I’m preparing for a talk next week at the Library Technology Conference in St. Paul, and this has me looking at a lot of library websites. And they STINK. Of course, this isn’t really news. People have written about bad library Web design before, and many have defended it because of staff limitations, budgets, etc. But simply designing a “good” website isn’t going to solve the problem. It’s like moving into a rickety old barn, finding it is a bit cold, and installing a programmable thermostat instead of looking at the gaping holes between the clapboards. You see, poorly-designed library websites are a symptom of a much larger problem that technology can’t fix. The problem is a library culture that gives lip service to user needs while really catering to librarians.

This is why doing a redesign of a website is far less a technical issue than a political one. I’ve been chatting today with colleagues who say that there are internal battles among different departments about who “knows” the user better. Unfortunately, this fight isn’t about who spends more time researching what users need; it’s an internal power struggle over whose personal vision for the library website will win out. It’s about librarian tribes and political infighting, and the end results is always a bad user experience. Always.

I am fortunate to work at an unusual library that empowers me to build to what our users need.1 I know what you’re thinking. “We build to what our users need! That’s not unusual!”” But the thing is, you don’t. You don’t know what your users need, because you don’t take the time to find out. You do what you think they need.2 Who, besides a librarian, would want all 48 links from the navigation repeated on the homepage?3 Now, librarians are not like library users. This is a fundamental rule of user experience design, and a mantra that we should all repeat, endlessly, as we do our work:

The user is not like me.4

Go ahead and say it a few more times.

“But we test with real users!” you say. Really? Doing one usability test every 3-5 years when you redesign your website doesn’t tell you anything. Handing out a survey by itself once every few years doesn’t do anything except waste paper. Done infrequently, these things just let you check off the “user-centered design” box when you write up the whole process for Library Journal. Understanding your users isn’t one step in a long process to make a great web experience. It’s the foundation of that experience. This is about building a relationship with your users. How many successful relationships have you seen that do a quick check-in once every 3-5 years? You should always be doing it. ALWAYS.

It shouldn’t be such a hard sell. After all, when problems comes up at our service desks, do we wait 3-5 years until the next time the desk is “redesigned” to address them? Why does a user on the Web count less than a user in person?

Check your gate counts. Now check your Web usage statistics. Is your catalog on your website? Your databases? Then guess what: 100% of your users come through the website. Now, how many come through the front door? Less than 100%. Yet where do your library’s priorities lie? You can find them written in the staff directory. How many people spend their time dealing with patrons in meat space? Now how many spend their time building great web experiences? See the discrepancy?5

Building great experiences on the web is 90% listening to your users. The rest is delivering your user’s needs with as little organizational interference as possible. Libraries are failing across the board, but it’s not too late.


  1. Mostly.
  2. Not all libraries ignore their user’s needs. Look at Brigham Young’s library homepage. It’s clear that there is a single focus: finding things.
  3. I’m not picking on any one library here. Do a Google search for “university library.” You’ll find link farms on 15 of the top 20 results.
  4. Through listening to Whitney Hess, I can say that this mantra is taught extensively at Carnegie Mellon’s Human-Computer Interaction Institute, although I can’t get more specific with the attribution. Anyone know?
  5. We can, of course, continue to ignore this human resources problem into the future, but these digital chickens will come home to roost. Why not plan for them, maybe install a programmable thermostat in their coop?