Twitter Annotations: developer and user nirvana or the biggest nightmare
So yesterday was big announcement day for Twitter at their Chirp Developer Conference which followed some recent announcement by the company that didn’t sit well with the developer community. The highlight that got most of the attention of course was the fact that Twitter finally has the beginnings of a monetization plan called Promoted Tweets.
However that wasn’t the only thing to grab the attention as two other things that they announced are going to have a huge impact not just on using and developing for Twitter but potentially for the way we communicate on the Web. The first one is what they call user streams which are intended to make Twitter a real time delivery system of messages. This will have an immediate impact on the whole Twittersphere especially once the developers can incorporate it in their clients and other Twitter related services.
As cool as that might be what followed was even more important and service changing feature being added to Twitter. I am of course referring to the Annotation feature that is suppose to go live for developers sometime next quarter. Mahendra Palsule at Skeptic Geek has a really good post on the feature that is well worth reading but first we need to understand what this Annotation idea is
The feature will allow developers to “add any arbitrary metadata to any tweet in the system.” So, just like a tweet can today be transmitted along with information about which other tweet it was in reply to, or what location it came from, or what application it was created on, now Twitter will allow developers to make up new stuff. Twitter is looking to see how developers use Annotations before it creates any sort of taxonomy for them, Sarver said.
As Mahendra notes in his post – this could be the greatest thing to happen with Twitter or it could end up being a big nightmare for the company.
The short story here is that Annotations could be the biggest treasure chest of goodies for everyone either using the Twitter service or developing for it. It is in effect exactly the SuperTweet that Robert Scoble talked about recently. While he may have been looking at the idea more from the advertisers point of view Annotations would be a boon to those using Twitter as part of their Social Media outreach programs. It would be able to let users know more information about who’s Twitter message they were reading or replying to.
Analytics and researchers would go nuts over it because of the amount of information that they could pull from a Twitterstream would double, triple or even grow by ten-folds. Mahendra provided a few ideas that really would only be scratching the surface
- Show me tweets from users above an influence-rank threshold
- Show me tweets from users who have at least x followers or x list memberships
- Show me tweets from a specific geo-location
- Only show me tweets that contain links or pics or videos
- There can be interesting mashups and visualizations based on such metadata.
Just as there is so much that could be realized from Annotated Tweets the flip side is that they would be next to useless in providing any of the above let alone be some sort of SuperTweets. The problem with the whole Annotations idea is that it is application specific which means that any annotations added to Twitter messages by TweetDeck would only be usable to people using TweetDeck. The same applies to any of the hundreds of clients and services built around Twitter. It would also apply to any that Twitter itself might add.
If this happens the promise of the power that Annotated Tweets could bring to the table gets blown away like a puff of smoke.
There is of course a possible solution that could put the nightmare aspect of this to bed but it would mean that developers would have to come together and create a common ground around Annotations. Consider this almost the ultimate type of open sourcing where the developers brainstorm and create a basic set of Annotation namespaces.
I realize that at some level all developers are competing against each other for attention and users not to mention against Twitter itself. However if there is one thing I think that all developers could agree on it would be that the user on Twitter is of prime importance. If this is indeed the case it would behoove them to work together around Annotations as by creating a common set of namespaces they will radically reduce a large portion of user frustration that is inevitable as confusion over what the hell Annotations are and why they aren’t a common experience.
Not only would this greatly benefit the user experience it would also be a good thing for Twitter the company as it would see that their concerns over user experience (such as the one that lead them to buying Tweetie) would be for the most part laid to rest.
The sad part is that I don’t think this will happen and as much as I really hope that I am proven I think we could be headed into some problematic time when it comes to Twitter and the clients we all like to use.