Label each press release stat against the survey data

About this exercise

Your PR agency drafted a press release around your pet-owner survey, and you need to approve the numbers. You want exactly 5 numbered lines, one per stat in the release, labeled Matches, Wrong, or Not in survey, with the survey figure after any Wrong label. Use only the survey data, plain text, and nothing else.

Common mistakes when you ask AI for this

• No fixed labels. The AI comments on each line in prose, which is hard to approve or reject quickly. Try: Give them: "Label each Matches, Wrong, or Not in survey."
• Calling missing stats wrong. The $1,100 figure isn't in the survey, and the AI may judge it by what it knows about pet spending. Try: Add: "If the survey doesn't include it, label it Not in survey."
• Not converting fractions. "One in three" can slip past unless someone compares it to 22%. Try: Say: "Convert phrases like 'one in three' to percentages before comparing."
• Not saying how many. The AI decides the length itself and usually returns more lines, one per stat than you can use. Try: Be exact: "exactly 5 numbered lines, one per stat."
• Getting markdown back. AI defaults to bold labels and headings, which show up as stray symbols in an approval email. Try: Ask for "plain text, no bold, no headings."

What a strong prompt includes

• Gives exact labels
• Names the source
• Checks number phrases
• 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)
• Announce a referral bonus with the steps to claim it (Marketing AI · HR and recruiting)
• Answer an angry 'I unsubscribed' email honestly (Marketing AI · Marketing and content)
• Answer your CMO's campaign question in one Slack message (Marketing AI · Marketing and content)
• Approve or return expense claims against the policy (Productivity AI · Finance and accounting)