Using the Kano model for product alignment
AI makes features faster to build. That makes it tempting to keep adding them. I use the Kano model to help teammates agree on what to build.
Product alignment starts with agreeing on the problem we want to solve.
Trying it with a calendar
I start with three categories: must-be, performance, and attractive. At work, I use Claude or Codex to make an interactive artifact where teammates can place candidate features. We compare those placements to agree on what to build.
Here is a fictional calendar tool with six candidate features.
Drag a numbered feature onto a curve. Or focus a marker: left and right change fulfillment, up and down change the assumed curve. Positions are hypotheses, not customer data. Changes reset on reload.
- Save events after closing the app
- Edit and delete events
- Find events faster
- Show a busy calendar faster
- Remind me when to leave
- Suggest a free time to meet
Choosing the first version
For this example, assume the MVP focuses on the must-be features someone needs to manage a schedule. We agree on basics such as saving, editing and deleting events before deciding whether to include travel reminders. We still weigh cost, risk and product strategy.
The original iPhone's smooth touchscreen interaction was exciting at first. Now I expect a smartphone to respond smoothly to touch. Folding Burritos describes this change as the natural decay of delight. As AI brings new capabilities more quickly, I wonder whether people will come to expect them sooner, and whether we'll need to revisit our assumptions more often.
Where I go from here
I start with what I know and what I assume. Models like Kano help me discuss those choices with teammates. As I learn from people, I can revise the assumptions behind those decisions.
Design is a series of informed decisions, guided by assumptions, with the goal of solving human problems.