UIEtips: Quick & Dirty Usability Testing: Step Away from the Book

Jared Spool

October 20th, 2008

At last week’s UI13 conference, the buzz was about getting started with usability testing. Folks I talked to had a frequent refrain: their group wants to start with some type of user research, but they can’t afford a full-blown scientific study. Neither the schedule nor the budget would let that happen.

That’s why Dana Chisnell’s session on quick-and-dirty usability testing resonated with so many of the attendees. They realized they could get a research effort off the ground without having to beg for a huge investment. In many cases, they could do it quickly and inexpensively, under the radar, yielding enough valuable information to make substantial improvements to their design.

In this week’s UIEtips, Dana shares how teams are using quick-and-dirty techniques and getting tremendous insights. I’m betting, after you read her article, you’ll see how you don’t need to follow “the book” to learn things that will improve your designs.

Read the article – Quick & Dirty Usability Testing: Step Away from the Book

Have you tried any quick-and-dirty user research techniques? How did they work for you? We’d love to hear about your experiences below.

9 Responses to “UIEtips: Quick & Dirty Usability Testing: Step Away from the Book”

  1. Max Soe Says:

    How do you actually acquire users for a client who is a corporate? For example, I may be working to improve a feature on an online banking site. How do I go about recruiting the client’s users?

  2. Dana Chisnell Says:

    Max,

    There are lots of ways to find participants for studies. If you’re sure you want current customers (rather than people who might be new to the site), you can try some things that I’ve seen work for my clients:

    - Request that a link be put on the site that invites people to sign up to take part in user research (it might send them to an email that goes to you or to a small questionnaire that gathers contact information and then you can call them later). This has the advantage that people have opted in. If it’s in a place on the site that people have to log in to use, you know they are already customers.

    - Hang out at a large branch (if your bank has branches) and ask people who are standing in line if they would like to participate in a short research session that will help the bank serve them better. Work with the bank manager to make sure he or she is comfortable with you doing this. You could even do short sessions right there in the branch if everyone is good with that.

    - Ask bankers in branches to help you recruit people who are opening accounts. Take the bankers out for coffee to thank them.

    If you have a couple of weeks before conducting your usability test, you might think about using personal and professional networks to help you find participants. Posting to craigslist.org might also work well, if you word the ad carefully.

    I don’t recommend getting customer lists from the client for two reasons. First, most banks have privacy policies that customers have agreed to that prevent customers from being contacted for anything but doing specific business. Second, the contact information — even at banks! — can be woefully out of date.

  3. Tara Schnaible Says:

    Working for a non-profit, we rarely have the luxury of time or formalized equipment for testing. Some aspects of formalized testing have made their way into our process, but our average recruit/test/result-presentation time-frame is about 1.5 weeks. We make certain to maintain tester anonymity (even when showing developers/team members video), we record mouse movements (and manually tally patterns), we formally present results via spreadsheets that breaks results down into “overarching issues” that have fixed severity/repeats/timing measures.
    Formality aside, where that testing deviates:
    @ We’ve found ways to use low-cost applications/sites to test (Webex sessions through Nethope, they can be recorded and as a member of Nethope we have free access to Webex) We use Adobe Acrobat/Axure models/early-dev websites/Paper-Prototypes, whatever we can, to get ANY INFO about basic ideas about terminology/navigation/etc as early as we can.
    @ We do not create formal/elaborate Word reports, our teams have to trudge through an Excel results sheet
    @ We test with 5-7 users, tests are oftentimes conducted remotely (see Webex/Skype) and we spend weeks at the onset of the project recruiting appropriate users from the project team/other representative people from local offices/etc.
    @ Due to remote testing and a need to make users feel at-ease (they are often strangers to the tester) – we are informal, friendly and chatty without leading answers/questions.
    @ Smallish tests of terms/one page can be done by walking around the office asking people 1 question that tests an interface feature we think might be challenging. Even tiny, impromptu tests can help provide guidance on whether a feature/widget/phrasing will work.

    The underlying theme here is that you don’t need great tools to eye-track, record mouse-clicks, etc. E.g. If you see a 20-something lean in to a screen on a website to read text, signs are pretty good your fonts are way too small. You can find out basic, valuable information in 5 minutes of questions and no budget.

  4. Mark Pawson Says:

    Run Tests at Tradeshows. There was two posts on this at IXDA.org
    http://www.ixda.org/discuss.php?post=23564
    http://www.ixda.org/discuss.php?post=18865

    Saul Greenberg and I just finished submitting a paper to the Journal of Usability Studies on a technique I used successfully with a colleague at a major Photographic show. Here’s hoping it gets accepted because when I did the research on this I found almost nothing for published material, even though based on the two links above, others have successfully done it.

  5. Samantha LeVan Says:

    When I have the time, I work on pet projects where I will take weeks or months to do formal research and usability testing. But on a regular basis, most everything I do is quick & dirty. I’d rather make sure some research is done, no matter the project schedule constraints, than say “sorry, I’d need at least a month”. It was hard for me to learn this outside of school, but once I realized my results were similar, although maybe less detailed, it’s become my favorite way to conduct user research. What’s more thrilling is that my clients are just as happy with quick and dirty results.

    I generally try to shoot for five users per round of usability testing, but if I see the first three all have exactly the same issues and I’m under pressure to return the results quickly, I’ll stop there and collect my thoughts, make a quick PowerPoint or Word doc and be done with it.

    One of my favorite ways to present data is in Visio. Just thought bubbles right where the problems are… works well when there are many detail issues to resolve, as opposed to wholistic design changes.

  6. Bev Says:

    I am very interested in the method Dana proposes, but I’m unsure how I can relate it to the type of client sites I design, which are mainly B2B (Business to Business) sites in areas like Telecommunications. These sites are aimed at business customers. Apart from trade shows, I’m unsure how I could quickly and easily find suitable participants for the type of sites I’m working on. Any suggestions?

  7. Samantha LeVan Says:

    Bev,

    Do the participants need to be familiar with the website? For usability testing, as opposed to contextual research, you could try people who are business professionals from other fields to get a sense of the navigation and “findability” of information. Coffee shops are a great place to find business people, particularly at lunch or just before work. For quick and dirty, the perfect user isn’t necessary. Closely related backgrounds and level of PC/Internet experience will still give you results to inspire design solutions.

    Samantha

  8. Dana Chisnell Says:

    Hey Bev,

    I’m totally projecting here, but there have got to be users — or people who are like your users — somewhere nearby. Even though your sites are business to business, there’s a human there somewhere. There might be someone just down the hall from you who fits, or comes close. Can you give us an example?

    Dana

  9. dış cephe Says:

    Thanks Dana.I agree.

Add a Comment