Information architecture not simple, even for experts
Subscribe now for $100 (23 issues) and save more than 37% off the cover price!
Get the latest news from Computerworld delivered via email.
Sign up now
You can train formally to be an information architect but it’s helpful to have a certain instinct for it in the first place; “the kind of mind that likes to tidy up stuff”, says Donna Spencer, author of a forthcoming book on the subject.
Spencer was visiting New Zealand last week from Canberra. She is an independent consultant currently advising Australia’s Department of the Environment. She and New Zealand practitioner Trent Mankelow, last week presented their views on information architecture to an event in Wellington organised by Mankelow’s company Optimal Usability.
In a web-and intranet-oriented business world “findability” is an important subset of usability, says Mankelow, quoting some admittedly dated statistics from IDC, pointing to billions of dollars lost by Fortune 100 companies from customers and staff not being able to find the right information on the website or in the company information system.
Information architecture can simply be described as “helping people to find stuff and then use that stuff”, says Spencer. It’s a matter of understanding the ways people categorise information in their own minds and there can be little commonality in that in large and varied groups of customers and staff.
Even the two speakers found areas of healthy disagreement.
Mankelow says one good way of organising information is according to the audience; a university site might have sections for staff, undergraduate and postgraduate students and prospective students.
Spencer says she finds this less useful, since many people in different groups might be looking for similar information; a task-oriented classification, around “what do you want to do?” is often more useful. This can then branch into more objective categories such as chronological or alphabetical order.
Before the talk, staff of prominent New Zealand companies discussed the merits of such hierarchical click-through schemes against search engines, which may yield a quicker result but may come up with totally irrelevant references. Internal search engines on company websites are often so disappointing, they agreed, that many users resort to Google with the site name as an additional search term.
A classic way of finding out how people’s minds categorise is to give them cards with the names of narrowly detailed topics and ask them to sort the cards into categories — either adopting previously named categories or defining and naming their own.
If reasonable agreement does not emerge from the first eight or so users, Spencer says, it’s probably pointless to continue with a larger number in the hope of finding commonality. Mankelow, by contrast, sees some point in taking larger samples.
The exercise can be done with real cards or by drag-and-drop online, but the latter loses any interaction among members of a group, says Spencer, admitting to surreptitious observation and note-taking of the conversations that go on during live card-sorting.
Once categories have been arrived at, appropriate labels must then be chosen for them.
“Use short, consistent and jargon-free language,” says Mankelow.
“I disagree with ‘short’,” Spencer counters, saying too short a label can be uninformative or confusing.
A member of the audience raised the question of copying information architectures from someone else’s website. There are certain good principles which will be productively refined by judicious cross-fertilisation of ideas, the meeting agreed, but on the other hand, a good information architecture represents considerable intellectual investment and ought to be protectable.
However, as with many ICT ideas, difficulties are seen both with copyright, which protects the expression, not the idea, and patent, where there is little New Zealand precedent in software and questions of prior invention and obviousness to be navigated.