- AI can write your email sequences — but only if you train it on your voice first
- The problem isn't the AI. It's the briefing. Vague inputs produce generic outputs
- A brand voice document + 5 example emails is all it takes to get usable results
- The workflow: AI drafts, you edit for tone — not the other way around
The Problem With AI-Written Emails
You can tell within two sentences. The slightly-too-formal opener. The bullet points that go on three beats too long. The sign-off that sounds like customer service, not a person. AI-written emails have a signature texture — and your subscribers feel it even if they can't name it.
This is the #1 objection I hear from business owners when I suggest using AI for email: "It doesn't sound like me." And they're right. Out of the box, it doesn't. But that's a setup problem, not a capability problem.
The AI isn't bad at writing emails. It's bad at writing your emails. That's a different problem — and it's one you can solve.
Most people approach AI email writing backwards. They open the tool, type a vague prompt, get a generic result, conclude that AI can't write emails, and go back to doing it manually. The workflow I'm about to walk you through flips that sequence entirely.
Start With Your Voice, Not the Prompt
Before you write a single prompt, you need a brand voice document. Not a 40-page brand bible — just a one-page cheat sheet that tells the AI what it's working with. Here's what goes in it:
- Tone descriptors (3–5 words): How would you describe your writing style? Direct. Warm. No-nonsense. Slightly dry. Practical. Pick the words that feel true and use them every time.
- What you never say: Every brand has language that feels off. "Synergy." "Circle back." "Exciting opportunity." List the words and phrases that would make you cringe if you read them in your own email.
- 5 example emails you've actually written: The AI learns faster from examples than from instructions. Give it real emails — ideally ones you're proud of — and tell it: "Write like this."
- Your reader in one sentence: Who are you talking to? Not a persona with a name and a backstory. Just one honest sentence: "I write to established business owners who are serious about growth and short on time."
That's it. One page. Drop this into every AI session before you start writing emails and you will immediately see a difference in output quality.
The 3-Step Email Drafting Workflow
Once your voice document is ready, the actual workflow is straightforward. This is what I use with clients who are writing weekly emails, onboarding sequences, and post-call follow-ups.
Step 1 — Brief, Don't Prompt
There's a difference between a prompt and a brief. A prompt says: "Write me a follow-up email for a sales call." A brief says: "I just had a discovery call with a 500k+ B2B services business owner. They're interested in AI workflow design but worried about team adoption. I want to follow up within 24 hours, acknowledge their concern directly, and invite them to a next call. Keep it under 150 words. Use the voice document above."
The brief does the thinking. The AI does the drafting. When you conflate those two jobs, you get generic output and you blame the tool.
The quality of your AI output is almost entirely determined by the quality of your brief. Time spent briefing is not wasted — it's the work. Most people skip it and wonder why the results are mediocre.
Step 2 — Edit for Tone, Not Content
When the draft comes back, resist the urge to rewrite everything. Read it for structure and content first. Does it say the right things in the right order? If yes, leave the structure alone. Now read it for tone. Where does it sound flat, formal, or unlike you? Change only those parts.
This is the key shift: you are editing, not writing. Your job is to restore voice to a structurally sound draft — not to start from scratch. Once you internalise this, the workflow gets genuinely fast.
Step 3 — Build a Prompt Library
Every time you write a brief that produces a good result, save it. After a few weeks you will have a library of proven briefs — for follow-up emails, welcome sequences, re-engagement campaigns, event invitations. You stop writing from scratch. You start from a brief that you know works.
This is how the workflow compounds. The first email takes 20 minutes. The tenth takes 5. The twentieth takes 2.
What AI Still Can't Do
AI is genuinely useful for email drafting. But it has real limits you should know about before you change your workflow.
It doesn't know what happened. If you had a great conversation with a prospect, the AI doesn't know the nuance of that exchange. You have to put that in the brief. If you skip it, the email will feel generic because the brief was generic.
It can't replicate humour reliably. If your brand voice has a specific kind of wit — dry, self-deprecating, slightly odd — AI will average it out into something inoffensive. You'll need to add those moments back manually.
It doesn't have judgment about timing. Sending a hard-sell email the day after a difficult call is a bad idea. AI doesn't know that. You do. The human judgment layer stays with you.
None of these are reasons not to use AI for email. They are reasons to stay in the loop rather than fully automating and walking away.
- Build your brand voice document before you touch a prompt
- Write briefs, not prompts — the brief is the real work
- Edit for tone, not content. AI handles structure; you handle voice
- Save every brief that works. That's your real asset