Write a saved reply that does not read like a saved reply
Builds a reusable reply for a question you answer every day, with clearly marked places to personalise so it never goes out generic.
Drafts a reply that acknowledges the real problem, says exactly what happens next, and promises nothing you cannot deliver.
Works with Claude, ChatGPT, GeminiUpdated 2026-09-27
Use this when a message has made you want to reply defensively. Angry customers mostly want to know that someone understood the specific problem and what will happen next, and this keeps the reply on exactly those two things.
Opening this in Claude or ChatGPT sends the prompt with its {{PLACEHOLDERS}} still in it. Replace them with your own details before you send, or copy the prompt and fill them in first.
Here is the customer's message:
<message>
{{MESSAGE}}
</message>
What we can and cannot actually do:
{{WHAT_WE_CAN_DO}}
Write the reply.
- Open by stating their specific problem back in one sentence, in plain words, so they know it was read. No "I understand your frustration".
- If we got something wrong, say what, once. No excuses about volumes or systems.
- Say exactly what happens next: who does what, and by when.
- Offer only what is listed as possible above. If they asked for something we cannot do, say so directly and give the closest thing we can.
- Under 150 words. No exclamation marks.
After the reply, separately:
- CHECK FIRST: anything in their message that should be verified before sending, such as an order number or a claim about what someone told them.
- IF THEY REPLY ANGRIER: the one thing the next reply should add.End of prompt: Reply to an angry customer without grovelling or stonewalling
Builds a reusable reply for a question you answer every day, with clearly marked places to personalise so it never goes out generic.
Reads a period of complaints for the causes underneath them, separates one-offs from systemic problems, and sizes each one honestly.
New prompts as they go up, plus founder essays on shipping AI systems. ~2 emails / month.
Prompts get you started. If you want this wired into how your team actually works, see what we do.