7 comments

  1. This comment isn’t directly related to the article (you’ve heard that one before haven’t you…).

    Reading the article made me wonder, how much empirical research is their into the value of facets, controlled vocabularies, etc in the domain of websites, intranets, and so forth. I can easily understand the value of these approaches in a library context, but how well do they translate to the tasks that we do on the web? I’d just be interested to know what research has been done.

    The main thing that got me thinking about this was the fact you tried to apply theory/knowledge to the search issue, but your informed design proved to be unsuccessful when you did your user testing.

    Sherlock

  2. Hey, I don’t mind tangents. To address a couple of your points:
    – I didn’t think that the user testing actually discredited the faceted org scheme. Instead it killed the wizard UI approach. I definitely included facets in the alternate design and in the final recommendations.
    – I have to admit that I didn’t do a comprehensive lit survey on facet research during the project. I felt that facets were a design pattern that had already well established (invented by S.R. Ranganathan in the 1930s) in the realm of classification and info retrieval. Of course, part of my not-so-secret librarian agenda has always been to try to apply ideas from LIS to web design.

    If anyone knows of some good research studies on applying classification/facets to web design, please feel free to share.

    -Chris

  3. Although I found the case study interesting, I have a few comments about your process and your approach to the results.

    First of all, you started the search discussion by describing this as an opporunity to “shorten this process for people in a hurry” — i.e. the goal of search was to make the process of shopping for a card super fast. Right off the bat this argues directly against having a multi-page search wizard. As someone who has designed many a wizard and many a search, the one thing I can say is that even if having multiple pages makes things quicker in the long run, users still have the perception that having multiple pages is slower and more draining when they are trying to do something fast. Wizards seem to work best when there is a redundant process that by its nature will take a while but that always follows the same steps and has a clear finish. This does not really relate to search. As a result, I was SHOCKED that the client agreed to a 12 person user study to compare the two ideas. Not only does 12 people seem like a bit of an overkill for deciding which of these two ideas works better, but having such a formal test with a two way mirror for observation with videotaping in order to make a decision about two formats that were still in rough paper prototype format seems like a particular waste of money. Note that I am not arguing against user study in this situation…I just think that you would have been able to have the exact same results testing 5 users in person without the fancy facilities and the incredible outlay of time.

    However, this takes me to my next concern about the article. In the end, you discuss your pain and angst at “losing” and triumph your decision to swallow your pride. As a fellow designer, this commentary was a bit disappointing to me. I view our role as designers is to be the lone champions of users in a world of people trying to program things that they don’t like or sell them things they don’t want. Growing one’s ego about being the design expert and therefore the one with presumably the best ideas is a negative approach to interaction design. When I work with clients, I always emphasize that I don’t have a monopoly on good ideas, I am just the one they hire to make sure that the good ideas get through and the user always wins. Of course in the end, you were happy with the results, but I am surprised that your article presented your shock at being incorrect.

  4. Julie-

    Ouch! I humbly accept your criticism. Your scolding is well founded. In my defence I’d like to offer a few comments:

    – Hindsight vision = 20/20 At the time, there were a number of design constraints that lead me to consider a wizard. In hindsight, I’m glad that I took the time to flesh out an alternate design. In the end I learned from both.
    – For the purposes of the article I’m perhaps overstating my angst. I think that everyone should have an experience like this at least once. It’s a wonderful learning opportunity when you are proven wrong, especially when you are really attached to a misguided idea. But I hope you’ll forgive me for the momentary pain I experienced in the process.
    – Spending on user testing was a little different in 2000 than now. Even so, one of the outcomes that the client specifically requested throughout the project was to have a nice audio/visual record of the tests. There was actually an advantage in that this is cheaper than travel/hotel costs for moving multiple people between San Francisco and Detroit.

    Regards,
    Chris

  5. Julie,

    That was very eloquently put. Nice website too (reminds me a tad of the 37 signals site).

    Sherlock

  6. Chris,

    It’s *very* refreshing and important to hear stories about how people other than us IA’s can be right about UI issues. You are a brave soul to write an article about how you “disproved” your own best idea (at the time) and had the clarity of thought to see that in the end you DID achieve your goal because the user’s won. While it’s great to hear IA success stories, I also know that the kind of lessons like yours happen more often than not. They are just as valuable for us as individuals and as a collective as the case studies outlining why the IA was right from the get go. Nicely done.

  7. Great story, I really appreciate you sharing the results of your search wizard testing (people often ask me about that).

    I’m finding more and more situations where facts apply and avoiding dead-ends is a huge plus. All kinds of e-commerce, especially high-ticket items like jewelry or expensive vacations. Even Internet Yellow Pages are taking this approach. So I think you were really on the right track.

Comments are closed.