Does usability stifle innovation?
I'm not generally one of those usability people who grabs whatever Jakob Nielsen writes and links to it. It's not that I have a problem with Dr. Nielsen's work, but I generally think that, if you're going to take the time to read my blog, then the least I can do is to take the time to have an occasional original thought. This week's Alertbox got me thinking, though, and it's a subject I find I keep coming back to, so pardon me while I think out loud.In a piece about user annoyance, Dr. Nielsen suggests that we should stop using drop-down menus for state lists (on shopping carts, in particular) and go back to straight text fields. Now, I'm not debating his data; I've been amazed more than once at how many people still have no idea how a drop-down menu works. What bothers me is a much broader question: when does usability risk giving into user habits so much that we stifle innovation?
Let me give an example that might help illustrate the point. You may be aware of the origins of the keyboard. The original mechanical typewriter had to be designed in such a way that it wouldn't jam, and this led to the QWERTY configuration. The key layout that we still use to this day is thought by many to be inefficient and suffers from usability problems. If, however, you were to test users on their current keyboard and an entirely new configuration (even if that configuration was more efficient, cognitively and mechanically), they would undoubtedly perform much better on the original keyboard. Why? Simply because that's what they're already used to. Of course, the switching cost (both in time and dollars) of moving to an entirely new keyboard might be high, but what if that new product really were better? By relying solely on our current habits and even testing data, we'd never know.
Ultimately, much of usability is about habit, and that makes sense. If I had to distill my Ph.D. in cognitive psychology into one sentence, it might honestly be "people are creatures of habit". When, though, do we cross the line from adapting systems to people into simply pandering to people's bad habits? At some point, if we only rely on what users already know, and don't allow them the learning curve to something new and potentially better, we reject innovation, and such a rejection is ultimately detrimental to future usability.
This is a difficult question, and I'm not suggesting that it has a one-sized-fits-all answer. As usability specialists, though, I think we have to sometimes take a risk and not constrain ourselves to what already exists. People are creatures of habit, but we're also creative and adaptable, and we need to have the opportunity to discover new and better tools. Without that opportunity, there wouldn't even be a worldwide web, let alone web usability problems to ponder.
Ron Denholm
· Wednesday, March 28QWERTY's usability had also been attacked by the amazing Microwriter, invented during the filming of Zulu by Cy Enfield. Users could potentially type much faster with this invention, but usability slowed down if you needed to insert symbols other than text.
Dr. Pete
· Wednesday, March 28Ron, that reminds me of a similar, one-handed chord "keyboard" I saw about a decade ago that never materialized. It's a fascinating idea, and really makes you wonder what it's going to take to replace something as ubiquitous as the keyboard.
AL
· Thursday, March 29I guess that applies to the clickless user interface that never saw the light. It's an amazing experiment, if you ask me. I myself would consider taking the plunge and adapting to such interface, but would everybody else be willing to? Very unlikely.
Mike Maddaloni
· Thursday, March 29Though not scientifically, I have observed both children and seniors use a GIU interface and surf the Web. Children, with no fear, have mastered all of the elements of UIs. For the seniors, those who took the time to understand the interface elements (e.g. the drop-down lists) are much more comfortable than those who just plain refused to, and admit it.
But what do I know - I have been using computers for almost 30 years, and I only use 2 fingers to type!
mp/m
Dr. Pete
· Thursday, March 29Wow, that's wild stuff, Al. I'm not sure I'd want to do away with clicking, but I love seeing people experiment, and every once in a while it takes a crazy idea to really change how people do things.
Dr. Pete
· Thursday, March 29I'm also guilty as charged, Mike. I am a very skilled hunt-and-peck typist, which I suppose means that I've become highly proficient and an innately inefficient task. I'd learn to touch-type, but I'm holding out for that new keyboard arrangement :)
Peter
· Thursday, March 29Interesting post. I do agree, sometimes the fact that we are used to something makes it usable... not the fact that it is actually usable. I remember in my first year of uni we used to come up with very wild (innovative?) interfaces, because our knowledge of usability was limited to non-existent. Nowadays there is a lot more sober, usable but at the same time, boring approach.
Michael Zuschlag
· Thursday, March 29It's a myth that usability engineering can only make things that users are used to. To make an innovative "unintuitive" interface: 1) Decide on an acceptable amount of training or experience for learning the innovative interface; 2) Give test users that amount of training/experience; 3) Test, evaluate, and improve the interface. Ultimately verify that performance of trained/experienced users on the new UI is sufficiently superior to performance on a conventional design to justify the effort of learning the new UI.
Dr. Pete
· Thursday, March 29Michael, I don't mean to suggest that usability makes innovation impossible, only that our tendency to rely heavily on testing and existing user habits sometimes makes it easy to get set in our ways. The process you've laid out is a good one, but it's also difficult and not enough people pursue it.
Greg Scowen
· Friday, March 30Good discussion.
Another self-confessed 2-finger typist... but I don't know that I want to learn a new keyboard layout, I have a hard enough time going back to QWERTY after my trusty Swiss-German QWERTZ.
There are more issues with drop-down boxes than I like to count. The state drop-downs often cause problems based on the lack of options for non-US residents too. A regular text-field is certainly a good option, but I suggest something better may even be available in a suggestive-text-field.
A list of possible options could be kept somewhere out of sight, and when a user starts typing in text it makes suggestions of what follows. Similar to Google's suggestive search feature.
Meanwhile... I agree that we are often constrained as usability designers by the expectation, or our perception of them, of our users. Is this something we need to overcome, or to accept? As user-requirements for example.
Dr. Pete
· Friday, March 30I did some data analysis when I used to use text-boxes for U.S. states, and I was appalled to find how many of my fellow U.S. citizens had no idea what their two-digit state code was (or, if given a longer box, how to spell their own home state). I'd really like to see Jakob's data on the errors, as I find it hard to believe that entry errors by clicking on the wrong item were higher than those when people were left to enter the data free-form.
tom sherman
· Friday, March 30I do believe that user familiarity is the most important aspect of designing an easy-to-use interface, and this is why the slick Web 2.0 interfaces of the world are hard for a lot of people to understand. They're new. We (the tech savvy) think they're cool, but they (the unwashed masses) think they're confusing. I think back to a friend's comment about GMail: "I don't really trust that it's working because I can't see what it's doing."
Re: Clickless interfaces and such, I think that usability improvements to interfaces will follow things like Enso Launcher (see http://www.humanized.com), which establish a familiar vocabulary for doing computer stuff -- rather than the Minority Report 3d gibberish.
Kendall
· Friday, March 30You can have innovation and usability if you also factor in designing with affordances (Donald Norman).
AL
· Wednesday, April 4I'd like to second Michael Zuschlag's comment and give a concrete example: the new Office 2007 interface. This is a classic innovation of a highly usable interface that, when coming from older version, takes the user an average of 3 days to learn. However, the user is rewarded afterwards with more productivity and easy-of-use.
Point in case: innovation + usability can happen at the same time.