Look beyond doing the same work faster. Pick one useful piece of work that wasn’t getting done, define what a good result looks like, and keep responsibility with the team.
Isn’t the point to save time?
It’s one point. Another is what someone can do now that they couldn’t do before.
1. Choose something a person would value.
A fictional test:
- The opportunity
- Turn an approved project handover into a short customer guide that never got written.
- The result
- A new customer can follow the steps and knows how to get help.
- Responsibility
- The team checks the instructions and owns the published version.
2. Evaluate the new value.
When a task wasn’t being done, don’t pretend there is a measured saving against it. Check whether the new output helps, what it costs to maintain, and whether it’s more useful than the next-best use of the team’s time.
Practical check: The team’s own expertise is what makes the result checkable.
Read more: If AI Frees Up Time, What Should You Do With It? · Where to Start with AI: Pick a Problem, Not a Tool
The example is fictional. This is general guidance, not advice for your situation.