Poly-hierarchical Navigation

After what has been a tough week, perhaps even a tough couple of weeks (for reasons which many of you will be aware), it was good to sit in a meeting today thinking a little further ahead than 60 minutes.

The directgov website is starting to look pretty good – apart from the neat MC Escher arrows in the orange bar at the top – and the folks who put the content together are highly capable and committed people. Their efforts may be going a little in vain though – I sat in another meeting today with a variety of suppliers who would tell me that they hold the torch of e-government close to their hearts but none of them could tell me what colour the top bar was in directgov, i.e. they hadn’t bothered to visit the site.

Directgov’s navigation is divided up into topics (e.g. Money in Government) and Audiences (e.g. Parents) and the idea is that you can navigate to content that you need through any of those devices – so being in Parents means you can access content that is technically owned by Money, such as benefit payments. Sounds simple so far – after all, that’s what links were designed for.

The editorial team want to go a bit further though: they want any content to be accessible from any place in the site and have that content branded and located as if it were owned by the part that you are in. At a simple level, say “parents” content is coloured green and “disabled people” is red – if you were in parents and looking at content that was technically owned by the disabled people’s team, the content would still show up in green so that you didn’t lose track of where you are and so that the experience remained the same.

Sounds easy, until you start thinking about what happens if a small part of the content needs to be changed to better fit the context of where it is (e.g. a title change, or a change in a few words)? Who would approve that change and how?

Who owns the content and what happens if the owner changes it so that it no longer fits properly in some other sections? How would you know that?

What would a search engine see? One instance of the content or many? Where would it point you to?

What would you do with related links? Would they point back to the original location of the content or to new content in the new audience section?

How does something like this scale? It seems fine for a few bits of content to appear in several places in the site (after all, news articles appear in various sections in the BBC – but they do appear exactly as they are without any changes), but what if we’re talking about the whole site being accessible from any one audience type or topic type? Does that make a mockery of the navigation?

If you do this, something that I think is called “poly-hierarchical navigation” does it mean that, actually, you haven’t got any navigation principles left at all? Some of these issues arise because we’ve built directgov with tabbed navigation on the top and then further, specific, navigation on the left hand side. Was that a wrong decision that has complicated things now? Or are we trying to solve the wrong problem?

During the meeting I pulled up Google and checked how many references there were to “Cold Weather Payment”, which was the example content we were using to work it through. There were 248. My theory was that we actually only wanted one, so rather than surface content in multiple places in directgov, we should first concentrate on getting it right in one place and then kill off the other references so that it was clear where the definitive content was. 248 is not a bad number. “Disability Living Allowance” has 16,900. That’s an awful lot of duplicated content.

This whole thing needs some thinking about – and sometimes, when you’re close to a project, you’re the last person that should be thinking about how to solve it. After all, if you came up with the design in the first place (I didn’t in this case, but the other folks in the room did), you’re likely to focus only on what’s good about what you’ve done and you may miss some of the flaws. So you need to do research and perhaps bring in some new thinking to see what to do next. If you’re doing that, you might as well take the opportunity to think through a whole bunch of other points about the design and inform your next set of deliverables so that you can see how to roadmap the next releases.

It was, in many ways, a relaxing end to a challenging week. I’ve got to get my thinking cap on to see what the right way forward is for this kind of navigation. I don’t know that anyone else is trying to solve it, and hopefully I’ve explained it clearly enough so that people will understand what I’m getting at.

Leave a Reply