Cold Email Prompt Generator
Build a specific cold outreach email prompt — recipient, goal, tone, and one real value proposition, ready to paste into ChatGPT or Claude.
A cold email that gets read depends on a genuinely specific reason for reaching out and a short, low-friction ask — not a generic pitch with the recipient's name swapped in. This tool builds a prompt for a cold email structured around a specific trigger, a clear value point, and one easy next step.
How to use it
A specific reason for reaching out this week, not just a general pitch, is what separates cold email from spam.
Kept to one clear point — a cold email with three asks reads as a sales pitch, not genuine outreach.
Aim for something the recipient can read in under 20 seconds on a phone.
Tips for better results
- Lead with the recipient's situation, not your company. Opening with what you noticed about them outperforms opening with a description of your company.
- Ask for one small, specific next step. A specific short time window converts better than an open-ended "let me know if you're interested."
- Keep it under 100 words. Cold emails lose most of their readers by the third paragraph — brevity is doing real work here.
Example output
“Subject: Quick question about your Q3 hiring push. Hi [name] -- noticed you're scaling the sales team fast this quarter. We help teams like yours cut onboarding time in half. Worth a 15-minute call next week?”
TL;DR
A cold email lives or dies in the first two lines. Most AI-generated cold emails fail because the prompt behind them is vague — “write a cold email about my product” gives the model nothing to work with, so it defaults to generic filler: “I hope this email finds you well,” a paragraph about your company, and a soft, forgettable ask.
Why generate a prompt instead of the email directly
Asking an AI model to “write me a cold email” skips the thinking that makes cold email work: who exactly you’re reaching, what they care about, and what a reasonable next step looks like. This tool forces those decisions into the input fields first, so the resulting prompt already contains the constraints that keep AI writing from sounding like AI writing — a specific value proposition instead of a vague benefit, and an explicit ban on hedge-phrases and hype words.
This matters more for cold email specifically than for most other writing, because the reader has no existing relationship pulling them toward finishing the message — every sentence either earns the next one or loses the reader. A prompt that front-loads a real, specific value proposition gives the model a reason to keep every line pulling its weight, instead of drifting into the polite throat-clearing that fills most unsolicited business email.
What each field is doing
- Goal changes the closing ask — a meeting-booking email ends differently than a dormant-thread reconnection.
- Tone shifts word choice and formality without touching the offer underneath it.
- Personalization detail is the field most people skip, and the one that separates a mail-merge-feeling email from one that reads like it was written for that person.
- Length matters more than senders assume — most cold emails are read on a phone in under ten seconds.
Filling in a project management tool for creative agencies, agency owners with 5-20 employees as the recipient, “Book a meeting” as the goal, and a value proposition of halving time spent on status-update meetings produces:
“Write a cold outreach email for a project management tool for creative agencies.
Recipient: creative agency owners, 5-20 employees.
Goal: Book a meeting.
Tone: Direct and confident.
Lead with the strongest value first, not the pitch: Cuts time spent on status-update meetings roughly in half…”
Paste that into ChatGPT or Claude and the model has a specific value prop, a real goal, and explicit rules against hype language — not a blank slate to guess at.
Using this across ChatGPT, Claude, and Gemini
ChatGPT tends to lean warm and slightly wordy by default — if a draft runs long, add “keep it under 80 words” before generating. Claude follows tone instructions more literally, so “direct and confident” often reads more clipped from Claude than from ChatGPT. Gemini sometimes needs the value proposition restated as a direct instruction rather than background context.
FAQ
Will this write the actual email for me?
No — it writes the prompt you paste into ChatGPT, Claude, or another AI tool, which then writes the email. This gives you a draft to edit in your own voice rather than a template that reads identically to everyone else’s.
What if I don’t have a personalization detail?
Leave that field blank. The prompt still works — it just won’t include a personalization instruction.
Can I use this for a follow-up instead of a first-touch email?
Yes — set the goal to “Reconnect after no response” and describe the original context in the value proposition field.
Does the tone setting change the actual offer?
No, only how it’s delivered — the offer and value proposition stay the same across tones.
How many follow-ups should I send after a cold email?
Two to three spaced-out follow-ups is a common approach — re-run this generator with the “Reconnect after no response” goal for each one, varying the angle slightly rather than repeating the same message.
