Most ChatGPT prompts for email marketing online are too generic to use. This library gives you 15 prompts I actually use with marketing teams: subject lines, welcome sequences, re-engagement flows, post-purchase emails, and the segmentation prompts that make all of them work harder.
Why most email marketing prompts you find online don’t work
Open Pinterest and you’ll find a thousand “ChatGPT prompts for email marketing” lists. I’ve tried most of them. The pattern is depressing. The prompts are written by people who have never sent a campaign that needed to actually convert. They produce safe, lifeless copy that no marketing director would ever send to a real list.
Here’s the thing nobody on those lists tells you: a good email prompt isn’t a clever sentence. It’s a brief. It tells ChatGPT who the reader is, what stage of the funnel they’re in, what action you want, and what your brand sounds like. If any of those four are missing, the output is generic. If all four are there, the output is usable.
Email marketing remains one of the highest-ROI digital channels. Industry benchmarks from Litmus and the DMA put the return in the $36-$42 range per $1 spent. [1] That order of magnitude is the reason we obsess over every subject line and every preheader. ChatGPT can save you hours, but only if you stop treating it like a magic button.
Below are 15 prompts I actually use with the marketing teams I work with. Copy them. Edit the bracketed parts. Test them against your own copy. Keep what works.
Subject line and preheader prompts
The first three are the highest-leverage prompts you’ll use. A 3% lift in open rate compounds across every campaign you send.
Prompt 1: The “10 angles” subject line generator
You are a senior email copywriter for [brand name and one-line description]. Write 10 subject lines for a campaign about [topic of the email] targeting [audience segment]. Vary the angles: curiosity, benefit, urgency, social proof, question, contrarian, personal, specific number, news, and value-first. Each subject line must be under 60 characters. No clickbait. Match this brand voice: [paste 50-100 words of voice sample]. After the list, briefly explain which 3 you'd test first and why.
What this prompt does that the lazy version doesn’t: it forces ChatGPT to give you a range, then tells you which to ship first. The angle list keeps the output diverse instead of producing 10 variations of the same line.
Prompt 2: Preheaders that actually pair with the subject
For each of these 3 subject lines, write 2 preheader options (max 90 characters each) that extend the subject line without repeating it. The preheader should answer "what's the very next thing the reader will see?" Subject lines: 1. [paste subject line 1] 2. [paste subject line 2] 3. [paste subject line 3]
Prompt 3: Subject line stress-test
I'm about to send this subject line to a list of [list size and segment]: "[paste subject line]" Critique it brutally. Score it 1-10 on: clarity, specificity, curiosity, voice match, and mobile preview (first 35 characters). Then suggest 3 stronger alternatives. Be honest, not polite.
Welcome sequence prompts
A new subscriber’s welcome flow is the most important email work you’ll ever do. People are most engaged in their first 7 days. Most brands waste this with a generic “Welcome to the family!” template. Don’t be that brand.
Prompt 4: 5-email welcome sequence outline
Design a 5-email welcome sequence for [brand]. Audience: [describe the new subscriber: what attracted them, what they expect, what stage of awareness they're at]. The end goal of the sequence is [specific business outcome, e.g. first purchase, demo booked, content engagement]. For each email, give: send timing, primary objective, subject line draft, one-sentence summary of the body, and the single CTA. Don't write the full body yet, just the strategic outline. Make sure each email has a distinct job; no two emails should overlap in purpose.
Prompt 5: Draft Email 1 (the “deliver on the promise” email)
Write Email 1 of the welcome sequence we just outlined. This email must: - Deliver whatever the subscriber signed up for (lead magnet, discount code, content link) - Set expectations: how often they'll hear from us, what kind of content - Include one human moment (a sentence that sounds like a real person, not a brand) - End with one clear next action, not three Length: 120-180 words. Brand voice: [paste voice sample]. Sender: a real person (first name + role), not "Team [Brand]".
If you want to go deeper on this, our guide on training AI on your brand voice walks through how to build a voice file you can reuse across every prompt.
Re-engagement and win-back prompts
Dormant subscribers cost you deliverability. Mailbox providers downgrade you if a chunk of your list never opens you. Re-engagement campaigns are uncomfortable to write because you have to be honest about the relationship. ChatGPT can help you find the right tone.
Prompt 6: The “are we still useful?” re-engagement email
Write a re-engagement email for subscribers who haven't opened anything from us in 90 days. The tone should be honest and slightly self-aware, not desperate. Goal: get them to confirm they still want to hear from us, or unsubscribe. Structure: acknowledge the silence in the first line, give them a clear "yes I want to stay" button and a clear "unsubscribe me" link. Mention 1-2 things they'd miss if they left, but don't oversell. Brand voice: [paste voice sample]. Length: 90-130 words.
Prompt 7: Sunset confirmation email
Write a final email for subscribers who didn't respond to our re-engagement email. Tell them: we're removing them from the active list, here's how to come back if they change their mind, no hard feelings. Keep it under 70 words. Make sure the tone is warm, not bitter.
