Writing and Sizing Twitter

Twitter LogoFor some time now, everyone’s been crazy for Twitter, a kind of hub for digitally checking-in with your friends, where everyone alerts one another of what they’re up to, sometimes as frequently as from moment to moment. If that’s a bit of an obscure description, it’s because there’s nothing else quite like it. Actually, ‘cute’ may be the best and most succinct descriptor I can come up with.

More Web service than Web site, I had a hard time remembering to post the short, punchy updates that are Twitter’s principal currency until the advent of Icon Factory’s free, desktop-based Twitterific utility for Mac OS X.

Twitterific puts a persistent kind of ‘heads up display’ right on your Mac OS X screen so that your friends’ posts are immediately available, and that you can easily add new posts yourself. No more having to load the Web site, or remembering to visit that tab in your browser where you’ve got Twitter.com running.

Smaller, Faster, Better

That Twitterific changes my interaction with the service is a good illustration of the difference between two software “postures” that Alan Cooper describes in his book, “About Face 2.0,” a must-read for any interaction designer. By virtue of its containment inside of a Web browser, Twitter — and most any Web site — is something of a sovereign posture application. Cooper defines this kind of software as…

“Programs that are best used full-screen, monopolizing the user’s attention for long periods of time… Sovereign applications offer a large set of related functions and features, and users tend to keep them up and running continuously.”


Even if your browser isn’t fully maximized, it tends to dominate most of your screen real estate, and it almost undoubtedly remains open and running most of the day. If you think of Twitter.com as a just one set of the functionality available through the browser (and admittedly, this is shading Cooper’s definition a bit), it’s plausible to argue that it demands a level of attention out of scale with what it truly requires. As cute as it is, it’s not a site that you often want to monopolize your screen, even for very brief periods of time.

Twitterific, by contrast, takes the exact same functionality, and presents it in an auxiliary posture, where it occupies much less screen real estate and only partial attention. Cooper writes…

“The auxiliary program is continuously present like a sovereign, but it performs only a supporting role. It is small and is usually superimposed on another application…”

The interface is more compact and, in its translucency, can be said to be less obtrusive. This simple but significant change in functionality presentment has a dramatic impact on the way I interact with Twitter: rather than relying on Twitter.com’s built-in reminders to “nudge” me back to the site, its persistent availability on my desktop makes it easy for me to monitor incoming posts and more effectively spurs me to add new ones. Since installing it, I’ve been using Twitter with much greater frequency than before.

Reading, Writing and Interfaces

There’s also an adjunct change that accompanies Twitterific, and that’s a slight alteration in the nature of my inputs to Twitter. It’s something that also happens to demonstrate another interaction principle at work: the concept that interaction is informed by language as much as by visual cues.


On Twitter.com, the main interface clearly asks “What are you doing?” That text is displayed in large type directly over the text input box, and it remains there after I’ve submitted a post, effectively prompting me for another post. Additionally, the stream of posts from my friends and me is labeled, in similarly large type, “What You and Your Friends Are Doing.”

The persistence and size of these labels are strong textual cues that always encourage me to answer the questions directly. Therefore, my early Twitter posts were straightforward but uninspired answers, e.g., “In a meeting,” or “Walking the dog,” or “Standing in line.” It got old pretty quick.

By contrast, Twitterific’s textual cues are much more subtle, and less persistent. It does ask, “What are you doing?” as a ghost label inside of the text input field, but that disappears as soon as I begin typing my answer. More importantly, it provides no label (aside from the user name) above each post.

Write What You Know

The result is that the posts that I make — and view — through Twitterific are less inclined to be literal answers to the question “what are you doing?” and more likely to be unorthodox, even smart-alecky non-sequitirs. Their display and their language logic are less constrained, freer to take different, unexpected forms.

This has allowed me to have a lot more fun with Twitter, to feel less constrained by the mundanity of Twitter’s single, standing question, and to treat posts as opportunities to say whatever I want. Have a look at my recent track record for a sampling. To be sure, they’re no more insightful or brilliant than what I was writing before, but they’re more fun to me, at least.

