ChatGPT Prompts for Email Marketing That Don't Sound Like Every Other Newsletter

The short version: ChatGPT can draft a solid email marketing sequence in minutes if you feed it your actual data, tone, and offer instead of a vague one-line request, and I'll give you the exact prompts I use for welcome emails, promotions, newsletters, and win-back campaigns below. The uncomfortable bit nobody tells you is that a better prompt won't fix a dead list, and a generic prompt will always read like a generic prompt no matter how clever the wording is.

Useful alongside this: ChatGPT Prompts for Writing Newsletters People Open.

Useful alongside this: ChatGPT Prompts for Email Marketing Campaigns: 25 That Get Opened, Rea.

Why most people get worse results from ChatGPT than I do

I get asked for email prompts more than almost anything else. People come to me with the same complaint: they typed "write me a marketing email" into ChatGPT, got something that reads like it was written by a committee of nobody, and gave up. That's not a ChatGPT problem, it's an input problem. I've written email copy professionally for over fifteen years, and the difference between a prompt that produces something usable and one that produces filler is almost always specificity. Not cleverness. Specificity.

When I rebuilt my own list after a rough few years, I wasn't precious about using AI to draft first passes. My subscriber count had dropped from around 60,000 pre-2020 to just under 4,300 by the time I started rebuilding in public. I used ChatGPT to draft the first version of almost every email in my relaunch sequence. Open rates on that sequence sat between 34% and 41%, which is well above the UK average of around 21-25% for the marketing sector reported by the Data & Marketing Association. That wasn't the prompt working magic. It was me giving the prompt real material: my actual voice, three years of context on why I'd gone quiet, and a small, warm list that still liked me. AI drafted the sentences. I supplied everything that made people open the email.

The prompt structure that produces usable drafts

Before the copy-paste prompts, here's the framework behind all of them. A good email prompt has five parts, and skipping any one of them is why most people's output sounds hollow:

• Who the email is going to, described like a real person, not "my audience"
• What you want them to do after reading it, one action only
• Why now, the specific trigger or context for this email today
• Tone, described with an example rather than an adjective ("write it like I'm texting a mate who runs a similar business, not like a brand")
• Constraints, word count, subject line requirement, no clichés, no exclamation marks unless you say so

If you're building prompts for other parts of your business too, the same five-part structure works for social captions and general business tasks. I use a near-identical framework in my ChatGPT prompts for small business owners, and it's worth learning once and reusing everywhere.

Welcome email prompts

The welcome email is the highest-opened email you'll ever send to that person, typically 4 to 5 times higher open rate than a regular newsletter, so it's worth the extra ten minutes.

• "Write a welcome email for someone who just signed up for [lead magnet name] because they want [specific outcome]. Introduce me in three sentences max, tell them one specific thing to expect from future emails, and end with one small, easy action they can take today. Tone: warm, direct, no corporate language, write like a real person emailing a new acquaintance. 150 words max."
• "Write a three-email welcome sequence for new subscribers to [business name], a [industry] business. Email 1: who I am and what they'll get. Email 2: my most useful free resource or tip, something they can use in under ten minutes. Email 3: a soft introduction to [product/service] with no hard sell. Each email under 200 words, one clear call to action per email."

Promotional and sales email prompts

This is where AI-written emails go wrong most often, because "sell my thing" as a prompt produces the same five sales tactics everyone else is already using: urgency, scarcity, a countdown, three bullet points of benefits. Ask for specificity instead.

• "Write a promotional email for [product/service], priced at [price], for people who have already [specific past behaviour, e.g. downloaded my guide/attended my webinar]. Lead with the problem they told me about, not the product. Include one real detail about the product (not a generic benefit list). End with a single clear call to action and a genuine reason for the deadline, not a fake countdown. 200 words."
• "Write a subject line and preview text for an email announcing [offer], aimed at people who opened my last three emails but haven't bought anything yet. Give me 8 subject line options, mix of curiosity, direct benefit, and question formats, under 50 characters each."

Newsletter prompts

Regular newsletters are where consistency matters more than brilliance. I send weekly, and the honest habit that keeps it going is a repeatable template, not fresh inspiration every Monday.

• "Write a newsletter for this week using this structure: one personal opening story (I'll give you the story), one useful tip related to [topic], one link to [content/product], one short sign-off. Here's the story: [paste your real story]. Keep the personal section under 100 words and don't turn it into a metaphor for the tip, just tell it straight."
• "Turn this blog post into a 300-word newsletter email. Pull out the single most useful idea, not a summary of everything, and write it as if I'm explaining it to one specific reader over coffee. Paste post: [paste text]"

If you're already repurposing blog content into email, it's worth repurposing it into social content at the same time. I cover the same idea from a different angle in ChatGPT prompts for Instagram captions that get read, since a lot of the "pull the one useful idea" logic transfers straight across.

