August 2, 2007
The examination of what makes for a good user experience is absolutely vital, and as you’ve probably gathered, I – like many others – think we ‘providers of tech’ often get basics wrong. In fact, depressingly, web/tech products seem to get things wrong more often than they get them right.
In order for this wrongness to happen, one of two things must have been an issue during the project process:
1) Belief that the end user isn’t terribly important
2) A ‘project interface’ issue – a gap in expertise between various departments or a (perceived) lack of cash or time.
It is nearly impossible to believe that the first is true – that project teams put together end-user tech without considering end-user needs. Sadly though, for one reason or another – usually IT teams thinking the technology is *everything* (“it’s really cool the way we’ve interfaced the Z60 box and the AR39 switcher. Besides, *I* understand the user interface so why won’t Dawn in Accounts…?”) – it does happen.
Gaps in project expertise are related but different, and easier to understand. Typically this occurs between designer and techy: each assumes the other is responsible for usability (or just ‘the user’) and in the end it turns out that neither focuses on this, the most crucial part of the product.
Now’s the time to bring in a real-world example. I buy my tickets for London on thetrainline.com for delivery to a ‘fast ticket’ machine at Bath Spa, as I do every week. The tech between the website and ticket machine is enormously impressive: a timetable lookup, a credit card transaction, a central database of bookings networked to any station around the country. Nice, and my thoughts go out to the poor people who had to do it.
So why is it that when I arrive at the (newly designed) ticket machine, I can’t see how to collect my ticket? There’s lots of station names, and a few other buttons, but nothing that tells me where my ticket actually is.
In the end I, like many frustrated users in front of bad UI’s both on- and off-line, start pushing random buttons. Eventually I try one that says ‘Tickets on Departure’. It works.
Sorry? What? Pardon?
‘Tickets on Departure’…..?
– ‘Prepaid tickets, press here’
– ‘Bought online?’
– ‘Pre-ordered tickets’
– ‘Collect your tickets’
– ‘Ticket collection’
– ‘Fast ticket collection’
…or any other sensible, obvious, meaningful choice of words. ‘Tickets on Departure’…
There’s some other badness going on – a touchscreen keyboard that isn’t arranged in QWERTY layout and the phrase ‘print journey’ rather than ‘print ticket’, but ‘Tickets on Departure’ is horribly bad. Crucially, it’s also right at the beginning of the process: your commitment is low, and peer pressure (the crowd building and tutting under their breath behind you) is high. Your natural response when that button isn’t right there in your face? Abort the tech and go ask a man instead.
I stood for a few minutes and watched other people with the machine. It turns out I’m not alone – the majority gave up, some asked a member of staff (who looked like he’d been asked before..). A couple pushed the last button remaining to them, as I had, and battled through.
Now I’m not big on mainframe stuff but the infrastructure required to do all that on/off-line talking feels to me like maybe 5 million quid, minimum? I’d imagine each ticket machine is probably £30k.
Seems a shame, doesn’t it? All that effort and money culminating in a crap UI which frustrates the very people it’s apparently built for.
Tell you what, I might see what I can find out about the project process in building these particular machines, and I’ll then post about what (if anything) I find…