Open the optimiser

Prompt: Reply to a Difficult Email

A prompt for drafting a calm, clear reply to a frustrated or confrontational email without either caving or escalating.

Sales and customer support ChatGPTClaude
Updated Sep 2026 PIS score 83
The prompt
Open in Studio 
You are helping draft a professional reply to a difficult email.

Objective: Write a reply that addresses {{THE PERSON'S}} concern about {{WHAT THEY'RE UNHAPPY ABOUT}} while staying calm and factual.

Context: Your relationship to this person is {{RELATIONSHIP — e.g. client, colleague, manager}}. What you can actually offer or commit to: {{WHAT YOU CAN OFFER}}.

Constraints:
- Acknowledge their concern in the first two sentences, specifically, not generically
- Do not apologise for things that were not actually your fault
- Do not promise anything beyond what was stated you can offer
- Keep it under 150 words

Output format: Return the email with a subject line, ready to send.

Before finishing, check that nothing in the reply commits to something beyond what was stated as available.

Their email:
{{PASTE THEIR EMAIL HERE}}

Example

Input

THE PERSON'S: client's
WHAT THEY'RE UNHAPPY ABOUT: a second delay on their project delivery with little notice
RELATIONSHIP — e.g. client, colleague, manager: client
WHAT YOU CAN OFFER: a revised delivery date one week from now and a 10% discount on the current invoice
PASTE THEIR EMAIL HERE:
  This is now the second time our project has been delayed with barely any notice. I need to know today whether you can actually deliver, or if we should look elsewhere.

Output

Subject: Update on your project timeline

Hello,

I understand — a second delay with little notice is genuinely frustrating, and you're right to want a clear answer today rather than another vague update.

I can confirm a revised delivery date of one week from now, and I've applied a 10% discount to your current invoice in recognition of the disruption. I don't want to promise anything beyond what I can actually guarantee, so this is the date I'm confident we can hit.
Show more 
Let me know if that works, or if you'd like to discuss it further.

Best regards

Why it works

  • Separating 'what you can actually offer' from the reply-writing step stops the model improvising a concession you never authorised.
  • Banning apology for things that weren't your fault keeps the tone calm and factual rather than defensive or over-apologetic.
  • A hard word limit forces a direct, specific reply instead of a padded one that avoids actually committing to anything.

Why the basic version fails

“Write a polite reply to this angry email” usually produces something either so apologetic it concedes points that weren’t wrong, or so defensive it escalates things further. Neither actually resolves anything.

What this prompt fixes

Separating “what you can actually offer” from the reply-writing step stops the model from improvising a concession on your behalf. The “don’t apologise for what wasn’t your fault” constraint keeps the tone calm without being a doormat.

Tips per model

  • ChatGPT is good at matching a specific tone if you describe the relationship precisely — vague relationship descriptions get generic replies.
  • Claude tends to respect hard limits (like the word count and “don’t promise beyond X”) more strictly than a vague request for brevity.
  • Paste the actual email rather than paraphrasing it — the specific words a frustrated person used often show you exactly which sentence needs the first, direct acknowledgement.