Recently I spent time with 10 users to gather reactions to proposed names for the a new feature on a shopping website. 5 fashion bloggers and 5 non-bloggers spoke with us for an hour about 5 different “features” we are building on the site.
Our main objective in speaking with users was to understand the connotations of each name and how they would imagine the feature behind each name. We also gathered great feedback on the feature itself which the team has already started acting on.
This research was meant to help settle a contentious debate between 5 very similar names. Here’s what I did:
First, because I wanted participants to focus on what they thought each name was, rather than which one they liked or didn’t like, I framed each name as a feature we had already built and for each name I asked the participants three questions:
What would you expect to see here? Why?
What might you be able to do here? Why?
What do you think of when you hear this name? Why?
From these answers I gathered impressions and sentiment based on the language used and the description of the imagined product.
Next I showed each name along with a word cloud of relevant positive and negative descriptors. I asked participants to mark the words they thought best described the qualities of the feature/name and explain their choices.
I also had them take each name and place them on a continuum of observe__vs__participate. This technique didn’t yield the best results, but it did work well to help participants articulate their thoughts.
After about 30 minutes of moving words around I finally showed them the new product and let them explore.
Lastly, I wanted to gauge memorability of the names. Since I didn’t have two weeks to wait and call them all back, I asked them at the end of the hour to recall the names we had talked about at the beginning. I had mixed feelings about these results. People tended to remember the most familiar names, and some of the names that were described as the most interesting, were not remembered at all.
What worked well
- Changing the order of the names for each participant to reduce bias.
- Framing the names as features we had already built. This took the distraction of the hypothetical away and allowed the participant to focus on their thoughts.
- The word cloud allowed users to recognize words they otherwise wouldn’t have thought of. This allowed us to dig deeper into the connotations of the names.
Here are some other resources I referenced: