David Housman

Usability (good God, y’all), what is it good for? Absolutely Somethin’!


Usability and Web Analytics are the peas and carrots of the user research world. Web analytics can tell you what is happening, but they don’t usually tell you why it is happening. Usability research tells you why something is happening in a small group.

Web Analytics and Usability need each other like Heads needs Tails, like Yin needs Yang, like Romeo need Juliet… like Elvis needed a peanut butter and ice cream donut sandwich.

In usability tests, user researchers invite participants to perform tasks on an interface while they are observed. They limit the impact of noise on the data by controlling the test environment. Usability tests provide one on one access to the participant, which makes it possible to inquire into opinions, suggestions, expectations, and reactions.

Usability tests offer the ability to collect information whenever you want during the study. We’re not limited to the constraints of html and JavaScript events. We can ask questions when someone furrows their eyebrow. I’ve run sessions where the participant cursed quietly, crossed her arms, and put her head down on the desk. In Web Analytics, this intense stress shows up as a high time on page. When people spend 20 minutes on a task and give up, SiteCatalyst makes that look like a number: just another exit. We also usually can’t control for user motivation in Web Analytics. In general, think about consulting your resident user researcher when you’re asking one of these questions:

  1. What are you thinking?
  2. What were you expecting?
  3. Why did /didn’t you do that?
  4. What did you expect to happen when you clicked _______?
  5. Where should that be?
  6. Was that easy or hard?
  7. What are the factors that affect someone’s actions?
  8. What should I do to iterate on this very new, very different concept?

So by now, I would expect most Web Analysts to be looking into online accreditation as a User Experience Professional, and pitching their implementation manuals and propeller caps out of the window. Well fear not Web Analysts: our skills are not entirely useless. The reason we don’t use offline user research all the time is a question of practicality. When a user researcher runs a usability test, the data collection is usually followed up by lots and lots of analysis. Usability tests take time to set up, to run, and analyze. As a web analyst, I can collect 1000 visits in a fraction of an hour. As a user researcher, it might take more than 2 years! It can be very difficult to tell how large an impact something might have, or how specific groups might react to a change. Talk to your web analyst if you have questions like the following:

  1. What is the impact of a change or factor?
  2. What do most people do?
  3. What’s do people do from very specific (i.e. 70 year old Males from New Jersey using FireFox)?
  4. How do users behave without artificial stimulation?
  5. What is the likelihood that something will happen?
  6. How does time affect something?

Leave a Reply