How to Use AI for Email Marketing That Actually Gets Opened

AI-written email marketing has a recognizable smell to it: over-polished subject lines, vague value propositions, and a tone that sounds like a company rather than a person. The goal when using AI for email is not to remove human writing — it is to write more human email, faster.

Start with subject lines. Give ChatGPT or Claude your email's main point, your audience, and three examples of subject lines that have worked for you before. Ask for ten options. From those ten, pick the two or three that actually sound like something you would write, then edit from there. This produces better subject lines faster than writing from scratch.

For email sequences, AI is useful for drafting the structural skeleton. Define the sequence goal (welcome, onboarding, re-engagement), the audience, and what action you want at each step. Ask Claude to draft the sequence. Review each email for anything that sounds generic, overly salesy, or inconsistent with your voice — those are the parts to rewrite.

For newsletters, build a consistent template: intro observation, main content section, recommendation or resource, and closing. Give Claude your notes or links for that week and ask it to draft within your template structure. The template keeps the AI constrained and produces a more consistent voice than a blank-slate prompt.

Kit (ConvertKit) and Beehiiv both have AI writing features built in, which is useful for staying in one tool. But the quality of drafts from standalone Claude or ChatGPT, then pasted into your email platform, is usually higher — the dedicated writing models are simply better than the bundled ones.

The most important principle: read every AI-drafted email out loud before sending. Anything that sounds awkward when spoken needs to be rewritten. Email that sounds like it was written by a human performs better, even when AI helped write it.