Power Twitter: Plumbing v. UX

Michael Arrington justifiably praises Narendra Rocherolle (83º) on the release of Power Twitter, but for perhaps the wrong reasons.

I agree that the search capability should be wired into the UX like Power Twitter has done, and I like all the little tweaks to the UX that have been made (although the 'peek' capability -- a mouseover of a contact's name shows that contacts last few posts -- is not working at the moment because of API limitations: all Power Twitter users are being trated as one user, at this moment).

But that's not the big blinding insight that jumps from Power Twitter. The big news is that what is displayed does not have to be limited to the lowest-common-denominator text stream that is burbling through the Twitter pipes. The text -- with all the various encodings we stick in the stream -- can be massaged by the tools, as if they are a lens on what the text means not what it says.

Power Twitter, originally uploaded by Stowe Boyd.

In the screenshot above you see that the bit.ly shortened URL I actually pushed into my Twitter stream (using the bit.ly bookmarklet, while looking at the TechCrunch piece) is expanded in the presentation by Power Twitter into the title of the post. Note that it doesn't simply expand the shortened URL into the full URL: it resolves it into a human-friendly text handle for the thing being discussed.

Likewise, all the other "codes" we are pushing through Twitter streams could be expanded in useful ways:

1. User Names -- Power Twitter's 'peek' feature is a good example of how the UX can be expanded to fetch more information about a person. I would also like to see their profile in this 'peek' interface.
2. Hashtags -- I'd like to see more information about hashtags, again, perhaps by mousing over.
3. Tickers -- $Stocktwits 'tickers' (see Twitter Is The New Bloomberg) is another convenient code that could be expanded in place, instead of making a user travel to the www.stocktwits.com website.
4. Other Codes -- Various people have proposed other codes that could be used by Twitter or other services to encode various sorts of information. I have suggested the use of '#*' as a metatag denoting reviews, '!' for tasks (as in assign to-dos to yourself or other people), others have suggested '+' for assigning attributes to users (@stoweboyd +whimsical +massive +discursive). In this case, Twitter could either build the associated services, lease them out to other companies, or allow users to choose which service they want to use. For example, I might opt for Remember The Milk for tasks, while someone else might opt for Google tasks..

The point I am (laboriously) getting to is that Twitter does not have to limit itself to displaying the actual text passing through the plumbing. They can opt to display pictures, expand conversations (a la Tweetree), resolve URLs, and decode all manner of other codes. We don't have to be limited to the 140 characters at the presentation level even if we are limited by the plumbing to that form factor.

And Arrington is right when he says that the average user doesn't want to go to 76 other sites to see these various acts of disambiguation taking place: we want it right in the center of our Twitter experience.

Twitter has ceded the client space to others, principally because the user experience at the website is so bad. (Who ever thought that separating 'sent' and 'received' direct messages on different tabs makes sense? After all, we got past that with email 20 years ago.)

What they need to do is to create an innovative new user experience, based on a dramatic redesign. Best would be in a client running on the desktop, but a web client is ok as a first pass. They need to consider how to continue to support innovation around various codes -- like $Stocktwits tickers -- while figuring out a way for users to have a consolidated user experience even if they are using (or receiving) multiple codes supported by multiple services. This is similar to the themes I expored in Fred Wilson on Making Twitter Smarter, And More Valuable, a few weeks ago.

(PS If you Twitteroids are listening, I return to SF next week, and I would be happy to sketch out these ideas in more detail, if you'd like.)