Large, well-funded companies are increasingly dominating the little world of Facebook applications. Slide and Rockyou for all-purpose applications, Zynga and SGN for social games, iLike for music. These companies have also moved aggressively as Facebook’s rival social networks began opening up their platforms to third parties over the last half a year or so.
Like mountain men in the industrializing West, a lot of smaller developers have been getting forced out — or forced to join larger companies. But some have not only hung on to their Facebook applications, they’ve also jumped to these new platforms. There, they also trap users with the most competitive game or the silliest poke.
As MySpace, Google and other social networking companies move in, I decided to ask one leading developer, Blake Commagere, how he was managing the changing world, especially the new frontier of the rival platforms. Commagere, who you may know as the creator of applications like Vampires and Werewolves on Facebook, tells me he has created leading applications on each network’s platform — with details on how he did it.
VentureBeat: First, what has the effect of OpenSocial been on the Facebook developer community? What do you think about Facebook’s move to open-source its platform today?
Blake Commagere: I personally wish Facebook had done this before big players like MySpace committed to open social – it’s unclear to me how many social networks will end up implementing both Facebook and OpenSocial. Bebo was in a position where they wanted a live platform on a timeline sooner than OpenSocial would be ready. Now that OpenSocial is ready and out there, I’m not sure why MySpace would bother implementing FB’s platform.
We now have two standards and if the huge markets that MySpace, hi5 and Orkut represent aren’t enough to convince all the Facebook developers out there to learn a new standard, I’d be surprised.
Frankly, Myspace could have come up with a completely different platform that didn’t conform to OpenSocial or Facebook and while a lot of developers may have complained (myself included), I would have still done everything necessary to port to MySpace. It’s a huge market – you’d have to be insane to ignore it.
VB: Do you think Facebook might merge its platform with MySpace and/or other social networks using OpenSocial?
BC: It would be interesting to know whether Facebook andOpen Social will take steps to ultimately merge their platforms – that would be a huge win for developers like me. In the meantime, I’ve already built a framework that allows my apps to run on both OpenSocial and Facebook, and I’ve considered open sourcing it, but I think it may be too constrictive to most application developers.
For example – if you aren’t using an MVC [software] architecture and PHP smarty templating, then conforming to my framework would probably feel like as much work as porting to another platform. It’s useful for me since I designed it for my apps to run on both Facebook and OpenSocial, but probably not useful to the rest of the world. What this means though is I’ve had to write an abstraction layer between my apps and the platforms since each platform has some differences – and while that abstraction layer was fun to create, the hope the development community had was that porting apps between social networks would be trivial, and we haven’t gotten there yet.
VB: So, how is that framework doing for you?
BC: My apps have taken a leading position on each platform – some of the platforms have more viral channels available and as a result the apps fare better on those platforms. I think the success of my apps has been consistent relative to other applications that have ported over as well.
As a breakdown:
On Bebo, the apps have done well – they are growing consistently, but not as quickly as they did on Facebook. It took me about 2 months to reach a million installs. A million users in 2 months is a great accomplishment and I think the only reason that that isn’t making headlines is many of us app developers have been spoiled by a million users in 10 days experiences and stories. The Bebo team did a great job of making ports easy – porting to Bebo took me 9 hours. Best. Decision. Ever. As far as differences in the network, the tone of Bebo is such that it allows for more anonymity in profiles. More anonymity results in a social graph that is not as strongly connected as the social graph in, say, Facebook and as a result, we see lower conversion rates through those viral channels. Think of it this way: You’re more likely to try out an application when you get the invite from “Blake Commagere,” a friend you know and have spent time with. Compare that to an invite from “EvilBoarder42,” a person you may not have met yet, but friended as a result of a mutual interest in snowboarding. Granted, on networks that permit more anonymity there are still tons of real friendships, but there will definitely be more weak links in the graphs the more anonymity is permitted.
My apps are doing okay on MySpace. They have not yet opened up newsfeed/notifications and so apps are relying on comment posts and bulletin posts to spread. Since MySpace has historically been very focused on self-expression, this is a good strategy, but widgets have been using those channels to spread for years, so the conversion rates are not as strong as they could be. Furthermore — MySpace does permit more anonymity than Facebook and as a result conversions do take a hit there. Once MySpace has newsfeed/notifications opened up, I think apps will explode in growth.
And, a very important point on the MySpace use of comments and bulletin boards as viral channels is this very important point: Over time any viral channel will result in diminishing returns from a user aquisition standpoint. (Think about the first time you got junk mail “oooh” with the last time “ugh”). The first times widgets (think slideshows from slide and rockyou) were using the bulletin boards on myspace as viral channels, they spread like wildfire. Now, since those channels have been leveraged significantly in the past, they aren’t going to produce the results they used to as users have been trained to look for unwanted messages in those channels, which is why we are all eager for MySpace to open up newsfeed & notifications.
On hi5, since they have opened up the notion of newsfeed/notifications, the apps have really exploded in growth! One thing I am eager to finish is a translation of my apps – I think that given the predominately international user base on hi5, I’ll see some huge benefits there.
VB: How have different implementations of OpenSocial affected your development time? What do you think about the new version of the OpenSocial technical specifications for its member social networks, version 0.8?
BC: Personally, I haven’t felt benefits or pain from the different platform customizations – most of the customizations seem to be just things leading up to 0.8 features anyway, rather than a departure from spec.
Lastly – I’ve only glanced a little at the v0.8 reference and it seems to be missing something that many app developers have wanted for a while – the ability to query for the name/etc. of an arbitrary user ID.
For things like leaderboards, or discussion boards, that is a critical feature – and it makes it hard to build a community around your apps within a social network without that type of functionality. If it is currently in spec, then I’m thrilled it is there and embarrassed that I overlooked it. But if it isn’t in the spec, um… why?
VentureBeat’s VB Insight team is studying marketing and personalization...
Chime in here, and we’ll share the results