Post-purchase and retention prompts
Retention emails get less attention than acquisition, which is exactly why the easy wins are here. Bain & Company’s classic research showed that a 5% increase in customer retention can lift profits by anywhere from 25% to 95% depending on the industry, because you stop paying acquisition costs to win the same customer twice. [2]
Prompt 8: Order confirmation that doesn’t waste the open
Write an order confirmation email for [brand], for a customer who just bought [product]. Transactional emails like order confirmations regularly post the highest open rates of any email type, with Klaviyo citing rates that can reach 50-60% [3], so don't waste this real estate. Include: clear order details, expected delivery window, and ONE secondary moment that adds value. Pick one of: a tip for getting the most out of their purchase, a quick "what to expect next" preview, or an invitation to a community/content. Pick the one that fits this product best and explain your choice briefly. Brand voice: [paste voice sample]. Length: 150-200 words.
Prompt 9: Cross-sell email that doesn’t feel salesy
Customer X bought [product A] 14 days ago. The natural next-step product for someone in this state is [product B]. Write a cross-sell email that: - Acknowledges they've had product A long enough to form an opinion - Doesn't assume they liked it (ask a soft question or reference common experiences) - Positions product B as the natural next step, not a generic recommendation - Has a clear single CTA Brand voice: [paste voice sample]. Length under 150 words.
Prompt 10: Replenishment / refill reminder
Customer's typical replenishment cycle for [product] is [N days]. They're [X days] into that cycle. Write a friendly reminder email that doesn't insult their intelligence. Include the one-click reorder link as the primary CTA.
Segmentation prompts that make every other prompt work harder
This section is the one most prompt lists skip entirely, and it’s the highest-leverage one. The single biggest reason ChatGPT email copy disappoints is that marketers feed it the same brief for every audience. Stop. Segment first, then write.
Prompt 11: Build a quick audience persona before you write
Here's what I know about a segment I want to email: - Demographics: [age range, role, region] - Behaviour: [recent purchases, content engagement, lifecycle stage] - Source: [how they joined our list] - Top objection (what stops them buying): [if known] Build me a one-paragraph persona for this segment. Then suggest: - The single biggest unmet need this segment has from us right now - The tone we should be using with them (vs our default brand tone) - One topic we should NOT email them about
Prompt 12: Segment-specific subject line variants
I'm sending this campaign to three different segments: [briefly describe each]. The campaign is about [topic]. For each segment, write 3 subject line variants that speak directly to their context. Explain what changed between segments.
A/B testing and analysis prompts
The work doesn’t stop when the email goes out. The best marketers I know spend almost as much time analysing results as they do writing campaigns. ChatGPT speeds up the analysis dramatically.
Prompt 13: Hypothesis-first A/B test design
I want to A/B test [variable, e.g. subject line, CTA copy, send time]. My current performance is [baseline metric]. My hypothesis is: [your guess about what will move the needle and why]. Help me design a clean test: - What's the single variable I should change? - What sample size do I need to detect a 10% lift with 95% confidence? - What's the minimum data point I need to call a winner? - What's the next experiment I should queue up if this one wins?
For more on this, our email subject line A/B test workflow walks through the full statistical sanity-checking process.
Prompt 14: Post-campaign analysis
Here are the results of my last campaign: [paste headline metrics: sends, open rate, CTR, conversion rate, revenue, vs baseline] Audience: [segment]. Hypothesis going in: [hypothesis]. Subject line and preheader: [paste]. Give me: - What worked - What didn't - The single biggest hypothesis I should test next - One question I should ask myself before sending the next campaign to this segment
Prompt 15: Quarterly email program review
Here are my email program metrics for the last quarter: [paste open rate trend, CTR trend, list growth, unsubscribe rate, revenue per email]. Act as a senior CRM consultant. Give me: - The 3 most important trends in this data - 1 thing I'm probably doing too much of - 1 thing I'm probably doing too little of - A 90-day action plan with 3 specific initiatives
Putting it together: a stack you can actually use
Don’t try to use all 15 prompts at once. The marketers I work with who get the most value pick three or four and run with them for a month before adding more. Here is the stack I’d start with if I were running a five-person marketing team:
Daily / weekly: Prompts 1, 2, 3 (subject lines and stress-tests). These pay back the minute you adopt them.
When a new sequence is needed: Prompts 4, 5 (welcome sequence design and drafting).
Monthly: Prompts 11 and 14 (persona refresh + post-campaign analysis).
Quarterly: Prompt 15 (strategic review with a fresh perspective).
That’s six prompts. Six prompts will replace about 60% of the writing work in a typical email program. Spend the time you save on the things ChatGPT genuinely can’t do: customer interviews, creative directions, and the tough strategic calls. For more on tool selection across the marketing stack, see our best AI tools for marketing teams guide.
Frequently asked questions
This guide was written by Hina Mian, Co-Founder of Future Factors AI, drawing on hands-on work with non-technical teams. It is updated periodically as the tools and the field move. Future Factors AI offers Bootcamps, Corporate Workshops, and Speaking & Consulting for teams getting practical with AI.
Sources
- [1] Litmus. The ROI of Email Marketing. 2022 (widely-cited industry benchmark).
- [2] Reichheld, F. (Bain & Company). Retaining Customers Is the Real Challenge. 2018.
- [3] Klaviyo. Order Confirmation Email Examples, Tips & Best Practices. 2024.
- [4] Litmus. State of Email Reports. 2025-2026.
- [5] HubSpot Research. Email Marketing Statistics. 2024.
