Why we don’t hide the front door handle inside the Batcave.
25 January 2008 I’ve just completed another round of usability testing for some clients. At times like these I find myself reflecting (yet again) on how deceptively difficult usable web design really is.The key problem this time was that old favourite, hidden functionality.
I was testing a B2B online application, effectively a VERY large catalogue site with ordering and online invoice payment functions. The owners had found that uptake was below expectations, and anecdotal feedback was that the site was ‘slow and complex’.
When I did the test sessions I found that users did in fact say this. What they objected to was the ‘pick and add’ cart shopping model. This is all very well in a B2C situation, where retail shoppers might buy 2-3 items and the repetitive steps of finding items and adding them one by one to the cart aren’t too arduous. But when you’re ordering 25-50 items for a shop it’s a bit of a chore.
But in fact, I eventually realised (thanks to a really experienced user), the site had been built with a solution. Users could ‘pre-load’ a range of commonly-ordered items into any number of ‘ranges’, one for socks, one for undies… In effect, these were ‘pre-loaded template carts’. Users merely had to check items in the range to load them into an active ordering cart. No more searching across the whole site for correct styles and sizes - if users added the item to their ‘range’ whenever they first ordered it, the job was then done for next time as well.
Magic! The only fly in the ointment of cleverness was that the developers then made this crucial function effectively invisible.
The main navigation on the left side had a heading called ‘My ranges’ - which in almost all cases was below the page fold line, due to an excessive proliferation of much less important links higher up the navigation bar. As a result, few users ever saw it, and if they did, its lowly placement gave no clues that it was something everyone wanted.
Worse, the home page of the application, which appeared once users logged in, had a handy three-part flow diagram showing the main steps to making your order. The three main user steps (apparently) were:
Product search > add to cart > checkout.Doh! What’s missing here, people? No wonder no one knew about creating a range… it’s being kept secret!
It’s easy to lampoon this kind of thing, but the simple truth is, if designers don’t spend time with users finding out what their experience of a site really is, then they’ll never realise when their ever-so-clever functionality is actually completely inaccessible.
Batman has lots of cool stuff in the Batcave, which he uses to fight crime. How useful would all that cool stuff be if the only door handle was hidden on the inside of the front door?
Labels: IA, user experience, web design
posted by Bruce on Friday, January 25, 2008,
,
Why can't the farmer and the cowman just be friends? or... Why every town needs its own Marshall
17 January 2008 Plenty of organisations are starting to realise they need to clean up their websites from time to time – so they hire in the usability posse. Like US Marshalls on the lawless Frontier, they call the Pinkerton Agency and hire some detectives from ‘out East’ who sweep in, clean up the outlaws who’ve been raiding the railroad, and then ride out again. Job done.Or is it…?
I recently went back and read over a website evaluation I wrote a year ago, and then had a look at the website itself. I was flattered, they’d implemented almost all my recommendations. Clearly my massive invoice had had an effect on them! But after a couple of minutes poking about I was forced to ask myself, is this a fully user-centred website?
Well… not really.
The problem, kids, is this. UCD is an iterative process. You do it, you wait for the dust to settle, you do it again. As many times as it takes: “How many times?”
The answer to that question is: “Well, partner, how much string have you got?”
So anyway, this site was much better. It now has one set of navigation, which shows the second level pages in each section. The page headings generally match the menu headings. The home page content focuses on user needs, rather than corporate self-inflation.
But the execution of the recommendations wasn’t done by people who know what user experience really means. The information architecture is still confusing, with 13 top level headings instead of the eight I recommended. The top level pages in each section don’t point clearly to the pages on the next level down – instead they’re as splattered with links as the survivors of a paint factory explosion.
But worst of all, apart from the homepage, the writing has not improved one jot. No topic sentences, no judicious placement of key words for SEO purposes, and more random bullets than a drive-by shooting.
Reading this kind of stuff is tiring, because your brain is doing two things at once, reading the words, and trying to remember them long enough to make sense of entire paragraphs at a time. Good online writing flows like Guinness, you don’t notice it happening at the time, but by the end of the glass, you know what you’ve been drinking.
I’m going to have to reform the posse and ride back into town to clean up the cattle rustlers we missed last time - while we were tidying up the railroad.
Maybe this time the clients will decide they need to hire their own lawman to keep the peace in Silver City when the Pinkerton Men have all gone home. There’s really no replacement for having a web content manager who can actually manage the content in a consistent and user-advocating way.
Contract resources can ‘make it nice’ for a brief period, but if the website is genuinely alive, it won’t stay ‘nice’ for long. The clients need to realise they are committed to an ongoing process, which they either manage properly in-house, or keep tap-tapping regularly on that telegraph to call the hired guns back again from St Louis.
Labels: IA, information design, usability, web design
posted by Bruce on Thursday, January 17, 2008,
,
