The folks at UIE are really good at what they do. I've recently attended a seminar of theirs focusing on usability and found it entertaining, informative, and immediately impactful to my work in user experience. And I enjoy their regular newsletter, too.
But I have to take exception with two false analogies in this latest newsletter. First is Jared's, who opines:
I agree in part: 20 years is quite long enough to figure out how to make a home video recorder simple to use. But I have to disagree with the analogy to toasters, which suggests that today's toasters are simple to operate, instantly understandable, and pose little or no barriers to use.
If you're talking about this toaster, then the analogy is largely correct. It's simple to operate: just drop your bread, bagel or waffle in, press down on the "plunger", and wait. No muss, to fuss.
Compare this to the toaster at my office. Granted, it's a toaster oven, so it's supposed to offer more functionality than your garden-variety pop-up toaster. But this thing has two shelves, four independently-operated heating coils, three knobs, at least two different ways of turning it "on", and settings for nearly 150 different "functions". What's sad is that in my office, filled to the brim with brilliant software coders, designers, and user experience professionals, no one can figure out how to use this thing. Someone finally brought in a plain old slice-style toaster (like the one above), just so we could get through our Monday morning bagels without a fist-fight.
So on average, I'd have to say that today's toasters are not all that easy to use. And I don't think you could say that, in the 110 years since the first electric toaster was invented, they have become any easier to use. Yes, the core functionality has been pretty well laid down, but frustration still abounds.
Second, a reader had previously written in to discuss the results of the CUE usability studies run by Rolf Molich. I love these studies, not least because they give me a free look in on how other usability folks run their shops.
This particular reader responded to Christine Perfetti's write-up thusly:
To which Christine responded:
The problem with this analogy depends on, well, you're definition of the word "underlying". Cars (like websites) have many different parts which interact with each other, and the end user (be it the driver or the site visitor) might have a problem with them that isn't necessarily tied to a single underlying problem. If one assumes that a site has a fixed number of identifiable, discreet "issues" that anyone with the proper tools and training can detect and resolve, one might be surprised to find different experts identifying the same problems differently. But if one assumes that usability issues are fluid, non-discreet events that are triggered by a series of variables that sometimes interact to confound some users.... then the results of the CUE studies shouldn't be all that surprising.
So too with automobiles: there may not be a single underlying problem, and while one mechanic may recommend replacing a valve while another calls for a whole new transmission, a third might suggest that both are unnecessary if the driver would just stop gunning it up to 120 mph while negotiating sharp curves.
I think usability testing is especially useful at identifying "problem occurrences". That is, in testing you can discover when a user has trouble, say, adding an item to their shopping cart. But that's merely observation of the symptom, and discovering the underlying cause is, like auto mechanics and medicine, part science and part art. Agreeing on a diagnosis (much less a prognosis) for a serious usability test takes consultation with other professionals, and iterative design.
Usability testing experts may find many different problems (symptoms) in a website (patient), but that doesn't mean they aren't all honing in on the same underlying problem (disease) and recommending solutions (prognosis) that will eliminate not only the problems they find, but others they don't directly observe.
