A good starting use case usually has 3 traits: it's repetitive and time-consuming, input data is available and clear, and results are easy to measure. Don't start with something complex and rare — even if it sounds more exciting.
A fast way to find a use case: ask each team lead one question — "what task does your team repeat every week that takes the most time?" That answer is usually the best starting point.
Try it now
- 1Ask each team lead: what repetitive task takes the most time each week
- 2Filter for tasks with clear, available input data
- 3Prioritize use cases where results can be measured within 4-6 weeks
- 4Pick 1-2 use cases to pilot first, not everything at once
Example prompt
"My team includes these departments: [list]. Here's the most time-consuming repetitive task each department reported: [list]. Suggest 2 AI use cases to pilot first, and explain why."