75. An AI product can charge per seat, per request or per completed task. How would you compare those pricing units?
MediumStrategy & prioritization
Case
Show model answer
Examine how each unit relates to customer value, predictability, operating cost and verifiable usage. A completed-task model can align incentives only if both parties understand what completion means and how disputes are resolved.
Test realistic bills and workflow variations with buyers and users. Compare margin exposure and adoption friction, including whether charging for attempts discourages valuable experimentation or hides costly failures.
What interviewers listen for
- Align units with observable value
- Test bill predictability
- Model margin and failure exposure