Despite not working in the field anymore I do keep seeing interface design issues popping up recently.
In my copywriting feed – feel free to subscribe to it on the right-hand side of this blog, it takes lots and lots and lots of copywriting feeds and syndicates them out – comes a piece on the rate at which most people scan Google pages. It averages out at 140 words in 6.4 seconds. This can surely only be enough to pick out a few words of interest and, as Chris Hoskin says in his post, it must be too quickly for a ‘rational’ decision to be made.
My take on it is that people will tend to return to that search page if the link they choose isn’t right for them, so it’s probably several ‘chunks’ of 6.4 seconds which isn’t quite as random, and through this repetition people start to make better decisions as they become more familiar with what’s presented.
So perhaps the real way to judge an interface really isn’t in its immediate intuitiveness but eventual familiarity. My father – who used to program in Mobol using punch cards – is trying to teach my 70+ year-old auntie how to use a computer. I think it’s admirable that she wants to get into it but he says it’s a real eye-opener. He has to teach her how to use the mouse, how to click, how to open and close windows and start applications.
These are second-nature operations to us graphics-interface-savvy users but you tell me: what exactly is the difference between single-click and double-click? Why do we single-click menus but double-click icons (I’m talking Windows here)? I can only come to the conclusion that you double-click a picture and single click text, but where does that leave the Start menu, with its combination of icons and labels? Why don’t we double-click hyperlinks? Suddenly it doesn’t seem quite as obvious anymore.
Way, way back I started on the path of enlightenment with a ZX81, graduating to a ZX Spectrum. The ‘interface’ for that – a very quirky command line – was really pretty dreadful, with arcane and obscure key combinations for picking out commands and operations. Yet I was able to churn out code really quickly with it. This wasn’t because the interface was well designed. It was because I got used to it.
The key is a combination of learning curve, consistency, and the adaptability of an interface to work around the user. If the curve is shallow and you can do things in several ways without getting lost, you’re working with a good interface.