Kit · 6 steps
Prioritise use cases and measure the return
A ranking of use cases with owners and baselines, and a one-page business case a CFO accepts.
The ZIP includes a README with the order, every file in Word/Excel/Markdown and each prompt.
- For
- Teams with many AI ideas and little budget.
- People
- You + a team of 3 to 5
- Level
- Intermediate
- Time
- ≈ 80 min
The steps, in order
-
Which AI use case goes first: value, ease and risk in one table
Rank ideas by value, ease and risk.
-
Measure this before you automate: compared with what?
Set the 'compared to what' before measuring.
-
From hours saved to growth: AI return scorecard
Add a business indicator, not just hours.
-
More productive, not more profitable: where does the saved hour go?
Decide what the freed-up hour is used for.
-
One-page AI business case (with its evidence gaps)
Build the business case on one page.
-
Prompt: hypothesis tree for an AI decision
Find the data point that would kill your proposal.
You are done when
- Every row has an owner with a name.
- Every number is either measured or labelled as an assumption.
- The last document ends with one owner, one date and one next action.