Usability Means Knowing Your Audience
During my recent redesign, I had to revisit all of the usual decisions, including what screen resolution to optimize for. Being a usability specialist, I tend to err towards the lowest common denominator, which means that I've been building sites around 800x600 for a few years now.This time around, though, I decided to be a bit more data-driven, and I was surprised to discover that less than 3% of my visitors were running at 800x600, with a whopping 45% at 1024x768. This surprised me, in part, because I'm used to dealing with client sites that have relatively non-technical audiences, where I really have to be aware of the fact that their habits are often 2-3 years behind mine. Of course, my blog has a completely different audience, and tends to be focused in the website development community.
Now, many usability people would say: "So what? Stick to the lower resolution and everyone is taken care of." The problem is, that's not really true. If I optimize for the lowest common denominator, just to appease 3% of my visitors, I'm actually making the user experience worse for 97% of my visitors. Practically speaking, that's a huge disservice to my audience.
Of course, in a perfect world, I'd find a solution that's flexible and fits everyone, but we all know that that's not always possible. I did opt for a compromise; at 800x600, this blog is still mostly viewable, with just the loss of the right-hand column. Generally speaking, though, should we really be catering to the least savvy among us, or should we be aiming closer to the middle? If pleasing 3% of the population makes the experience worse for 97%, I have a hard time calling that good usability.
Dr. Pete
· Saturday, September 1I completely agree with that, Kevin. I don't mean to imply that we shouldn't aim to meet everyone's needs, just that, in the usability field, I think we sometimes overcompensate. We have to be mindful of when our efforts to please everyone all of the time end up pleasing no one. As you say, it comes down to the difference between theory and practice, and practice dictates making reasonable compromises.
Terry Bleizeffer
· Sunday, September 2I gotta disagree with you both on this one. Trying to meet everyone's needs is not a noble but unattainable goal, it's a mistake. Even if we define "everyone" as "everyone in our potential market", I still think it's a mistake to try to meet everyone's needs.
Have you ever met with marketing to try to clarify what types of users you were trying to go after, and walked out frustrated because marketing cheerfully told you that they were going after "everyone"? The problem isn't that going after everyone is hard, it's that going after everyone is impossible. You can't design a product that meets the needs of the specialized/advanced/expert/early adopter/compete-on-function customers AND meets the needs of the newcomer/average/compete-on-price customers, unless there are no other competitors in one of those battles.
Choosing what subset of the market you want to go after is not a "necessary evil", IMO. It's just good user experience.
Dr. Pete
· Monday, September 3You make a good point, Terry. In the context of the post, I was thinking a bit more narrowly in terms of the technological constraints of the user. In the broader sense, though, you can't build a positive user experience without a strong sense of who your users are.
Kevin
· Friday, August 31This is interesting, because great usability, to me, means making a site one-hundred percent usable to both the person running Lynx on a 386 and the person running Firefox with all the bells and whistles. Realistically, however, given time and budget constraints, this is simply not reasonable. The most one can hope for is to design the site for its largest audience, and to degrade or enhance it for the next largest group. For example, if 60% of the traffic received are from browsers without javascript, 30% from browsers with javascript, and 10% from text-based browsers, I would first perfect the site for browsers without javascript, then add javascript support, and finally make it function in text-based browsers, time-permitting. Good usability practices may dictate making a site function well for everyone, but it only sounds good in theory...