Check a vendor's sales claims against its own case study

About this exercise

A scheduling software vendor emailed you big claims and attached a case study. You want exactly 4 numbered lines, one per claim in the email, each labeled Backed, Stretched, or Not in case study, with the case study's actual figure after any Stretched label. Judge only by the case study, in plain text with nothing else.

Common mistakes when you ask AI for this

• Asking if it's a good product. The AI evaluates the software in general instead of checking each claim against the evidence. Try: Say: "Label each claim against the case study only."
• No fixed labels. The AI hedges with 'roughly accurate', so you can't tell a stretched number from a solid one. Try: Give them: "Backed, Stretched, or Not in case study."
• Not asking for the real figure. Without the case study's number, you can't push back on Thursday's call. Try: Add: "After Stretched, give the case study's figure."
• Not saying how many. The AI decides the length itself and usually returns more lines, one per claim than you can use. Try: Be exact: "exactly 4 numbered lines, one per claim."
• Getting markdown back. AI defaults to bold labels and headings, which show up as stray symbols in your notes for the call. Try: Ask for "plain text, no bold, no headings."

What a strong prompt includes

• Gives exact labels
• Names the source
• Asks for the real figure
• Says exactly how many
• Asks for plain text

Practice this now ↑

Related exercises

• Alert sales to a competitor's price change, facts only (Research AI · Product managers)
• Answer a buyer's HOA questions from the bylaws (Research AI · Real estate agents)
• Answer a coworker's travel policy question in one Slack reply (Research AI · Office and admin)
• Answer a family's visiting questions from the current policy (Research AI · Nurses and healthcare workers)
• Answer winter break housing questions from the current policy (Research AI · College students)