All of which proves the point that interfaces are written as much as they are rendered. Outside of the alteration in software posture described above, the text is the single most significant change between Twitter.com and Twitterific. Everything else stays the same; we are, in fact, dealing with exactly the same application logic on the back-end. And yet there’s a subtle but material change in the nature of the resulting input nevertheless. At a low level — and admittedly, Twitter is a small-time application, at least for now — this is fascinating stuff, and a good reminder that more than just visual design determines good experiences.

  1. It is quite cool to see many UIs be “distilled” into concentrated, almost ethereal windows or presences, almost appearing as a/in a glance. Interesting way of dealing with the heavy steady clutter on most users’ desktops.

    And as you point out, this often changes the interaction model pretty drastically. I imagine this trend will continue to accelerate bigtime as more and more phones improve their web access as well.

  2. Although not knowing you, from reading your twitterings I had to add you there. Hope you don┤t mind!

  3. Very interesting observations! I think of “Twittering” as “microblogging”, meaning I think of it as a place where I have 141 characters worth of space to note things that are somewhat interesting, but not notable enough for my blog. Since I’ve applied that personal definition, I find it easier to ignore the “What are you doing?” question that, as you rightly say, often results in banal responses. Perhaps I’ll try downloading Twitterific to see if my stance towards it changes again.

  4. It’s a small subset of writing that I call “Small thoughts” — the kind of random single sentence thought that typically runs through our minds daily by the minute or second or all at once.

    Twitter is clever in that it brought that idea to the internet masses. I haven’t gotten on board yet but it’s something I mentioned and have posted a few times to my own site (the small thoughts that is). What hooks you obviously is the social aspect and getting a small glimpse into what people are instinctively sharing or thinking about at any given moment — both random and compulsively throwing out there.

    I think I’m going to have to finally try it out.

  5. while I like the idea of Twitter it’s really not that unique. I find the best way to explain it as the Facebook status message.

    while Facebook does not IM people it does update the feed on the main page.

    Even if you think that the two are completely different things, if you watch what people post on both Twitter and the Facebook status message they are very similar types of messages.

  6. I’m the author of Twitterrific.

    Your observations about sovereign vs. auxillary posture are spot on. One of our design goals was to ensure that the application could be as unobtrusive as possible. And adaptable to different work environments/conditions.

    Unfortunately, web browsers ofter fail in this regard. It’s an all-or-nothing affair (e.g. you see the window/tab or you don’t.)

    In my mind, the most interesting thing about this project was its initial motivation: lack of a good notification mechanism.

    Web browsers are inherently limited in this regard. They don’t have any way to draw your attention. A sound, an animation, or changing the state of an element on the page doesn’t help if the window is buried in the dock or task bar.

    As web applications begin to mature, this problem will be addressed. Hopefully in a consistent and user friendly way.


  7. My initial thought on Twitter was “Why should I care what my friends are doing right now?” and, even more, “Why would any of my friends give a shit about what I’m doing right now?”

    I went to the Twitter site a few months ago to figure out exactly what the buzz was about. After 10 minutes looking at the site and reading every word on the pages, I still had no idea what the hell it was.

    I sent an email to them basically complaining the site (like many other sites) did a pretty poor job of explaining “what the hell Twitter is” on their home page, or any other page for that matter. It looks like they fixed it now — the new home page is really good at this.

    Still, it’s a little hard to understand the “value proposition” beyond lending a little bit of a sense of self-importance to each user. It’s another “the world revolves around me” app. Maybe I still don’t get it.

    Maybe it’s just the kind of thing you need to actually use to understand or appreciate. I remember when AIM came out a lot of people wondered “I don’t understand. Why not just use email?” But when they tried it once or twice, most people I know switched instantly. The momentum of the worldwide adoption curve led to ubiquity pretty quickly, and now I can’t live without IM.

