the spice girls

Look at those shoes!  Now look at the rush of social networks looking to follow Facebook’s lead and by offering API’s and turning themselves into platforms for third party applications.  This is unlike the shoes, in that this type of platform is not useless and I’m betting it is no fad.
It is of course true that Facebook was first to market with this initiative back in May and that everyone is now jumping on the bandwagon and copying them, but for my money that is more because the last 4-5 months have shown what a great move Facebook made, rather than a case of everyone following the latest fashion.
You probably know this already, but just in case, there are now 5,500 third party applications running on the Facebook platform.  It is understandable that other social networks want to emulate that success.
Myspace has announced that they will launch their platform next week, rumours are that Google will follow suit in November, Ivan Pope reports that hi5 has similar plans, I have heard Michael Birch say that Bebo has similar plans and finally LinkedIn is also heading in the same direction.
All of which is great.  It is great for consumers who will get more apps and a better experience out of their social networking and online time, and it is great for entrepreneurs and VCs as the wave of innovation that is coming from Facebook spreads to other platforms.
In particular I am thinking that if a successful Facebook app can build audiences in other social networks then they will eliminate the problem of being dependent on Facebook.  That will a) increase their value, and b) give them leverage in any negotiations they may need to have with Facebook (or any of the other networks).
People mean lots of different things by the word ‘platform’ though.  Marc Andreessen wrote a great post on this a couple of weeks ago.  He defined three different types of platform, rising in complexity to code and utitility for developers.
  1. The Access API – developers can pull data out of the application and mash it up in their own apps running on their own servers which users access from the developers own site.  E.g. Flickr.
  2. The Plug-In API – developers can write apps to run on their own servers but which users access from inside the parent platform/site.  E.g. Facebook
  3. The Runtime Environment – developers write apps that run on the servers of the parent platform/site and which users access from inside the parent/platform.  E.g. Andreessen’s Ning and Second Life.
This brief summary doesn’t even get 10% of the way into what is a very complicated subject, and if you are interested you should read Andreessen’s post in full.
That said, I am starting to question his view that level three platforms are the way of the future.  (Not something I do lightly given that his accomplishments and mine wouldn’t get on the same chart, even with a logarithmic scale.)
My question comes from the observation that if I am running the same application on a number of different platforms (say inside Facebook, Myspace, Bebo, hi5 and maybe even a social network running on Ning, or even Second Life??) then wouldn’t I want to run and host that in one place centrally, on my servers?
That way I can have one set of code for the application logic which is translated for the different social network platforms by a layer of presentation logic (and as Arrington points out providing that presentation layer might be a big opportunity in its own right).
All of this is becoming important as Myspace has said that their API will let developers run apps on their servers, which sounds like a Runtime Environment as per Andreessen’s definitions.  I haven’t seen any details about how the other networks plan to implement their ‘platforms’.
  • ivanpope

    And also Tagged are promising a ‘platform’. I forgot to blog it, it’s up there now!

  • And also Tagged are promising a ‘platform’. I forgot to blog it, it’s up there now!