Library › AI Evals › Product managers

Grade 4 AI onboarding emails against 3 criteria

Practice: Asking for a format you can use. About 5 minutes.

About this exercise

Your AI writes welcome emails for new Tasklane signups, and you're grading this week's sample. On Nov 1 the Starter plan was renamed Basic, so 'Starter' now fails C1. Grade the 4 emails in exactly 4 numbered lines: '1. PASS' or '1. FAIL - C2 - "exact words from the email"', naming the first criterion it fails. Plain text, under 90 words.

Common mistakes when you ask AI for this

  • Missing the rename. The AI doesn't know Starter became Basic, so it passes Email 2. Try: Tell it: "Since Nov 1, Starter is called Basic. Starter fails C1."
  • No evidence behind a FAIL. The AI writes 'fails C2' without showing why, so nobody can check the grade. Try: Say: "Quote the exact words from the email that fail it."
  • Paragraphs instead of lines. The AI writes a review of each email with suggestions. Try: Show the shape: "2. FAIL - C1 - \"quote\"", one line per email.
  • Getting markdown back. AI defaults to bold verdicts and a table. Try: Ask for "plain text, no bold, no tables."

What a strong prompt includes

  • Shows the line shape
  • Asks for a quote as evidence
  • Gives the plan rename
  • Says exactly how many
  • Asks for plain text

Practice this now ↑

Related exercises