A3: Guerrilla Usability Testing

PrintEmail

Usability testing is the process that allows those redesigning part or all of a website to see what users see.  By asking a user to perform a few set tasks using the website, one can observe their interaction with a website to reveal how the website confuses users and suggest improvements.  For all of its usefulness, usability testing has a reputation for being an expensive and time-consuming process.  The presenters, Jason Eiseman of Yale Law School Lillian Goldman Library and Roger Skalbeck of Georgetown Law Library broke the process down and showed how it can be easy and relatively inexpensive while still getting the same useful information. 

Eiseman and Skalbeck are both veterans of a website redesign process involving extensive usability testing.  They did a great job of giving examples from their own experience, and were also careful to cite the advice of experts, including Steve Krug and Jakob Nielsen.  I found this very helpful.  Too often, Annual Meeting programs focus too much on what worked at an individual institution. 

Eiseman and Skalbeck started their presentation with a usability test demonstration using a volunteer from the audience.  They were careful to run their demonstration test exactly the way they run a real usability test, which gave them an excellent example for the rest of the program.  They recorded this demonstration using Silverback, the same software Yale used in their actual usability testing.  This also allowed those in attendance to see the recommended software in action.  Later, Eiseman and Skalbeck replayed most of the recording to summarize their presentation.  Those listening at home can probably skip this part.

After the demonstration, the presenters gave simple instructions and tips for how to perform usability testing.  These tips were insightful and non-obvious.  For example, they caution against trying to prove a personal theory about how users interact with the website.  The presenters were careful to say that usability testing is better kept open-ended and used to discover how people use the site.  This was an interesting and not necessarily intuitive conclusion.  Usability testing seems to fit with the idea of testing a particular hypothesis.  This advice would help librarians design their usability testing questions to discover the best information about how their users interact with the site.

The presenters’ other most surprising point was the idea of testing “early and often.”  Rather than testing the website once when it is finished or nearly finished, Eiseman and Skalbeck recommended conducting many small tests during the redesign process.  Eiseman conducted usability testing on a simple wireframe illustration of how the website might look and just asked users where they would click if they could.  This allowed him to test how people would use the site before any of the links and other functionality was ready. 

Testing multiple times throughout the redesign process seems intimidating, but that is where the “guerrilla” part came in.  Eiseman and Skalbeck discussed how to simplify the process and make it less expensive.  They simplified the process, in part, by only testing with a handful of users.  To show how usability testing could be less expensive, the presenters discussed tools ranging from the expensive ($1500!) Morae software to the less-expensive mac-based Silverback to a simple paper and pen. 

I am glad I attended the presentation, and I would recommend downloading both the program and the handout.  The handout is a bibliography of sources on usability and usability testing including the Steve Krug book, Rocket Surgery Made Easy, which was referenced several times during the presentation, one website, and the software discussed and/or used in the program.  More helpful than just a list of resources, Eiseman and Skalbeck provide their valuable and frank opinions on each resource.  On the other hand, the handout does not include Eiseman and Skalbeck’s advice or instructions on how to conduct the testing.  Thus, downloading the handout alone will not give all of the useful information from this session.

Although I recommend downloading and listening to this program, I must note that visual aspects of this presentation were very helpful and important.  One of the best illustrations was a series of clips of several Georgetown users getting confused and making the same mistake when trying to use the website.  This illustrated why it is not necessary to see dozens of users to know that something needs to be redesigned.  It was an excellent point, but one that was nearly entirely visual.  This is not to say that the audio-only download will not be worthwhile, but those listening to the audio may be somewhat frustrated.

I would definitely recommend this program for librarians at any institution currently contemplating or undertaking a website redesign.  It provides great advice from librarians who have been there, serves as an introduction to the top books and software tools, and shows how to think creatively about a process that can seem intimidating.

Comments

From: 
Email:  
To: 
Email:  
Subject: 
Message: