Count a portfolio's RAG statuses correctly and name the worst project

About this exercise

Count the projects by RAG status using the rule in the task: Red = a milestone date missed or more than 10% over budget; Amber = at risk but no miss; Green = otherwise. Output a markdown table with columns Status, Count, Projects and then one line: Worst: the single project with the largest overspend percentage, with the percentage. Compute percentages from the budget and spend figures. Use the latest figures. Output only the table and that line.

Common mistakes when you ask AI for this

• Using the old forecast. The AI keeps Dune at 98. Try: Say: "Dune's forecast is 101."
• Dropping the missed milestone. The AI treats Cobalt as green after completion. Try: Say: "The miss still counts this month."
• Wrong worst. The AI picks the biggest dollar overspend by feel. Try: Say: "Compute overspend % = (forecast - budget) / budget."
• Wrong table. The AI changes columns. Try: Say: "Columns: Status, Count, Projects."
• Trusting the AI to count words. AI is bad at counting. Ask for a total word limit and it often lands over it. Try: Cap it: "only project names in the Projects cell."
• Trusting the AI to count words. AI is bad at counting. Ask for a total word limit and it often lands over it. Try: Cap each part instead, for example: "each line under 12 words."

What a strong prompt includes

• Columns
• Rule
• Formula
• Updates
• Worst line

Practice this now ↑

Related exercises

• Age a returns queue in business days and flag the ones over the limit (Customer Support & Success AI · Customer support teams)
• Alert sales to a competitor's price change, facts only (Research & AI Search · Product managers)
• Announce a colleague's promotion with approved facts only (Leadership AI · Managers and team leads)
• Announce a manager swap in Slack with two sentences and one link (Leadership AI · Managers and team leads)
• Announce a new team member in an email with exact first and last lines (Leadership AI · Managers and team leads)