The other day Twitter killed my Twitter app.
This isn’t a post written by someone who is bitter and angry. It is from someone who loves Twitter and really enjoys developing with the Twitter API. As a Twitter fan-boy I also want to see Twitter continue to grow and advance. The problem is that I have notice a trend- Twitter taking ideas from independent developers and integrating them into their own service, rendering our apps obsolete.
Now you might argue, perhaps rightly, that Twitter should have been fulfilling the same service my app provided from the beginning. In fact, my app really only was a work-around for Twitter’s own limitations- and didn’t effectively do much. Still, one site ranked it among the top-ten Twitter apps of 2010 and at least one of my users thinks that FeedTwit was better than Twitter’s current solution. If I was more cynical, on the other hand, I would think that Twitter sold someone the exclusive rights to do what my app already did.
At the same time, my app wasn’t the only pirated independent Twitter app recently. I used to love http://tweetree.com/ – but Twitter integrated so many of their ideas into New Twitter that there is no advantage to using them any longer.
My point: it is not in Twitter’s interest to keep pirating ideas from its developers.
Why Twitter should encourage independent developers
- Encourages creative thinking and innovation– this is perhaps the biggest advantages for Twitter in rolling out the API to independent developers, like me. However, when Twitter takes our ideas and integrates them into their service, they take away the reason for us to innovate. Why should I put a year and a half into developing my Twitter app when I can be afraid that suddenly Twitter will make my app obsolete?
- Independent data– my app provided valuable, independent data from my users. In my case I could tell you what time of day or what day of week people were most active on Twitter. This was useful not only to marketers- to know when to share their message to receive the most eyes, but also to other developers- to know when to shut down their apps for service. I can’t provide this valuable information any longer.
- Less for them to support– my Twitter app was far from perfect. It seems there was always something going wrong or breaking. One time, while testing, I even accidentally sent 10,000 DM’s (and accompanying text messages) to my personal cell phone. But if I spend my time on working through these issues, Twitter can instead focus on providing a stable network which benefits everyone.
What should Twitter do instead?
- Keep it simple– I personally think the Twitter website should be very simple (like before New Twitter). Let independent developers customize its use based upon the laws of supply and demand: if an idea is good it will survive, if not, it will die. I understand why Twitter would want to innovate on their own website: fewer and fewer people are using Twitter.com and they will (eventually) need eyes on their website if they ever want to make money from advertising on their web page. Let me propose another means of generating revenue- charge developers for white-listing their app. This would allow people to develop apps to a certain point and then force us to monetize it in order to pay for the white-listing.
- Keep it broad– recently Twitter has said its going to discontinue supporting XML. Previously it has threatened to devalue RSS. With limitations like this it not only requires developers to limit the ways they can interact with the API but it limits which developers can participate (without learning a different way of interfacing with the API). Why would Twitter want to put all their eggs into one coding basket?
- Keep it consistent– recently Twitter identified that it was going to run out of numbers to represent each particular status. The fix to this was called “snowflake” and it was a real fiasco. Many apps broke (including my own) despite Twitter’s warnings to make sure “snowflake” wouldn’t affect it- I didn’t think it would, but I ended up being wrong. A lot of heartbreak could be solved if they simply used the versioning system they have already built into their API. This would have kept an API consistent through the transition to larger status IDs.
What is my next step?
Back to square one. I initially called my app FeedTwit because it was designed to convert RSS feeds into Twitter direct messages. Only after I rolled this out did I realize most people were using it to convert their Twitter mentions into DMs (and then text messages). Now I am back to only supporting RSS feeds. Even if there isn’t nearly as much interest as before, you could still use FeedTwit to:
- Get notifications of mentions for other Twitter accounts you manage.
- Keep up with a Twitter list
- Keep up with other social networks, for instance LinkedIn, through your Twitter account.
- Get news, weather, and even traffic reports to your cell phone by Text message, from your Twitter account.
- Track a keyword or hashtag discussions on Twitter.
- Take any RSS/Atom feed and send it to your Twitter account as a direct message.
I have changed the focus of FeedTwit at this point. Now FeedTwit does something that Twitter recently stopped doing: providing RSS feeds for individual Twitter accounts. Ironically I am now adopting ideas that Twitter has abandoned.
UPDATE: Apparently the FTC is now investigating: http://www.bloomberg.com/news/2011-07-01/ftc-said-to-have-begun-antitrust-inquiry-into-twitter-s-developer-policies.html
UPDATE: Yet another idea Twitter “borrowed” from third-party developers: http://support.twitter.com/articles/20075772
How have you seen Twitter take over ideas from independent developers? Am I being too harsh on the Twitter API team? Add your comments below.