According to McKinsey's 2025 State of AI Global Survey, 88% of organizations reported regular AI use for at least one business function in November 2025.
But nearly two-thirds of respondents also said they hadn’t yet begun scaling AI across their organization, meaning 2026 is still a year of experimentation.
As many brands experiment with ChatGPT for email marketing, they may find they’re getting competent drafts that remain basically generic because they don’t reflect who their customers are or how they behave.
ChatGPT doesn’t have access to your subscriber list by default, which means it doesn’t take into account real browsing behavior, purchase history, or stated preferences. It doesn’t write to anybody in particular, unless you go to great lengths to describe each audience segment in painstaking detail.
In this guide, we describe how to best use ChatGPT for email marketing drafts. Then, we expand on how to personalize those drafts with real customer data, so you can get the most out of your efficiency gains with AI.
What is ChatGPT for email marketing?
ChatGPT is an OpenAI chat interface built on top of its GPT large language models. Marketers use ChatGPT for a variety of email marketing tasks:
- Email strategy: ChatGPT can help marketers organize content calendars around major announcements or seasonal campaigns.
- Ideation: ChatGPT can help marketers connect strategy to content through brainstorming and suggesting relevant offers.
- Drafting: ChatGPT can come up with a first draft that marketers can use as a basis for personalization with real customer data.
ChatGPT may be used as a standalone tool for email marketing tasks, but it may also be used with a Model Context Protocol (MCP) server, which connects brands and their marketing data to the AI interfaces they already use. MCP servers are best used for getting email performance data through ChatGPT conversations.
How to generate an email marketing draft with ChatGPT
ChatGPT email drafts are helpful when you can save real time on generating a copy foundation. Where they fall short is personalization with real customer data.
Here’s where you can start with generating a quick first draft from a prompt as a brief:
- Subject line and preview text variations: Describe your campaign goal, and ask for at least 8–10 variations you can test.
- Body copy: Describe your offer, your audience, and your angle. You won’t be able to personalize based on real-time customer data, but ChatGPT should still get a sense of who the draft is for.
- Seasonal angles: If it’s relevant, specify any seasonal details to include references to different holidays or milestones.
- Past examples: If something has worked in the past, let ChatGPT know so it can work in elements of success into your current draft.
- Tone adjustments: Fine-tune your draft with additional prompts that adjust for tone. For example, you may want to see longer or shorter variations that sound friendlier, more exclusive, or more urgent.
Here's what a simple prompt can look like for a campaign:
For an email campaign, write 8 subject lines and 150 words of body copy for customers who bought our Merino wool base layer in the last 6 months and haven't ordered since. The offer is 20% off matching Merino socks, and it expires on [date]. The socks are 80% Merino, come in 3 colors, and we don't sell them as a bundle. The tone should be direct, without exclamation points, and no puns. Don't mention free shipping, because we don't offer it. Here are 3 subject lines that worked for us last quarter: [paste].
Remember to be as detailed as possible. ChatGPT often fills in any blanks by making up facts, so you’ll want to eliminate hallucinations as much as possible by providing it with real information and explicitly stating any guardrails it should keep in mind.
How customer data improves AI email marketing
Customer data is the backbone of personalization. And according to Klaviyo’s Future of Consumer Marketing Report, 74% of consumers expect personalized experiences from brands. The time you save drafting emails with AI is only worth what you can deliver on personalization.
If you want to save time with AI and see actual revenue gains, your AI email drafts need to be drawing on customer data that's current: what customers are clicking on, what's sitting in their carts, or what they’re telling support.
That kind of context ideally exists in a unified customer profile inside a CRM with both a built-in customer data platform (CDP) and built-in agentic AI for marketing. If your CRM, marketing automation platform, and AI marketing assistant live in separate silos without a unified view, AI email marketing will inevitably fall flat because your drafts won’t incorporate the customer segmentation that’s so necessary for personalization.
Here’s what you can expect to see when your AI for marketing is built into your CRM, so your AI email drafts are based on real customer data:
- Copy that reflects real scenarios: Drafts draw from existing customer segments, your product catalog, and actual performance history.
- Offers that match actual segments: A first-time browser and a lapsed VIP get fundamentally different offers that encourage unique next steps for each segment.
- Subject lines and CTAs tuned to real behaviors: Your subject lines and CTAs are informed by past behavior that actually converted, rather than ChatGPT’s best guess.
How to incorporate mobile messaging to your AI email marketing
When you’re working with an AI-first B2C CRM, not just copy/pasting outputs from ChatGPT into your marketing automation platform, the same real-time customer data that can personalize AI-generated emails can also decide whether a given subscriber should even receive an email in the first place.
This is where AI capabilities like channel affinity and audience optimization come in. Channel affinity predicts which channel a customer is most likely to engage with next, based on their historical behavior with your brand, while audience optimization automatically refines your campaign audiences to remove likely-to-unsubscribe profiles before sending.
As part of your AI email marketing process, here’s how you can incorporate other channels, like text messaging, mobile push, or WhatsApp:
- Draft AI-generated email and text messages in one AI session. Prompt your AI marketing agent to write toward overall campaign goals, and generate messages for email, texting, WhatsApp, and push at the same time.
- Use channel affinity to automatically prioritize by customer preference. If a customer is more likely to respond via text message or WhatsApp, your built-in AI marketing agent can use AI-powered channel affinity to send a mobile message first with an email follow-up.
- Reduce message fatigue with audience optimization. Customers with low engagement on a given channel don't get drafts generated for it, which cuts down on fatigue and unsubscribes that come from blasting every channel at once.
Corkcicle, for example, uses channel affinity to target campaigns by channel engagement. Their most efficient segment is now customers who’ve engaged in the last 90 days who prefer email. “Channel affinity allows us to be efficient and targeted and has provided great results,” says Erica Olsen, ecommerce marketing specialist for Corkcircle.
Standalone ChatGPT vs. AI email marketing built into your CRM
Marketing activity | Standalone ChatGPT | Built-in AI email marketing agent |
|---|---|---|
Customer data | No access | Full access to real-time data |
Personalization depth | Limited to prompting | Based on real segments, product catalog, and performance history |
Sending | No sending capabilities; must send in a separate marketing automation platform | Drafts launch-ready emails within a marketing automation platform |
Compliance guardrails | Opaque ChatGPT guardrails, no custom brand guardrails | Custom guardrails based on product exclusions, legal disclaimers, banned phrases, and regional compliance |
Omnichannel messages | Manual prompting and sending | Builds email and text message campaigns and sends automatically to each customer’s preferred channel |
Measurement | No measurement, drafting only | Performance measurement with embedded marketing analytics accessible through an MCP |
Draft and launch personalized AI-generated emails with Klaviyo
Klaviyo is the autonomous B2C CRM that brings together real-time customer data, built-in agentic AI, and marketing automation so brands can power more valuable customer experiences across email marketing and more.
Klaviyo Composer is an AI marketing agent that can identify your biggest growth opportunities and turn them into ready-to-launch campaigns in minutes. With Composer, you can:
- Send emails based on real customer data. Every output is informed by your real segments, catalog, and performance history.
- Draft more than just copy. Composer builds the audience, the copy, channel coordination, and timing, all in one conversational session.
- Send omnichannel campaigns and flows. Build multi-step cross-channel message sequences with channel-appropriate content, timing, and sequencing in a single generation session.
- Measure what each send actually earned. Klaviyo Marketing Analytics reports across channels, so you know what needs to improve for your next campaign.
Personalize your AI-generated emails with Klaviyo. Get a demo