The Cold Email Prompt That Converts (Copy-Paste Formula)
Most ChatGPT cold emails get 2-3% replies. This framework gets 15-20%. Here's why.
Part of our content marketing prompts series. See also: [LinkedIn viral post prompt](/blog/linkedin-viral-post-prompt), [SEO blog prompt](/blog/seo-blog-post-prompt), and [general prompt writing framework](/blog/write-better-prompts).
Why ChatGPT Cold Emails Usually Fail
ChatGPT defaults to generic: 'I noticed your company, we do X, let's talk.' Every AIDA element (Attention, Interest, Desire, Action) is checked but none land. The email is correct but hollow.
High-converting cold emails use emotion, specificity, and urgency. ChatGPT doesn't do that unless you structure it.
The AIDA Structure (For ChatGPT)
- **A = Attention.** Open with a spike — a compliment (genuine), a question (real), or an observation (specific).
- **I = Interest.** One concrete reason why *they* care (not why your product is cool).
- **D = Desire.** Show what success looks like without you vs. with you (contrast).
- **A = Action.** One ask. Reply, call, visit link. No pressure. Easy to say yes.
The Prompt Structure (Template)
Write a cold email from [YOUR NAME] at [COMPANY] to [PROSPECT NAME] at [PROSPECT COMPANY].
Use this structure:
ATTENTION (line 1):
[Specific observation about their company / genuine compliment about their recent work / question that shows you've done research]
INTEREST (line 2-3):
[Why this person cares: Explain the pain point YOU know they have, using language they use]
DESIRE (line 4):
[Show contrast: Before (current state) vs. After (with solution). Don't mention your product yet.]
ACTION (line 5):
[One ask: 'Reply with your thoughts,' 'Jump on a 15-min call,' 'Forward this to the person who owns X']
Constraints:
- Max 100 words total
- No "check us out" language
- No "best-in-class" or superlatives
- Sound like a human who's done research, not a marketer
- Use contractions (I've, you're, they'll) to feel conversational
Real Example: SaaS Outbound
Prompt:
Write a cold email from Sarah at NavedNaik.com to James at TechStarups.io.
ATTENTION: James launched the 'AI for Operations' guide last week — it hit 8k upvotes on HN.
INTEREST: Most ops teams spend 60% of time on manual prompts and workflows. You know this problem.
DESIRE: Show contrast: 'Right now, your ops team manually writes prompts for each task (slow, inconsistent). With structured prompts, they move 3x faster and stay on brand.'
ACTION: 'Curious if this would work for your community? Happy to show you how 3 ops teams did it.'
Max 100 words. Conversational tone. No 'best-in-class'."
Result (ChatGPT output):
"Hi James, Just finished your 'AI for Operations' guide — 8k upvotes on HN is well-deserved. I work with ops teams solving one core issue: they're spending 60% of time on manual prompt-writing. It's slow and inconsistent. Structured prompts change that. Three ops teams we worked with moved 3x faster and kept everything on brand. Thought you might see this working for your community. Want a 15-min walkthrough? — Sarah"
That's 75 words. Specific. Conversational. Cites their work. Shows contrast. Easy ask. Expected reply rate: 12-18%.
Pro Adjustments by Industry
- **For B2B SaaS:** Lead with recent product launch or feature, tie to pain point, show time/cost savings.
- **For Agencies:** Lead with recent client work or portfolio piece, tie to their client outcome, show portfolio value.
- **For Consulting:** Lead with industry insight or research you noticed, tie to their business model, show revenue impact.
- **For Services:** Lead with specific case study, tie to their project type, show outcome (quality, speed, cost).
Testing and Iteration
Generate 5 variations with this prompt. Send to 25 people (5 of each variation). Track replies. Keep the highest-reply version. Reuse its structure.
One team went from 3% to 18% reply rate by testing 3 ATTENTION hooks with 75 prospects each. Data wins.
The Mistake That Kills Cold Email Success
Most people use the AIDA structure but forget to personalize the ATTENTION and INTEREST sections. They generate one version, send it to everyone, and wonder why the reply rate is 2%. The secret: Change the ATTENTION and INTEREST for every prospect list. One hook for B2B SaaS founders. A different hook for agencies. Different hooks for different industries. Lazy personalization = lazy reply rates.
Scale Your Outreach
Once cold email is working, the next step is lead qualification. Try our [Lead Capture tool](/automation/lead-capture) to automatically qualify inbound leads while you're building your email list.
Questions
What reply rate should I expect?
Baseline: 5-7% (if you follow AIDA). With strong research and relevance: 12-18%. Outliers: 20-25% if the list is warm and your solution is a perfect fit.
How many emails should I send before judging the prompt?
Minimum 20. Ideally 50. Each list has different response behavior. Don't optimize after 5 replies.
Should I mention price in the cold email?
No. Price invites 'too expensive' objections before you've qualified. Save it for the call. Keep cold email focused on relevance.
Can I use this for LinkedIn outreach too?
Yes, but shorten it. LinkedIn opens have 60% lower open rates than email, so compress AIDA to 2-3 sentences max.