Email is still the highest-ROI marketing channel — for every £1 spent, the average return is £36. But writing great emails is hard and time-consuming. AI makes it fast, consistent, and data-driven. Here’s exactly how to use it.
The AI Email Writing Framework
AI doesn’t just write emails — it helps you plan them, A/B test subject lines, personalise at scale, and analyse results. Master these five phases and your email output will never be the same.
Step 1: Define Your Campaign Goal
Every email has a job. Tell the AI exactly what that job is:
I’m writing a promotional email for [product/offer]. The goal is [click to purchase / sign up / book a call]. My audience is [describe]. The email should be [warm/direct/educational]. Write a campaign brief including: goal, primary CTA, secondary CTA, tone, 3 subject line options, and a 4-paragraph email structure.
Step 2: Generate Subject Lines (A/B Ready)
Subject lines determine open rates. Always generate at least 10 options and test two:
Write 10 subject lines for this email: [brief description]. Include: 3 curiosity-based, 3 benefit-focused, 2 urgency-based, and 2 question-based. Flag the strongest 2 for A/B testing.
Run your chosen two against each other for 4 hours, then send the winner to the remaining list.
Step 3: Write the Email Body
Write a [promotional / welcome / re-engagement] email for [audience]. Goal: [CTA]. Tone: [conversational / professional]. Structure: opening hook (1-2 sentences), problem agitation (1 paragraph), solution reveal (1-2 paragraphs), social proof (1 sentence with a testimonial or stat), CTA (1 button with 5 word label). Max 200 words. Do not use jargon.
Edit the output to add your brand voice, real customer names, and specific data points. AI gives you the bones — you add the personality.
Step 4: Personalise at Scale
True personalisation goes beyond [First Name]. Use conditional content blocks driven by subscriber data:
I have 3 subscriber segments: [segment 1 description], [segment 2], [segment 3]. Rewrite this email intro paragraph tailored to each segment’s specific pain points and goals.
Most ESPs support conditional content — show different paragraphs based on tags or custom fields.
Step 5: Analyse and Optimise
After each campaign, paste your metrics into the AI:
My email had a 24% open rate, 3.1% CTR, and 0.8% conversion. Industry average is 21% open / 2.5% CTR. What does this tell me and what should I test next?
AI turns raw numbers into actionable recommendations in seconds.
Recommended Tools
Klaviyo + AI features — E-commerce email automation
Mailchimp Content Optimizer — Built-in AI suggestions
Beehiiv — Newsletter-first with AI writing assistant
ChatGPT — Subject line testing, copy refinement
