Library › Data AI › Product managers
Count support tickets into the top three themes
Practice: Catching mistakes and made-up facts. About 5 minutes.
About this exercise
You're bringing last week's support tickets to backlog grooming. You want exactly 3 numbered themes, ranked by how many tickets mention them, each in the form 'Theme (N tickets): example #ticket', under 60 words, plain text. Count each problem once: skip the duplicate ticket and the spam, and leave out one-off requests.
Common mistakes when you ask AI for this
- Asking to 'summarize the tickets'. The AI writes a paragraph per ticket group with bold headings and no counts. Try: Be exact: "Exactly 3 numbered themes. Plain text."
- 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 line instead: "each line under 12 words."
- Counting the duplicate. The follow-up ticket looks like a new export problem, so the AI counts it twice. Try: Add: "Count each problem once. Skip #107 (duplicate) and #105 (spam)."
- No line shape. The AI writes long theme descriptions without a count or an example. Try: Show it: "1. CSV export broken (4 tickets): #109".
What a strong prompt includes
- Asks for counts
- Skips duplicates and spam
- Sets the number of themes
- Shows the line shape
- Asks for plain text
Related exercises
- Alert sales to a competitor's price change, facts only (Research AI · Product managers)
- Approve or return expense claims against the policy (Productivity AI · Finance and accounting)
- Approve part of an overtime request with the math (Writing AI · Managers and team leads)
- Average daily call activity after fixing a dialer glitch (Data AI · Sales and customer success)
- Average handle times after removing outage calls (Data AI · Customer support teams)