    Speaking of Instant Messaging, when I see Twitterific, I wonder: How is it any different from instant messaging? It seems just like sending an annoying IM to all your buddies every few hours, except using yet another channel. Again, this seems like more of a nuisance to me. But hey, again, that’s what I thought about AIM at first.

    Good point about sovereign vs. auxillary. I’ve been talking about these concepts for a while without using the terminology. I’ve been calling the “sovereign” approach the “full-focus state“.

  8. Chris: the question of why someone would care about what Twitter provides is basically what I meant when I say that, right now anyway, it’s a small-scale application. The only reason I can give is that it’s fun and mildly amusing, much like a sliding tile puzzle.

    Craig: Thanks for chiming in! You guys did a very nice job with Twitterific. Your work — and the points you made here — also go to another issue: as Web applications and their attendant services mature, we’ll see more and more seamless desktop clients, like Twitterific. Personally, I think this is the way to go.

  9. Are all the people insane?
    It’s the major sign of stupidity 2.0.
    People love to say what are they doing all the time? Even worse, somebody is looking all the time what other people are doing all day?????
    What a group of idiots!!!

  10. Well, Gustavo, Twitter isn’t for you, then. 😉

    I’m surprised I like Twitter: I really don’t do much IM. But I’m interested in how Twitter binds small virtual communities. I find myself interacting more deeply on a traditional level (work collaborations, email, and calls) with people who are on my Twitter list.

    In addition to the social aspect, we’re using Twitter for group notifications: server problems, calls for help and advice, etc.

    The key is who you invite or follow. I’m fortunate in that my Twitter group is fairly disciplined. 3-4 twitters per person per day is plenty. Beyond that, big, chatty groups become unmanageable. Twitterrific is poorly suited as an IM client.

    A strength of Twitter is its platform independence. We have people accessing via web browsers, the Twitterific client, and mobile devices.

    Anyway, it has promise.

  11. I agree with Gustavo that the desire to publicly display oneself is a (modern) phenomenon that deserves some reflection. On the other hand, the ability to do this is significant at the level of being able to display _presence_ status, ie: what am I doing (or planning to do) where. This can be very helpful to colleagues, collaborators and clients in helping mediate communication and facilitate connection. It can also be helpful for alerting people to something happening _now_ such as an interesting radio broadcast.

    I also agree with Joshua Kaufman’s point about flow, and we need some new interaction models to be rolled out to address this.

    Khoi didn’t mention that Twitter also supports reading and writing via instant messenger clients/protocols, including Gmail and AIM — another auxiliary posture alternative to the browser. In the next version of MacOS X one could of course create a Dashboard widget that embedded the Twitter home page. There is already a nice Dashboard widget, Twitgit created by Ben Ward.

    Finally, a related product is Jaiku (jaiku.com) which aggregates your Weblogs, pictures and bookmarks (Ma.gnolia, del.icio.us, etc) and creates a status badge for your site.

  12. I’m with Naz and Stefan. It’s all about the “small thoughts.” Twitter is essentially enabling the kinds of people who like to have many clever away messages on IM, continuously updated throughout the day.

  13. Why twitter? Well, I don’t do it to make new friends, that’s for sure. I do it because it gives one person specifically and quick note about where I am and what I’m doing. I find it pleasant that when I call my wife mid-day, she will say, “So, how did the meeting with Alan go?” based on something I wrote on twitter earlier that day.

    Does anyone else read my twitter entries? I don’t care if they do (like a postcard), but they’re primarily intended for her.

  14. @essl: It’s been suggested that Twitter could make a nice mobile to-do list.

    @khoi: there’s nothing quite so surreal as sitting in a barber shop in London and getting a text message that tells me you think the sun is super bright 🙂

  15. Twitter essentially decouples away/status messages from IM, which seems like a fairly cool idea.

    I’d love to see an IM client that embraces this and hooks its away system up to Twitter; to me it’s the exact same usage paradigm.

  16. I treat Twitter as the IM status message in steroids… not only can I read clever status messages, I get a history of those too, and even an RSS feed.

Thank you! Your remarks have been sent to Khoi.