Tag Archives: birdfeed

Twitter in 2 years

Marco predicts that third-party Twitter apps will lose half of their users within the next 2 years:

“We won’t even be angry at Twitter — we’ll move to the official apps voluntarily, and we’ll look back on all third-party clients like we look back on Tweetie, vanity link shorteners, and third-party image hosts today: as relics of a quickly abandoned past before we all started using Twitter’s better, newer features.”

During WWDC this year, Buzz Andersen gave a great talk at a small venue outside the conference. With the hindsight of several years, he talked about building Birdfeed, the challenges of competing with Tweetie, with his own struggle at perfection, and many more insights on the rise and fall of third-party Twitter apps.

It left me with a lot to think about, and I loved the old stories, screenshots, and related nostalgia. But in closing out the questions & answers, one statement in particular struck me as a nail in the coffin for third-party developers: Buzz revealed that even he now uses the official Twitter app.

Birdfeed, etc.

He snuck it in under a commentary on Alex Payne’s excellent “last post about Twitter”:http://al3x.net/2010/09/15/last-thing-about-twitter.html, but we now have a “Birdfeed postmortem of sorts from Buzz Andersen”:http://log.scifihifi.com/post/1144913920/in-a-way-the-twitter-platform-has-come-full. I’m particularly interested in where Buzz thinks the Twitter app market is going:

“Does this mean that there’s no longer room for third party Twitter clients? My suspicion is that people will continue to make them, but it seems to me that they’re already on the road to becoming increasingly uniform and commoditized as the Twitter experience is more sharply defined by Twitter itself (as my Birdfeed collaborator Neven Mrgan has suggested to me, Twitter clients are going the way of email clients).”

It seems nearly everyone thinks competing with Twitter’s official app is a bad idea. “Here’s Tapbots revealing Tweetbot”:http://tapbots.com/blog/business/say-hello-to-tweetbot shortly after the Tweetie acquisition:

“But this news changes things for us. We probably won’t be able to charge for the app anymore. Who’s going to pay for a Twitter client when ‘Tweetie’ is free?”

I think the problem isn’t trying to build any Twitter app. The problem is building a mainstream Twitter client. The official iPhone and iPad apps, plus the redesigned web site, are so good that it’s futile to go head to head with them. You can’t undercut on price, and they are so well coded you’d need a talented full-time team to out-engineer them.

As I said in “my last Twitter post”:http://www.manton.org/2010/09/next_generation.html, the trick is to look past the API. What would I want Twitter to add to the platform if I had my way? Design an app around that and you might have something interesting.

The theme in Buzz’s post that resonated the strongest with me is the emotional drain that building an app like this can have. The competition is intense. There’s a feeling that if you don’t have every little feature when you ship, you’ll be laughed out of the App Store. That is certainly on my mind, especially as “I intend to push the price”:http://www.manton.org/2010/06/10_ipad_apps.html in this market, and even with beta feedback I’m still not sure how my app is going to be received in the real world.