Tweets in your information architecture

If there’s one thing we’ve learned over the past 30+ years of user interface development, it’s that a given action should be invoked via multiple gestures. For a given command there is no single gesture that everyone can agree on is perfect for the task. Personally, when I’ve used a command more than about 3 or 4 times I start looking for its keyboard equivalent. Other people seem content to use a mouse for everything, but my carpal tunnel starts buzzing if I use the mouse too much. There are context-sensitive (right-click) menus, scroll wheels, pinch gestures and taps on the iPhone – each is a different input mode.

In the realm of communication Twitter is another mode. It is neither email nor IM nor blog, and while it is very popular with a younger crowd it is not for everyone. Frankly you will never find me twittering because I don’t suffer under the illusion that anybody wants to hear my every passing thought. But one of my colleagues has been trying to persuade me – for many months now – that it’s actually quite useful, and he’s been making progress. I don’t dismiss it out of hand any more, at least. For instance, he showed me how a support person could monitor a filtered stream of tweets and each time certain unique phrases show up- for example “I’m having trouble with product XYZ” – then he has the option of being contacting the person and offering to help fix the problem. There’s something really attractive about somebody contacting you out of the blue to help you solve a problem rather than having to wait on the phone for somebody in support to pick up your call. Makes you wonder how long it’s going to last, however. I can measure call rates and customer satisfaction for incoming calls, but it’s harder to track unsolicited (outgoing) offers of support to people. Putting on my management hat, especially if I were a manager of a busy call center, my feeling is that something that you don’t track never happens. So maybe you have a more lenient – less trackable – rule that if there are no incoming calls to Support, your job as a support rep is to troll the tweet feed and see if anybody needs help. Yeah, I don’t think so. Frankly I’d rather just tweet my friends.

The other thing my colleague showed me is that it doesn’t have to be your job at all to help people with their problems – there is a community out there willing to help for free. Once again, they just filter the feed and figure out whether there are any damsels in distress. I imagine this is a great way to meet new people, especially those of the female persuasion. It’s a little like the open source community. Somebody offers things for free, nobody understands why they are doing it, but somehow a community builds up that seems to just do this sort of thing. In the early days it was a mystery why people did it, but these days people do understand a little better what motivates those in the open source community. It’s recognition, a bit of fame – the same reason why some people are willing to be called the “biggest loser” and wear the badge proudly. It may someday be parlayed into real monetary gain, if you get enough of it. Looking at Twitter, I would say this is a huge problem. I am sure there are popular tweeters that got popular because they have a real gift for the medium, but I should think the vast majority of the popular feeds became famous in another medium. Helping a random person solve their problem is not going to make you universally popular. So twitter is not like open source. The motivation is not the same.

So where is the real benefit of using twitter? I think that if you have a small community of friends and you like to stay in touch, then there is some benefit to chucking occasional tweets into the pond and letting your friends vicariously experience some of your thoughts and feelings. Your friends will occasionally offer you help when you need it, because they are your friends and it’s the thing you do for friends. It can’t be your only interaction medium. Long distance relationships fade over time, no matter how much you try to maintain them. Twitter is riding the crest of a big hype wave right now. Once the hype is over and things return to steady state in the twitter world, I see the most significant long term usage pattern being close friends keeping in touch.

So if I am right, then what does that mean for an information architect? Can this information be used in an organizational context? Are there bits of it that can be folded into defect reports, feature requests, or source code change orders? In terms of the content I can’t think of anything. It’s just too hard to fit any useful data, without context, into 140 characters. Of course for the people who are doing the communicating there is plenty of context – it’s the game you saw at the bar last night, or the company picnic, or that boring meeting you just wasted an hour in. Or that you are currently in the middle of. Context fades away pretty quickly, leaving you with a few barren words that are no longer laden with meaning.

The way I see it, the only way that twitter information can be mined for potentially useful information is in determining social circles. User interfaces will get better when the program knows more about you, and information about your social circles might be one of those useful bits of information that it could use. For example, “Alex, I recommend you use feature X because your friend Chuck seems to find it to be very effective.” I don’t know exactly how it will be used until such features become commonplace, but I do know that Google, for one, would love to know everything they can about you. They’re on to something.

Just to reiterate – and perhaps fend off mail – I’m not saying that Twitter has no place in an organization, just that it doesn’t have much use in its information architecture.

Emails probably have sufficient context and information content that they should be considered part of the information repository that should be managed. In my next post I will talk about some interesting research going on in this area.