Library › Marketing AI › Marketing and content

Score headline options from reviewer ratings and recommend one

  • For Marketing and content
  • You practice Catching mistakes and made-up facts
  • Time about 5 minutes
  • Level Advanced

About this exercise

Three headline options for the Orchid Studio homepage were rated by reviewers. Write a prompt for a markdown table with columns Headline | Clarity avg | Specific avg | Total, then one line starting "Recommendation:" with the headline letter and a reason under 15 words. Average each score across the valid reviewers only, show one decimal, and total = clarity avg + specific avg. Ignore any reviewer who didn't read the brief. Under 80 words, no other text.

Common mistakes when you ask AI for this

  • Including Jun. The AI averages all four reviewers, which makes A and C look better. Try: Add: "Ignore Jun, who didn't read the brief."
  • Recommending on mobile wrapping. The AI weighs the wrap worry, which wasn't a criterion. Try: Say: "Decide only on the Total."
  • Skipping the format. The AI writes a narrative review of the headlines. Try: Name the columns and the Recommendation line.
  • Letting markdown through. AI adds bold text and headings by default. Try: Add: "Plain text only, no bold, no headings."
  • Trusting the AI to count words. AI is bad at counting and often runs past a word limit. Try: Cap each part (for example "each line under 15 words") so the total lands well under the limit.

What a strong prompt includes

  • Names the columns
  • Excludes the bad reviewer
  • Defines the total
  • Fixes decimals
  • Asks for a Recommendation line

Practice this now ↑

Related exercises