The State of the Platform: Update

by Jesse Farmer on Wednesday, May 7, 2008

My article about The State of the Facebook Platform has been spreading through the blogosphere like a game of telephone. Lots of people have chimed in with their own opinions.

I wanted to write a follow-up post to clarify my opinion and address some of the responses.

What I'm Claiming

My claims are simple and uncontroversial. I observed two things: one, the activity level in the Facebook forums is a fraction of what it was four months ago; two, Facebook apps launched today are much less likely to succeed.

The trends for these two observations are highly correlated and exhibit the same peak around February 2nd, 2008. What happened around that time? One, Facebook began instituting increasingly demanding and arbitrary developer policies. Two, other networks began launching fully-featured competitors to Facebook's platform.

From the high correlation, the timing of events, and comments from people working in the industry, I concluded that developers are less interested in Facebook today because there's less return on their investment of labor.

What I'm NOT Claiming

I'm not claiming that the Facebook Platform is unhealthy. Nor am I claiming that it was a bad idea for Facebook to implement the policy changes they did.

I'm certainly not claiming that any of the data implies either of the above. Indeed, it's still possible to find success on the Facebook Platform. It just requires more effort than it used to.

Also, most emphatically, I'm not talking about Facebook users. The article was only about developers and their decision to create software for Facebook, not about Facebook as a whole, which is still seeing phenomenal success.

Other Hypotheses

The most common alternate hypotheses for these trends was summarized by Jeffery McManus:

This is not a terrific metric for developer activity — it doesn’t measure what you purport to measure. Developers generally view and post to forums when they have problems; if fewer developers are posting to the forums, it may mean that there are more developers who are having less trouble.

I call this the "documentation hypothesis" and addressed it briefly in my original article. I think it's an unappealing explanation for a few reasons.

First, if it were true, we'd expect to see spikes in forum activity whenever a new issue arose on the Platform, especially since Facebook's changes tend to be radical and out of nowhere. The decline in activity is virtually monotonic, however, and the data shows no such spikes.

Second, even if it were true, it doesn't explain the correlation between forum activity and application success, nor does it explain the sudden decline beginning around February 2nd. As an explanation it just isn't sufficient.

Is the Trend Good or Bad?

I understand the tone of the article was bearish, but I was writing it from the perspective of a developer deciding whether or not to commit to the Facebook Platform. There are lots of perspectives, though.

Facebook's Perspective

I believe Facebook is making these changes intentionally. They have a love-hate relationship with companies like Slide. Strategically speakingthese companies got in at the very beginning and quickly cordoned off sections of the social graph for themselves, largely out of Facebook's reach. Messages on FunWall don't go through Facebook, for example.

This is clearly not in Facebook's strategic interest, but they can't just boot these companies out because a significant number of Facebook users would throw a fit. From Facebook's perspective these trends in developer engagement are good because it allows them to reassert control and improve their image as the "high-quality social network."

Facebook Users' Perspective

Let's face it, most Facebook users don't like to be pestered by applications. For them these changes are good. And judging by Facebook's traffic stats it isn't hurting them one bit.

Advertisers' Perspective

For advertisers these developments are universally good. If the bar for application development is higher it means the applications that succeed will be of a higher quality. Nobody wants to advertise on "What color barf are you?" and Facebook doesn't want that application to be front-and-center, either. It just looks bad.

Developers' Perspective

For developers this is a mixed bag. Facebook's cavalier attitude about platform policy means that you're playing on shifting ground. On top of that the changes they've already made mean it's harder for applications to succeed, on average.

Still, for companies like SGN and PlayFish, who want to make quality applications, this means that they don't have to worry about competing with win-at-all-cost, spammy applications.

I just wouldn't recommend developing only on Facebook, as they've shown they're willing to change and bend the rules at a whim and for their own benefit. You know that as soon as Facebook decides they don't like what you're doing they'll do everything in their power to hinder you. Hedge your bets.

Hype Cycle

Don't forget about the hype cycle, either. All technologies go through a phase of inflated expectations followed by a trough of disillusionment.

I'd say we're right in the middle of the trough of disillusionment. Companies like Zynga, SocialMedia, Slide, RockYou, and SGN are going to slug through the slope of enlightenment.

Will we have a social operating system or a revolutionary social commerce system waiting at the end? Probably not. Will we have innovative casual gaming platforms? I'd take that bet.

I'm interested in hearing other perspectives, too, particularly investors' perspectives. Does anyone have any insight on that?

Update: Nick Gonzalez, formerly of TechCrunch and now of SocialMedia, makes a similar point about the hype cycle. I also like the Darwinian nature of his post's title: "The Thinning of the Herd." Heh.