:Ben Metcalfe Blog

Twitter and the need for a developer exit strategy

Today’s announcement of an official BlackBerry client for Twitter is interesting for a number of reasons. The app itself is actually the one that RIM have been building in house for sometime. If you look at a BlackBerry you will see that apps like Google Talk, AOL Messenger, Yahoo Messenger are all written by BlackBerry manufacturer RIM, but distributed as ‘the official BlackBerry client’ by those companies. Clearly, this is RIM’s modus operandi and they like to be in bed with the service they are building for.

I therefore wonder if Twitter didn’t actively go out and look to make an official app, but instead RIM propositioned this to them as the way they build these apps and Twitter decided it was worth making the move.

However, what is significant is I wonder if this development will encourage them to now create ‘official apps’ in other areas, especially on niche or emerging devices and platforms. Fred Wilson’s blog post, while he is now playing down it’s significance, seems to suggest this.

Twitter needs create the platform exit strategy for developers

If you want to start ‘filling the gaps’ in your proposition yourself, you need to create a way for developers to remain loyal and focus their interests elsewhere in your ecosystem.

Let’s look at Facebook. They used their platform to add value to their proposition and fuel the rate of growth. When they met the inflection point where that was no longer necessary they regained control slowly restricting the app’s ability to communicate with the user and relegated the apps to a separate tab that no one would ever click on.

To pre-empt the frustration this might cause they very astutely created Facebook Connect – which enabled developers to continue to work on the opportunities of the Facebook platform but away from Facebook itself (which is what Facebook wanted). In many ways, this was an exit strategy for developers who had invested time and experience in the platform.

If Twitter begin to effectively close up the opportunities in the platform by creating ‘official apps’ across the board, or ‘plugging the holes’ themselves then they too need to create the exit strategy for developers else they will get pissed.

Fred Wilson suggests developers look at social gaming, enterprise, discovery, and analytics. Those are great ideas, but Twitter will need to do certain things and release certain APIs to create these ‘exit strategies’ to encourage developer focus away from clients and into those areas.

Here are some of my thoughts on what Twitter needs to do to make these ideas viable and valuable for developers:

Social gaming
Spymaster showed us that social gaming is too noisy on Twitter, at least as it exists is today

  • Twitter need to create a filter or similar mechanism to allow social gaming to exist so users can participate without fearing they are creating noise for the followers. I actually think makes this the toughest one to solve.

Enterprise
Twitter is already building product in this area and that will concern developers.

  • Twitter needs to be more transparent with what it is and isn’t doing in this space so developers can feel safe investing their time building apps Twitter isn’t going to compete with.
  • An official solutions marketplace and preferred consultant network, while stodgy, would help startups connect with enterprise

Discovery
Rate limiting is still hurting this area, despite fire-hose/garden-hose being available. Discovery typically needs volumes of data.

  • Remove rate limiting for certified discovery apps
  • Increase who can get the true fire-hose as it doesn’t appear to be open to anyone
  • Create a Twitter Query Language (TQL?) to allow new types of data interrogation to occur that doesn’t require the user to maintain a mirror of the twitter status database.

Analytics
Twitter is also working on projects in this space, which again will understandably concern developers. It also has the issue of (potentially) needing volumes of data like Discovery.

  • Twitter should be transparent with their roadmap here too
  • Release more API’s around user metadata, such as rankings and other computable values from the master DB
  • Removing rate limits, creation of a TQL would help massively here too

What it comes down to is that developers have created mainly presentation-layer focused apps because the API set available to them has encouraged this. If eyeballs and page views are related to monetization, then I can understand why Twitter might be wanting to reign in what is going on this space.

However, before it can urge developers to go build in different spaces, in different parts of it’s value chain, Twitter needs to lay the foundations for this to occur. Creating these new APIs and being more transparent with its own roadmaps in the Enterprise, Discovery and Analytics spaces are all very necessary.

Friend and Seesmic founder Loic Le Meur has some wise perspectives on this.

(Disclosure: I am a shareholder in Seesmic and NutShellMail, both of which are primarily focused in this space)