I ‘discovered’ a new information seeking behavior today and realised that I have never heard any discussion of particular design strategies for it. The behaviour was that of repeating a visit to a particular site or page.

Here’s how it happened:

I was doing some rewriting for an IA workshop. In this workshop, I talk about information seeking behaviours, give examples and ways to design for them. I include the 2 traditional ones – known item and exploratory information seeking. Pretty straightforward. I also talk about when people don’t know what they need to know, and how to design for that.

I was thinking about examples for each so looked at my browser history to remember what I had been doing recently. I realised that almost all of my web use didn’t fall into one of my ‘categories’ of information seeking – almost all of it was returning to things I had already visited (posts and articles I’d already read, sites that had design features I noticed, products I had previously looked up). In a week of browsing, there were very, very few times that I had started an information-seek from scratch.

That gave me a big surprise as I realised how much I revisit information, and how much I rely on memory, my URL history and del.icio.us to re-find information. It all made me wonder:

• how prevalent this behaviour is
• what design strategies we could use for this type of task

I was thinking, as I frequently do when I write an article, about how blogging has affected the way that I write.

I noticed a few years ago, when the content of many of the blogs I read was pointers to other information, that bloggers like to create posts as quickly and easily as possible (no news there). One of the most common blog entry patterns is:

• find an interesting article
• find an interesting sentence or short paragraph in the article
• drop the paragraph into the post, with a short lead-in and a ‘thank you’ or ‘from’

This means that, in order to have your article linked to and read, it is essential to have a good sentence or paragraph that represents the point you want to make. While this is a general principle of good writing, it is particularly important when writing for distribution via blogs.

It is something that I do after finishing writing an article – go back and make sure that there is something that is easy to pick up.

So in that vein, here are the magic paragraphs from an article that I recently co-wrote with Tina Calabria � 10 ways to continuously improve your intranet:

The amount of work involved in designing a new intranet or redesigning an existing intranet is minor compared to the time needed to maintain an effective intranet over the longer term. In fact, it is common for the initial excitement of a new intranet to fade away as the reality of day-to-day maintenance and the challenges of improving the intranet become apparent.

For this reason, intranets tend to go one of two ways after launch – they either stagnate with few new features added over time or become masses of unstructured content and functionality created in a random fashion.

I haven’t written anything substantial here for ages – not because I haven’t had good thoughts, but I’ve been terrorised by spam.

My various anti-spam measures have been fighting with one another. I was running MT-blacklist on this weblog and Knowspam on my email. I couldn’t get email notifications when people left comments (unless they were on my good addressees list on knowspam) so had to rely on finding comments with MT. But MT (2.64) only showed 5 comments at a time, so despamming was pretty tiresome. Anyway, the upshot was that managing spam was starting to take up a lot of time.

So I decided to upgrade MT on the vague marketing promise of ‘better comment management’. That went reasonably well, but then I found out that I needed a new version of MT-blacklist. Tired, I set comments to ‘moderate’ and decided to finish up later.

So I got home today and found 16,000 comments waiting for approval. Yes, I thought that it was a bug in the interface as well, but it was real (taking away MT-blacklist left me open to the spambots). Hours later and I have removed all of the comments (400 at a time), installed a new version of MT-blacklist, configured, re-configured, stuffed around, deleted files and finally got everything working.

I hope…

Now I’m too tired to write. Maybe on a spam-free day I’ll have something to say.