Win-back and re-engagement prompts

Every list has dead weight. Roughly 20-25% of most email lists haven't opened anything in six months, and that group is dragging your deliverability down every time you send, because inbox providers like Gmail read low engagement as a signal you're spam. A win-back sequence isn't optional if you want the rest of your emails landing in the inbox rather than promotions or junk.

• "Write a re-engagement email to subscribers who haven't opened an email from me in 4 months. Be direct about it, don't pretend nothing happened. Ask one honest question about whether they still want to hear from me, and make unsubscribing easy and guilt-free in the copy. Under 120 words."
• "Write a final win-back email, the last one before I remove inactive subscribers from my list. Tell them plainly this is the last email unless they take an action, give them one clear button/link to click to stay, and don't apologise for the list clean-up, frame it as good for them too (less inbox clutter)."

Segmentation and personalisation prompts

This is the part where I have to say something people don't want to hear: ChatGPT cannot personalise an email for someone it knows nothing about. If you paste "write a personalised email" with no data attached, you get an email with the word "personalised" bolted onto generic copy, usually just their first name in the subject line. Real personalisation needs real segments, built from actual behaviour, purchase history, or survey answers, fed into the prompt as data.

• "Write three versions of the same promotional email for [product], each written for a different segment: Segment A bought from me once over a year ago and hasn't since. Segment B has opened every email for 3 months but never bought. Segment C is a repeat customer. Adjust the opening line and the offer angle for each group, keep the core offer and price the same."
• "I've attached survey answers from 50 subscribers about their biggest struggle with [topic]. Group the answers into 3-4 themes and write a short email for each theme, addressing that specific struggle directly. [Paste survey data]"

Where I've built full automated workflows around this

Once the prompts work, the next question is always the same one clients ask me: can this run without me typing it fresh every week? Mostly yes. I've helped teams connect ChatGPT output into their email platform through simple automation, so a new blog post or product update triggers a drafted email automatically, which someone then edits and approves rather than writing from scratch. I go through real examples of what that looks like, including the time saved on a weekly basis, in AI automation for marketing teams. For a five-person marketing team, that setup has typically saved 6 to 8 hours a week that used to go on first drafts alone.

If you'd rather have someone build that workflow for you instead of piecing it together yourself, that's the kind of thing I do hands-on with clients, and you can see how that engagement usually works on my AI implementation coach page.

The editing step nobody wants to add

Here's the uncomfortable truth about all of the prompts above: none of them are finished emails. Every single draft ChatGPT gives you needs a human pass, and skipping that pass is the single biggest reason AI-written email marketing underperforms. Not because AI writes badly, it doesn't, but because it defaults to safe, smooth, slightly hollow phrasing when it has no specific facts to work with. Read back the drafts above and you'll notice they lean on your input for every good bit, the real story, the real price, the real segment data. Take those out and you get filler.

My rule with clients: read every AI-drafted email out loud before sending. If it doesn't sound like something you'd say to a customer standing in front of you, rewrite the sentence, don't just tweak a word. I do this with niche-specific prompts too. When I built out sets for photographers and yoga teachers, the prompts that worked best were the ones asking for one specific, unglamorous detail (a class time, a shoot location, a booking deadline) rather than the ones asking for "engaging, professional copy". Specific beats polished every time in email.

A quick step-by-step for your next campaign

• Write down who this email is for in one sentence, using a real subscriber if you can picture one
• Paste your last three sent emails into ChatGPT and ask it to describe your tone in a few phrases, then reuse that description in future prompts
• Use the five-part prompt structure above: who, what action, why now, tone, constraints
• Ask for 3 to 5 variations of the subject line and pick the least "salesy" one that still says what the email is about
• Read the draft aloud, cut anything that sounds like it could apply to any business, and add one real detail back in

Related guides live in the ChatGPT Prompts and Limits: 30 Guides to Fun Prompts, Photo Edits and Daily Caps.

The free prompt builder does the working out for you.

Related: email marketing write for us.

If you write on this topic, see Write for Us Digital Marketing Blog.

Related: writing for us on news.

Frequently asked questions

Can ChatGPT write a full email marketing campaign on its own?

It can draft one, but it can't run one. ChatGPT has no access to your list's actual open rates, past purchase behaviour, or unsubscribe reasons unless you paste that data into the prompt, so treat every output as a first draft that needs your real information and a human edit before it goes out.

What's the best ChatGPT prompt for subject lines?

Ask for 8 to 10 options across different formats (curiosity, direct benefit, a question, a number) rather than one, and give it the actual email content first so the subject line matches what's inside rather than overselling it.

Will AI-written emails hurt my open rates or deliverability?

Not because they're AI-written, but generic, low-specificity emails do get lower engagement, and low engagement is what damages deliverability with providers like Gmail. The fix is specific, edited copy, not avoiding AI altogether.

How much should I edit a ChatGPT-drafted email before sending it?

Assume at least one full rewrite pass every time. Read it aloud, remove any sentence that could apply to a competitor's business unchanged, and add back one real detail, number, or story that only you could have written.

Further reading

• the UK Information Commissioner's guidance on direct marketing
• OpenAI's own platform documentation