[WEB4LIB] Re: Website redesign survey
K.G. Schneider
kgs at bluehighways.com
Thu Mar 10 13:19:55 EST 2005
LII (http://lii.org )has done user surveys for two years running. Now we are
preparing for a usability review of our forthcoming new site. In fact, the
review is really for the new site plus one, since the new site isn't a
radical change (this usability review is being done so late it's primarily
catastrophe insurance), but I digress.
We've had about 4,000 respondents both years we've surveyed. We found our
user surveys to be extremely helpful for planning purposes, and in that
sense, survey responses are great for building in usability from the ground
up and providing the kind of useful insights you can't get any other way
(and the kind of insight it's helpful to get before you spend a year running
in the wrong direction). As Mike points out, a survey a complementary tool,
and of course a survey is skewed to whoever will take the time to do it (be
brief! And look brief! I bail out of online surveys if they appear to be
long and meandering). But yeah, we learn a lot from our surveys. The best
lessons were the curve balls--things we hadn't thought about or anticipated
response about. While you do need to swallow your pride, you also will find
out things that are gratifying and fun to know.
A survey is also a marketing tool, of course, and a way to float trial
balloons, and a reminder that we care about users, etc. etc. (And so I tell
myself to get myself into gear for planning this year's survey!)
I wish I knew more about good survey development. However, a survey designed
by a rank novice is better than nothing. (Did I mention keeping it short?)
Btw, one thing task-oriented usability testing can't do is give you the idea
of how people use your site over time. The survey tells us that people use
LII for collection development, staff and user training, and personal
development, that they like our style, that they appreciate our personal
responses to enquiries, that the weekly blurbs in the newsletter are
effective. You can't get that out of a series of one-time tasks. (However
important those tasks are..!)
Karen G. Schneider
kgs at bluehighways.com
More information about the Web4lib
mailing list