Minimum Usability Testing: Now there’s no excuse not to do it.
(originally posted at idealware.org/blog)
According to the author of the web usability bible “Don’t Make Me Think” Steve Krug, testing one user is 100% better than testing none and testing one user early in the project is better than testing 50 near the end. To that end, I’d like to suggest testing 5 or 10 more often and found a couple of services that make it possible.
While I am afraid that too many of my Idealware blog posts have been focusing on cheap or free resources, I do believe that the easier it is for nonprofits to do usability testing regularly the better their sites will be – so here is one more post about a couple of low cost options.
Strictly a quick an dirty sort of testing, I stumbled across this crowd-sourced usability service on a blog and had to check it out. For $10 you get to set up your own questions or scenario and receive responses from 10 reviewers.
The results are, as you might expect for that sort of money, a mixed bag from a wide generic set of users. Although I really don’t believe any site should be targeting anything so broad as “general public” it can be helpful to have this kind of input. I would see this as useful for organizations doing a redesign to try out a couple of different directions or make a specific design choice (red or orange for that donate button?).
The site provides a bit of guidance for how to get the best and most useful responses and I have to say, man, is it fast – less than a day turnaround time to collect 10.
The reviewers come from a service called Mechanical Turk which, although I have been around the web a while, I had never heard of before. Its an Amazon program where users offer and accept small tasks best done by humans for small fees. Its feels a little creepy, and I do have some doubts about ethics and ramifications, but apparently it has a big following and fan base on both sides. Feedback Army acts as oversight and interface in order to maintain quality control, so only reviews they judge to fulfill your requirements filter back to you.
Overall I found the feedback to be helpful in my test case and will probably use this again when I need an “outside perspective” on design decisions.
A comment from the original blog post pointed me to Usertesting.com, which is a bit more expensive ($20 per tester) but which blew me away with what you get for your money. I submitted my site for a trial and received an incredibly well thought out and thorough review video in which you see the user’s screen and mouse movements while listening to a narrative of what they are thinking as they complete your request and answer your questions. It gave me a lot of new ideas and insight about where I was missing the mark by being too close to the subject. In addition you get a standard written report answering basic usability questions that is also useful.
Jakob Nielsen, the godfather of usability testing, makes a good case that testing 5 users will yield the highest value of information in the most economic way – so for $100, a nonprofit could conduct some of the worthwhile usability testing that seems to always fall through the cracks.
And of course if you ask the right questions you can also test new ideas and designs with your own supporters using a survey tool like Survey Monkey for free, so there really is no excuse not to do usability testing at this point.
More on quick and dirty usability testing and why you need it, as well as what to test and how, can be found on Jakob Nielsen’s site:
Fast, Good and Cheap:
Usability for $200
Usability on Nonprofit Web sites