Wednesday, May 14, 2008

How We Do Usability

Last week, I discussed a little bit about what usability is (which is what I do for a living). Hopefully that gave some conceptual framework to this week.

Ok, so what exactly do I do? Well, a few things. One thing we do is look at existing applications or sites and help make suggestions to make them better, often by doing redesigns. The other thing we typically do is to start with a project at the beginning, trying to design it as usable from the beginning. Do I have a preference between those two? Well, due to my critical nature, I kind of like looking at existing sites and finding the problems, but honestly, it's much better for the users if we are involved from the start.

To do our work, we have 2 tools: 1) experience (ours and others) and 2) usability testing. To be perfectly honest, a lot of usability work is common sense and experience. By doing the work and the reading, we get a lot of knowledge about how users do things. For instance, it's simply common sense that you DON'T put a ton of text at the beginning of a website or an application. Why? Because people don't read it. All of that marketing speak and junk that CEOs and marketing people love to have to extol the virtues of their brand or company or whatever. Users have learned to scan through websites and applications to find what they want. Very rarely to people on the web browse anymore. Ten years ago when the web was pretty new, sure. But now, people have a purpose when they come. They're wanting to do something or find something; they don't want a brochure. So common sense says, don't have a ton of text at the beginning of a website or application. And yet, how many places do you still see this? Way too many.

The other tool in the bag is Usability testing. What's that? It's very easy. You design your site or application, or you take an existing one that you suspect has problems. Then you get a typical user, sit them in front of it and basically say, "How would you use this?" or "How would accomplish this task?" And then you get another one, and another, and so on and so on to make it statistically significant. If the large majority of people can accomplish what you ask, you know that you've got a pretty good system. If they can't, you need to redesign. Usability testing is the absolute best way to determine how usable a system is.

Because here's the thing: even with all the years of experience that our group has, we still get things wrong. It's very easy to get caught up in getting things done and meeting deadlines that you forget to take certain things into account, like how a user will get to a particular area that you've designed. That happened to me on a project a few weeks back. The other thing is that, especially on a redesign of an existing system, users will have a certain mindset about how to accomplish a certain task and you have to take that into account. Otherwise, they're not going to be able to use what you're creating.

So, those are the tools. Next week, some hurdles to usability. Some ways that people ignore or don't take it into account or marginalize it. Hopefully, those of you out there are finding this slightly interesting and eye-opening. Usability is a growing field and one that people don't really notice until they come across a site or application that just doesn't work.

1 comment:

Rob Cox said...

[sarcasm]You mean you don't just do whatever is easiest for the developer?[/sarcasm]

Template Designed by Douglas Bowman - Updated to Beta by: Blogger Team
Modified for 3-Column Layout by Hoctro