Campaign Website Navigation: A Usability Test
A special word of thanks to Optimal Workshop for helping me out with this test. They have amazing usability testing tools and if you build websites, it will better inform your decision making.
The Test
After reviewing the website navigation menus of 50 US Senate campaigns, we learned that 11 different terms were used to describe the volunteer page.
I decided to test the three most common terms — Volunteer, Get Involved, and Take Action — to see what impact, if any they had on a user’s ability to successfully navigate to the volunteer page.
Success Rate
With Chalkmark (from Optimal Workshop), I asked users “Where on this page would you click to help out the campaign with your time?” (a wordy question that purposely avoids the words being tested) then randomly showed them one of three homepage variations. The only difference between the three variations was the name for the volunteer page.
Thanks to the awesome subscribers of my newsletter, I had 97 participants* take the test. 54% successfully navigated to the correct page in a single click (more on why first click is important here).
But the word choice played an important factor in determining a user’s success rate:
Users who saw “Volunteer” were 10% more likely to get to the volunteer page with their first click than those who saw “Take Action” and 5% more likely than those who saw “Get Involved.”
Completion Time
Another factor to consider is how quickly the user was able to complete the task successfully. Here’s a breakdown of the results:
Now we see that “Get Involved” was the term that, on average, helped users successfully find the volunteer page the quickest.
What this means
When it comes to answering the question, “Which is better?” we’ve got something of a hung jury, since we don’t know what impact a faster time has on a completion rate. But there is a way to find out: keep testing.
If this were a real campaign with a real website and real traffic, you could run an A/B test to measure which navigation term yields the most volunteer signups.
But we did see one way you can use a tool like Chalkmark to narrow in on the right question to test and to help inform design decisions with data rather than opinions.
*Obviously I would have preferred more volunteers to collect more data points, but I think the results are still interesting and it’s a great example of a question that can be solved by data rather than opinions.
