Thinking About… Navigation

Over the weekend, Pete pointed out a piece about “mystery” navigation on websites that doesn’t tell people where they’re heading. He suggested it might be rather relevant to me, seeing as d4d™ does exactly that. (And I recently discovered that the rollovers for showing where you’re going when the mouse hovers over the button doesn’t actually work in IE – I’d always kind of assumed it did, and thus hadn’t tested it. Mea culpa)

Anyway, it’s something that I started off as a project on here, something where I was interested to see how people used it. Initially it actually had the look of buttons – using nasty Javascript rollovers and images – where you only knew where you were going once the mouse was already over the button. But it was ugly, so when I went over to CSS for the main bit of the site, I just kept it with text but still on the same theory, not knowing where you were going ’til the mouse was over the button.

You see, my theory is that most people actually navigate based on position, rather than necessarily reading the button that tells them where they’re going. (And yes, I realise there’s a fallacy to this for first time visitors – just bear with me, OK?) I’ve worked on sites where the navigation was truly dynamic, and altered itself so that the “most used” buttons for user ended up at the top of the navigation tree (Man, you should’ve seen the database behind that bugger) and so the nav was different for every regular user, and also changed as you used it.

However, it turned out people hated this – about 75% of the regular users would already have their mouse pointing at the place on the screen that held the button for where they wanted to go, and when that changed, it annoyed them. It’s the same kind of theory that annoys people when supermarkets change the shelf layout, or store position for certain items. People like patterns. People stick to patterns (for the most part) and thus reinforce those patterns.

So d4d™ went the other way – if you knew where you were going, it didn’t matter what the button said – and in fact it could say nothing at all. In many ways it’s an experiment that’s worked – but it’s also one I wouldn’t put onto a commercial site I was doing. When it’s commercial, you really pretty much have to make sure you’re open to everyone, with a particular focus on the first-time visitor. If they don’t like what they see, or find it hard to get round, then they won’t become a repeat visitor.

(I’ve currently also been trying to persuade work about this one, that dynamic nav might sound cool and look cool, but long-term it pisses people off and is epically counter-productive)