Build a Support FAQ from Raw Tickets
Turn a messy pile of support tickets into a structured FAQ that reduces repeat questions.
The Prompt
I want to build a customer-facing FAQ from my existing support tickets and common customer questions. First, I am going to paste a list of support tickets, common questions, or complaint summaries. They may be messy, repetitive, or in no particular order. Your job is to make sense of them. [PASTE YOUR SUPPORT TICKETS, CHAT LOGS, OR QUESTION LIST HERE] Once I provide the raw data, do the following: **Step 1 — Cluster the questions:** Group similar questions into themes. Name each theme clearly. (e.g. "Billing and Refunds", "Getting Started", "Technical Issues", "Account Changes") **Step 2 — Identify the top 10 most common questions** based on frequency or similarity in the data. **Step 3 — Write a clean FAQ entry for each question:** For each question, write: - **Q:** The question as a customer would actually ask it — conversational, not corporate - **A:** A clear, direct answer in 2–4 sentences. No jargon. Assume the customer is not technical. - **Escalation trigger:** One sentence describing when this question should be escalated to a human rather than handled by the FAQ or a bot. **Step 4 — Flag gaps:** Identify any questions in the data that you could not answer from the information provided. List them so I know what documentation I need to create. **Step 5 — Automation recommendation:** For each FAQ theme, tell me whether a chatbot or automated response could handle it fully, partially, or not at all — and why. Format the final FAQ as a clean document I can publish directly to my website or load into a support bot.
Variables in [BRACKETS] should be replaced with your specific details.
How to use this prompt
-
1
Click Copy Prompt above to copy the full text to your clipboard.
-
2
Open your AI tool of choice (Any AI recommended for best results).
-
3
Paste the prompt and replace anything in [BRACKETS] with your specific details.
-
4
Review the output and iterate — add context or constraints to refine the results.
-
5
Once you have the output you need, save it as a template for future use.
Want this running automatically — without lifting a finger?
This is what our agents do, 24/7, across your entire operation.