For a long time my newsletter was a hobby that felt like a second job with no salary. Two hundred subscribers, most of them friends, and a growth chart that looked like a flat line drawn by someone falling asleep.
Then I changed how I used AI. Not to write the thing — readers can smell that instantly — but to handle the unglamorous machinery around it. Eighteen months later it's past eleven thousand subscribers, and I still write every word myself.
Here's the actual system, tool by tool, with nothing left out.
AI grows a newsletter best when it does the operations, not the writing. The voice stays human; the leverage comes from everything else.
What I delegate to AI:
What I never delegate: the core ideas and my voice.
When AI tools got good, my first instinct was the obvious one: have it write the newsletter. Faster output, more issues, more growth, right?
Wrong. Open rates dropped. Replies — the lifeblood of a small newsletter — dried up. People subscribe to a person, and AI prose, however polished, reads like a person who isn't there. It's smooth and forgettable, the inbox equivalent of elevator music.
So I made a hard rule that hasn't changed since: AI never writes a sentence my readers see. That single constraint is the reason everything else works. The personality is the product. AI just multiplies its reach. That trust-first instinct is the same one behind building an audience from zero — people subscribe to a person, not a content machine.
Here's where AI earns its keep. Every issue I write is a dense block of ideas, and I used to let it die the moment it hit inboxes. Now I mine it.
I paste the finished issue into an AI assistant and ask for repurposing — but specifically, in my voice, using my actual lines as raw material:
I edit all of it, but the blank-page tax is gone. One hour of writing now feeds a week of promotion. That repurposing engine — basic content automation, really — is what took growth from linear to steady-climbing. The Content Marketing Institute has long argued that repurposing one strong asset across formats is among the highest-leverage moves a small team can make, and this is exactly that idea in practice. It pairs naturally with the content system I run as a one-person operation.
I write once. The AI helps me show up everywhere. That's the entire growth lever.
The single highest-leverage thing I do with AI is subject lines, because nothing matters if the email isn't opened.
I don't ask AI to "write a subject line." That gives generic mush. Instead I give it the issue, my three best-performing past subject lines as style anchors, and a clear instruction: generate fifteen options across different angles — curiosity, benefit, contrarian, specific-number, story.
Then I pick. The AI is a brainstorming partner that never runs dry, not the decider. I've found subject lines I'd never have written this way, and my open rate climbed several points over a few months just from having more, better options to choose from.
The key is feeding it your winners. Generic in, generic out. Show it what your audience opens, and the suggestions get sharply better.
A newsletter that's been running a while is sitting on a goldmine it forgot it had: the archive.
I used AI to read through eighteen months of back issues and cluster them by theme. Out of that came three tight lead magnets — a short guide pulled from related issues, condensed and reorganized. I rewrote them in my voice, but the structure and the "what belongs together" work was AI's, and it would have taken me days by hand.
Those lead magnets became my best signup drivers. People give you their email for a focused, useful thing. The content already existed — AI just helped me see the shape hiding in my own back catalog.
This is the move most newsletter owners miss. You don't always need new content to grow. You need to repackage what you've already proven people want, and email automation handles the delivery once they sign up.
The least glamorous, most underrated use: list hygiene and segmentation.
A bloated list of dead addresses tanks your deliverability — send to people who never open, and inbox providers start routing you to spam for everyone. So I use AI-assisted analysis on my engagement data to:
Cleaner list, better landing, higher open rates, which feeds back into growth. It's boring and it compounds. Most of the leverage in any email marketing platform lives in this unsexy layer, not in the writing.
Let me be honest about the trade-offs, because the breathless "AI does everything" takes annoy me.
| Task | Time before | Time now |
|---|---|---|
| Writing one issue | 3 hours | 3 hours (unchanged — on purpose) |
| Promoting it everywhere | 2 hours | 20 minutes |
| Subject-line testing | guesswork | 10 minutes |
| Repackaging archive | days | an afternoon |
Notice the writing didn't speed up. That's the point. AI gave me back the hours around the writing so I could write more often and promote consistently — the two things that actually grow a list. The work that builds the relationship stayed slow and human.
If I had to point to the single highest-ROI thing AI helped me build, it wouldn't be any individual issue. It'd be the welcome sequence — the automated set of emails every new subscriber gets in their first two weeks.
Most newsletters waste the most valuable moment they'll ever have with a reader. Someone just signed up. They're more curious about you right now than they will ever be again. And the typical response is to drop them into the regular broadcast and hope they stick. Most don't.
So I built a proper onboarding sequence, and AI did most of the heavy lifting around it. I fed it my best-performing past issues and asked it to identify the five pieces a brand-new reader absolutely needs to understand who I am and why I'm worth reading. It mapped out a logical order — a story arc, really — for introducing myself across the first few emails.
Then I wrote each one in my own voice, but the structure — what goes where, what to lead with, how to pace the asks — came from that analysis. The sequence does a few things automatically: it tells my story, it delivers an immediate useful win so they trust me fast, and it gently invites a reply so the relationship becomes two-way from day one.
The numbers moved in exactly the direction you'd hope. New subscribers who go through the sequence are dramatically more likely to still be opening months later than the ones I used to drop in cold. They've already had several good experiences before the first regular issue ever lands.
The first two weeks decide whether a subscriber becomes a reader or a number. Don't leave that moment to chance.
The beautiful part is that it runs entirely on email automation. I built it once. It onboards every new person, around the clock, with no effort from me — and it's quietly responsible for a big chunk of the retention behind that growth chart.
If your newsletter feels like a second job with no salary, try handing one dreaded operational task to AI this week and protecting your voice for the writing — and follow along as I share more of the system.
Q: Won't readers be able to tell if AI is involved at all? They can tell when AI writes the content — that's the part to protect. They can't tell, and don't care, that AI helped you brainstorm subject lines or schedule social posts. Keep the voice human and the operations are fair game.
Q: Which AI tool should I use? Almost any capable AI assistant works for this — the technique matters far more than the brand. What changes results is feeding it your best past work as examples, not which model you pick.
Q: Isn't repurposing just spamming the same thing everywhere? Only if you copy-paste. Done right, each platform gets a genuinely native version of one idea. People rarely see all of them, and the few who do mostly appreciate the reminder.
Q: How much of my growth was AI versus just consistency? Honestly, consistency is the engine — AI is what made consistency possible by removing the operational drag that used to make me quit. They're not separable. The tooling is what let the discipline survive.
AI didn't write my way to eleven thousand subscribers. It cleared the runway so my writing could actually take off.
Protect the voice. Automate the machinery. Let the two compound.
Don't ask AI to be you. Ask it to handle everything that isn't you.
What's the one operational task around your newsletter that you secretly dread? Start there — that's the hour AI should be buying back first.
One idea a week to a published issue in under an hour. The boring system behind a newsletter I never dread sending.
No following, no network, no luck. Just an unglamorous system I ran for eighteen months. Here's exactly what I did.
One person, output that looks like five. It isn't about working more hours — it's about a kind of leverage teams rarely have.
Comments
Sign in to join the conversation
No comments yet. Be the first to share your thoughts!