AI Prompts for Cold Email Sequences

What's the difference between a cold email sequence that books 10 meetings a month and one that gets a 2% reply rate and mostly angry unsubscribes? It's almost never the product. It's sequence structure, timing, and the ability to say something genuinely relevant to the person on the receiving end.

ai prompts cold email sequence workflows can solve the second and third problems systematically—but only if you understand what makes a cold sequence actually work before you start prompting.

Quick-Start (Copy This Right Now)

This is the 5-email cold sequence framework that works for most B2B outbound scenarios. Fill in the brackets and generate each email separately for best results:

I'm building a 5-email cold outreach sequence for [your role] at [company] targeting [specific ICP: job title, company type, size]. My product/service: [one sentence]. The primary problem I solve for this persona: [one sentence]. The outcome they typically see: [specific result with numbers if possible].

Design a 5-email sequence with:
- Email 1 (Day 1): First touch—lead with relevance to their role, introduce one specific problem
- Email 2 (Day 3): Add value—share a relevant resource, insight, or case study. Don't just follow up.
- Email 3 (Day 7): Different angle—change the framing entirely. Try a pain point they might not associate with your product.
- Email 4 (Day 14): Social proof—one specific customer story relevant to their situation
- Email 5 (Day 21): Breakup email—short, direct, gives them an easy out

For each email: write the subject line, the full email (under 120 words), and one sentence explaining the strategic goal of that specific touch.

Tone: direct, peer-to-peer. No bullet points. No subject lines with "Quick question" or "Following up."

What this does: The "one sentence explaining the strategic goal" instruction forces the AI to be intentional about each email's function rather than just producing five variations of the same message. That strategic clarity is what separates a sequence from a series of follow-ups.

⚡ Pro tip: Before running this prompt, do 10 minutes of ICP research with this: "List the top 5 business problems [specific job title] at [company type] is likely dealing with right now, based on current [industry] conditions. For each problem, write one sentence on how it might affect their priorities around [your product category]." Use the most urgent-sounding problem as your Email 1 hook.

Understanding the Variables

[ICP specificity]: The more specific your ICP description, the better the output. "Marketing managers" is useless. "Director-level demand generation managers at B2B SaaS companies between $10M-$50M ARR who are likely running Google and LinkedIn paid programs" is usable. AI can't write a relevant email to a vague persona.

[The outcome with numbers]: "Better results" is not a CTA-worthy promise. "Customers typically reduce paid CAC by 20-30% in the first 90 days" gives a prospect something concrete to evaluate. If you don't have this data yet, the prompt still works—just be honest about what the outcome looks like without fabricating a number.

[Subject line constraints]: "Quick question" and "Following up" are the two highest-volume cold email subject lines in every inbox. They signal immediately that what follows is a mass email. If you're seeing these in your AI output, add this: "Rewrite each subject line so it sounds like it was written specifically for this recipient—reference their role, their company's situation, or a relevant industry context."

⚡ Pro tip: For the Day 3 "add value" email, use a separate prompt to generate the content: "Write a 3-bullet insight email on [topic relevant to ICP]. Each bullet should be one specific, counterintuitive observation that a [ICP role] would find genuinely surprising—not generic advice they've seen before. Source or frame each insight as something you've learned from working with [customer type]."

Step-by-Step: Building Each Email in the Sequence

Email 1 — First Touch:

Write a cold email first touch from [sender name, title] to [prospect role] at a [company description]. Opening context to reference: [one specific, verifiable thing about them—company news, job posting, LinkedIn post, tech stack]. Problem to lead with: [specific problem statement]. Do not mention the product until sentence 3. Subject line: under 7 words. Body: under 100 words. CTA: one specific, low-commitment next step.

Email 3 — Different Angle:

Write cold email #3 in a sequence to [prospect role]. My previous emails addressed: [summarize email 1-2 angles]. This email must take a completely different angle—either a different problem, a different framing, or a different type of value (e.g., if previous emails were problem-focused, make this one insight-focused). Do not reference the previous emails directly. Keep it under 80 words.

