Yesterday, I was privileged to do a guest post for Louis Gray and chose to do something a little bit different and wrote about the issue of whether APIs can be used to generate revenue for services such as Twitter, It was something I had been thinking about for a while but the news that they had throttled the unauthenticated API calls which are the life blood of many third party applications brought it in to context:
"It has seemed obvious to me for a while that an ideal aspect of a business plan for a social networking service such as Twitter would be to charge partners for premium access to the API, but once you have started down the free path, is it possible, or wise, to backtrack and start charging?"
You can read the full post over at Louis' blog here.
I was a bit worried about this post as it could have been a potential banana skin; instead it garnered some good responses both in the comments and on FriendFeed so I wanted to follow up on a few issues raised as a result.
"The route to success is recognising that Twitter's power is as a service"
Jeff Sonstein raised the above point that a successful way forward would be for other services to aggregate the data from Twitter for their own users - much as FriendFeed is doing. Jesse Stay argues that developers may not be willing to pay and go elsewhere such as identi.ca and that, instead, Twitter should consider premium features that they can offer their users for a fee.
As I mentioned in my post, services like FriendFeed are just as guilty as having no clearly defined business plan; the social web is running on venture capital but how close are we to running on empty? It is unlikely that many services will be willing or able to pay for API access and, although, both users and developers may move to a free alternative how then will thatservice support itself - we will enter into a vicious cycle.
Sudha Jamthe suggested that Twitter could "go to traditional media companies to build new services upon their API, similar to what Sphere did with NY Times" but I don't see the potential for the same kind of relationship here. Admittedly, Twitter has a large body of data but a lot of that is quite frankly useless and banal - even some relatively interesting content is worthless when taken out of it's original context.
Perhaps the main use for the bulk of Twitter data is in trend analysis which could, in turn, be used to plan advertising campaigns or for targeted marketing. Is this why Twitter bought the best search tool and has limited the API requests for all the rest? Could they make money by selling their analysis results or by charging for full access to enable companies to perform their own?
The folks at Twitter have never completely ruled out advertising so full trend analysis of your conversations and your friends could provide the only way to effectively target you as an individual and provide ads that you are actually likely to act upon.
Whatever the future brings I feel it will be quite different from the way we are operating now. Users may still get free access and developers might only be charged a small amount (perhaps an initial one off fee to get a licence to use the API) but advertising or data analysis is probably the way way to get most bang for the Twitter buck.