There is a general belief in Silicon Valley that once you have the users, revenue will follow. So that’s what Twitter decided to optimize for, back in 2006. One innovative way of doing that was getting third-party developers enthusiastic about the platform and offering a simple to use API that they could use to build their own twitter clients. And developers did. Twitter grew into a massive service, in part thanks to the great twitter clients that made using twitter a nice experience.
Twitter raised a number of rounds of funding, which was easy because many venture capitalists primarily look engagement numbers, and those were looking good. Recently, however, things have started to change. A company cannot continue on living just by raising more and more venture capital. At some point it will have to grow to be self sustaining, so that’s what Twitter started to work towards. How do you earn money from a free service? Advertising! Where do you advertise? On any user-facing part of the service.
A significant part of users had since moved to using twitter almost exclusively in some third-party client that Twitter had no control over. Most people _never_visited the website. So, twitter needed to move its users back into its own clients and started acquiring popular clients such as Tweetie and Tweetdeck. That wasn’t enough, so they had to limit the public API more and more. As a result, third-party twitter clients start to die completely or have to raise prices to be self sustaining. It is to be expected that eventually, most third-party clients will be discontinued due to increasingly stringent rules on API usage.
This is not good for users and even worse for the third-party client developers.
Still, none of this should come as a big surprise when analyzing Twitter’s incentives. Initially the incentive was to get as many users as possible, recently it shifted to generating revenue. All of Twitter’s users got used to using the service for free, the chances of successfully switching to be a paid service are likely to fail, so: advertising. For advertising you need control over the user experience, and so we are where we are today.
Enter App.Net. App.Net decided to do a little bit of incentive hacking. They asked themselves: how can we build a Twitter-like service where the service’s incentives better align with what users want? Their solution: go paid. A percentage of the income goes toward running the services, another to third-party application developers, creating an incentive to develop great App.Net clients (users get a monthly email asking them to rate the clients that they use).
Clearly, this is a very different incentive than the Twitter’s. Sure, for App.Net, too, getting as many users as possible is beneficial, and it will try everything in its power to do so. More users is more income, but it benefits current customers too, as Metcalfe’s Law says: a single fax machine is useless, but the value of every fax machine increases with the total number of fax machines in the network. App.Net wouldn’t need the millions of users Twitter has. It can be sustainable with far fewer. App.Net tries to emulate Twitter’s growth model by opening up the API for developers to develop third-party clients, but supercharges it by paying developers to do so. From the get-go, App.Net has a sustainable business model and it profits only from getting new sign-ups that stick around (they pay monthly or yearly).
App.Net is an interesting experiment in incentive hacking. Are people willing to pay for a service to ensure its incentives align with their needs? Time will tell.