03

From AI task to institutional value

Value has to carry up, from the task to the institution.
  1. The taskFaster or better?
  2. The workflowDid the whole process improve?
  3. The people servedAre they better off?
  4. The institutionDoes it advance a priority?

Example: AI meeting summaries save 21 minutes each (task), go out the same day (workflow), free staff time for student outreach (people served), and support the student-success goal (institution).

Outcome hypothesis (required for each use case)

If we use [AI capability] in [defined workflow], we expect [measurable outcome] because [mechanism], provided [quality, access, and risk conditions] are met.

Specify the comparison baseline (the existing workflow, a redesigned non-AI alternative, or another option) and include implementation and transition costs. Where practical, use controlled or phased approaches to strengthen attribution.