Interview x amount of users. It should be a large enough sampling to see trends and patterns in responses to your interview q’s. One designer leads, the other takes notes. Sometimes rounds of interview are helpful to regroup and rework the questions in between.
In interviews, ask why. A lot. Get to the real *why* behind their behavior. Let them talk freely, but guide the conversation and keep it on track. Avoid leading questions and confirmation bias.
Don’t deviate from your questionnaire too far. You’ll need parity between user's responses, but follow your curiosity. Have a conversation. It could lead to new ideas. 💡
With another designer, plot and cluster users on a wall that have similarities in their perspectives, workflows, etc. Don’t overthink it or try to label the clusters too soon. It’s okay if there are outliers.
Do not do this alone. Talk out loud about what you heard/observed.
Start to bubble up the individual behaviors and mental models into a generalization about the group. Don’t lose track of user data though, it’s important to keep the red thread.
Use 2x2s to plot and stress test defining traits of your persona groups. You can drop groups if there’s not enough healthy tension for them to serve as vital, useful lenses on their own. Swap out axis for things like trust, confidence, comfort, safety, motivation, activity, etc.
Now it's time to document and name your personas. Consider using adjectives that represent their driving behaviors or “the greatest gift” you could give that persona rather than defining them by their demographics or a meaningless fake name. ✨
I got a little 🔥 for not explicitly mentioning observations alongside interviews. Yes, observe. Talk to people. Watch them use your product/service in their environment.
I also got a little 🔥 for suggesting personas be created by designers, and not with the wider org. I'll stand by that and respectfully disagree that everyone needs to do the tactical work of making the sausage.
I've given 1 example per quadrant here for simplicity, but in reality you'd want to see patterns & trends across x amount of observations (what you saw & heard) for it to become an emerging insight & implication. ✨
Let's do another one!
What's fun is to see the tension between observed and expressed needs. That's where the kernel of a good idea starts.✨