Email 5 — Breakup:

Write a breakup email for a cold sequence targeting [prospect role]. It should: (1) be under 50 words, (2) acknowledge that now might not be the right time, (3) offer to come back in [timeframe] if they'd like, (4) leave the door genuinely open without being passive-aggressive. The tone should make the prospect feel respected, not guilty.

What this does: Each email in a sequence has a distinct job. Separate prompts for each email produce better output than asking AI to write all five in one pass, because each prompt can specify the unique constraints and goals for that specific touch.

⚠️ Common mistake: Writing a cold email sequence that's actually just the same email sent five times with slightly different subject lines. If someone looked at all 5 emails side by side and they follow the same structure, cover the same problem, and make the same ask—that's not a sequence, it's persistence. True sequences move the prospect through a logical arc.

Pro-Level Variations

Personalization at scale:

I have 20 prospects in the same ICP ([describe]) but each works at a different company. Write a cold email template with [3] personalization variables in [brackets] that I can quickly customize for each prospect. The email should feel personal even when the brackets are only 5-10 words each. Core message stays consistent; personalization creates relevance.

Trigger-based email:

Write a cold email triggered by the following event: [prospect just posted on LinkedIn about X / their company just announced Y / they're hiring for Z role]. The email should reference this trigger naturally in the first sentence and connect it to a problem my product solves. Don't be creepy about the monitoring—frame it as genuine relevance, not stalking.

⚡ Pro tip: Test two version of Email 1 simultaneously on your next prospecting push—one with a problem-first opening, one with a specific insight as the hook. Track reply rates for 30 days. The winner becomes your primary template; then test a variation of the winner. Systematic testing beats both intuition and "best practices" for your specific ICP.

Troubleshooting Common Issues

Reply rate under 3%: The problem is almost always specificity. Your emails sound like they could be sent to 500 people. Add more ICP-specific language and a more specific trigger or context.

Getting replies but not meetings: Your CTA is too broad. Change "let me know if you'd like to chat" to a specific day/time proposal or a specific question that requires a yes/no answer.

Lots of unsubscribes from Email 3+: Your sequence is wearing out its welcome. Either your value emails aren't providing actual value, or your timing is too aggressive. Try extending days between touches.

Your Turn

Build your first sequence with the Quick-Start prompt above. Run each email as a separate generation for the most focused output. When you find a sequence structure that books meetings consistently, save each email prompt in PromptABCD with notes on the ICP, the hook angle, and the conversion data. That's your organization's outbound playbook—built from what actually works, not what someone said should work.

Measuring Your Sequence and Iterating

A cold email sequence that doesn't get measured is just an opinion. Use this prompt to close the feedback loop:

My 5-email cold email sequence targeting [ICP] had these performance metrics:
- Email 1: [X]% open rate, [Y]% reply rate
- Email 2: [X]% open rate, [Y]% reply rate
- Email 3: [X]% open rate, [Y]% reply rate
- Email 4: [X]% open rate, [Y]% reply rate
- Email 5: [X]% open rate, [Y]% reply rate

Diagnose: Where is the sequence losing people? What does the drop in open rates between email 1 and later emails tell us about subject line effectiveness vs. content engagement? Which email should I rewrite first, and what specific change would most likely move the needle?

What this does: Lets real performance data drive iteration rather than intuition or generic best practices. The drop pattern between open rates and reply rates tells a very specific story about where your sequence is breaking down.

⚡ Pro tip: If your reply rate is higher on later emails than earlier ones, your sequence may be front-loaded with too much selling. Prospects who don't reply to Email 1 often aren't against the category—they're not yet warm enough. Try moving your most compelling value proof to Email 2 or 3 and see if reply rates shift earlier in the sequence.