AI to Production · tool 1 of 9
Choosing an AI Use Case
Choosing an AI Use Case: picks problems where AI can create value, and avoids expensive demos.
Best for: Before any AI project · Time: Half a day
Domain: Across all domains · Focus area: Initiating
How it works
Start from a business problem and a process, not from a model. Score each idea on value, data availability, how far errors can be tolerated, and whether it fits into real work.
- Ask what happens when the AI is wrong.
- Check whether simpler software solves it.
- Find an owner who will change the process.
How to apply it
- List problems with their cost today.
- Score value, data, error tolerance and adoption.
- Choose one with a clear owner.
- Write the baseline metric.
In practice
Summarizing support tickets for agents is a safe start. Automatically refusing refunds without a human isn’t.
Pro tip
If a rule or a spreadsheet would do it, use that. It’s cheaper to run.
Use it
- Template: AI Use Case Scorecard (blank), filled example
- Case study: Zillow Offers
Quick check
What should an AI project start from?
Sources
Last reviewed 2026-11-02 against the PMBOK® Guide, 8th edition · Practitioner review pending · How this guide is made