First, I interviewed the product team to understand their vision and ambitions for the product. Their passion for Pi’s abilities to improve people’s lives was truly inspiring.
Next, I wanted to learn what active users loved about Pi—things that, in their view, made the chatbot helpful and sticky. I interviewed 20 users and analyzed their behaviors, needs, and pain points. This was eye-opening, as I realized how much people enjoyed talking to Pi and how different they thought it was from the other AI assistants.
The most interesting finding, however, was that Pi was helping people in more ways than the product team initially envisioned. The interviews with users revealed four mindsets—four distinct approaches to why and how people were interacting with the chatbot.
I used the findings from primary research and the data from competitive analysis to craft value propositions that were true to Pi’s purpose and would resonate with potential users. Eventually, the product team redesigned Pi’s interface to include messaging that inspired people to speak with the chatbot and learn how to get the most out of it.