What is a sensible first scope for an agent program?
Why B is the answer
A contained, high-value use case with a clear owner
Start small and measured, then scale what works.
In detail
- A contained, high-value use case with a clear owner lets you learn safely.
- Avoid a broad launch with no clear owner.
Use case
An illustrative scenario, not a real incident.
A company starts with one agent for invoice queries, with one owner and one measure. Later work builds on what it learned.
See it in different ways
flowchart LR A["AI strategy"] --> B["Plan"] B --> C["Ready"] C --> D["Govern, secure and manage"] D --> E["Build and run"]
stateDiagram-v2 [*] --> Pilot Pilot --> Measure Measure --> Improve: low ratings on a topic Improve --> Measure Measure --> Expand: targets met Measure --> Rollback: serious problem
flowchart LR S0["One use case"] --> S1["Measure"] --> S2["Improve"] --> S3["Scale"]
Key point: Start contained, then scale.
Why the other options are wrong
- A. Every process that has ever been documented
- That scope is impossible to manage.
- C. All departments, on the same day, with no pilot
- Launching everywhere at once multiplies the risk.
- D. The riskiest process in the company, to prove the value
- The riskiest process is the worst place to learn.