Many personas end up as a poster: a photo, an age, a hobby. The team looks at it once and goes back to deciding on gut feeling. A persona is only useful when it helps answer a product question.
About 150 interviews and usage data
At Gameram, I built behavioral personas from about 150 interviews. I mapped user attributes, tasks, journeys and the motivations behind them, and combined it all with product usage data.
The personas start from what people do and why, not from who they are.
Combining them with data is what gives a persona weight. The interview shows the reason. Usage data shows how many people behave that way.
Validated with a survey
At Turno, a mixed-methods study combined interviews and survey-validated personas. It found a 25-day average onboarding time, and the finding led to tutorial improvements.
Validating with a survey takes a persona from a handful of conversations to the real size of the user base. That’s what lets a team decide where to start.
What I take from this
- Start from behavior. Tasks, journeys and motivations say more than demographics.
- Combine with data. A persona without numbers is a good story. With numbers, it becomes a way to prioritize.
- Use the persona to decide. If it doesn’t change any decision, it’s decoration.
A good persona doesn’t describe who someone is. It describes what they’re trying to do.
Some details have been omitted for confidentiality.