{{rh_onboarding_line}}
👋 Hey!
A corporation decides to start a newsletter roughly once a year. The idea usually survives about six weeks. Nothing kills it. No executive vetoes it, no analysis proves it wrong. It stalls on the first practical question anyone asks; sadly, that question has no clean answer: whose budget does this come out of?
Same week, Taylor Lorenz says she has lost about a quarter of her income because readers swapped her recommendations for a chatbot and a local publisher in Atlanta pushed newsletters from 2% to 28% of company revenue, mostly because one person decided the newsletter was the business.
Ready to put your name on it? Let's dive in! 👇
Meme of the Week
This Week's Hot Takes
Your company approves the newsletter twice and staffs it zero times. Acquisition sits with demand generation. Retention sits with customer marketing. One send does both jobs, so the credit splits and the cost lands nowhere.
One local publisher took newsletters from 2% to 28% of revenue. Rough Draft Atlanta's Keith Pepper rebuilt editorial, audience development, and ad sales around email instead of bolting a daily send onto the existing website.
A deliverability expert broke his own SPF record and watched opens slide. Al Iverson moved DNS in a hurry and dropped one sending IP. Opens fell from a steady 50% to roughly 40%, plus one spam complaint.
Kit will now build your landing page from a chat prompt. The new Kit MCP plugs into Claude or ChatGPT, returns a styled publish-ready page, and edits existing ones on request. Paid plans only; tweakable inside Kit before publishing.
beehiiv can email your list the second a podcast episode drops. Podcast Automations pull episode titles and descriptions into the send. Gated episodes get audio previews, and a new importer pulls your back catalog over by RSS.
Growth Hack of the Week
Create on Monday, publish all week
How: Block Monday morning and do the analysis before you write anything. Run last week's posts through an AI assistant (via MCP if possible) and ask which ones drove subscribers rather than views, what the winning hooks had in common, and which formats have quietly stopped working. Then schedule the entire week in one sitting through a scheduler, splitting output 60/40: 60% formats you already know convert, 40% experiments plus revived old winners. Spend the rest of the week only collecting. Screenshots, hooks, structures worth trying next Monday.
Why: Esteban Perez runs ADHD Weasel this way and ships roughly 110 posts a week with a two person team, feeding 4 million monthly views and 36,000 free subscribers plus 830 paid at $10 a month inside 18 months. Chenell Basilio's breakdown makes the useful point: the volume is a scheduling artifact. He also keeps links out of most posts, then drops the link in a reply 10 to 30 minutes later, once the algorithm has already decided the post is worth showing.
Expected Result: A full week of distribution that ships without daily decisions, and a Monday habit that tells you where your subscribers actually come from. Perez found Substack Notes converting to paid at a far better rate than Threads despite fewer raw views, and moved his attention the following week.
Spotlight: User Mag by Taylor Lorenz
What works:
She named the part of her product that AI ate. Most of User Mag used to be recommendations, and Lorenz says scouring the internet for the interesting thing is something a model now does better than she does. Readers who cancel tell her, in her own exit replies, that they get their information from AI instead.
She moved the value to the part that cannot be scraped. Her fix is more original reporting, on the argument that LLMs still cannot uncover what a reporter uncovers. The pitch to readers is blunt: you are paying for access to her brain, not for a link roundup.
She is loud about her AI use rather than hiding it. Roughly $300 a month on tools, disclosed on purpose. Withholding it, she argues, stops readers from evaluating what they are reading in context. She has little patience for peers who trash AI in public and ask how to use Claude in the group chat.
She followed the money on coverage. Politics was bad for advertisers, so she is shifting toward consumer tech, and she says plainly she is not the only one making that trade. Whether you like the trade or not, she made it deliberately instead of drifting into it.
Creator quote: "I've lost about a quarter of my income." She expects half by the end of the year if the trend holds.
Your takeaway: Open your last five issues and mark every section a model could have assembled from public sources. That is your churn risk, and for most of us it is the curation. Keep it, because readers still want it in one place. Stop pricing on it. Charge for the thing you had to leave the house to get.
Tool of the Week
CleverReach: plain, dependable email marketing that just sends
Why You'll Love It: CleverReach covers the whole loop with nothing you will never open: a drag and drop Editor, Segmentation, Automation, Analytics, and an HTML embed block for the bits you want to hand code.
Best For: Operators who want an ESP rather than a creator platform with a social layer bolted on, and anyone leaving an all in one who only needs lists, segments, automations, and clean reporting.
A/B Test of the Week
Test: Put your whole ad budget into one large newsletter placement, against Paved's wide-then-deep structure. Split round one across 5 to 10 newsletters, hold back a chunk of the budget, then renew only the placements that cleared a bar you set in advance.
Result: Paved published the framework its best performing advertisers use, and the mistake it names first is not overspending. It is spending so little that the result tells you nothing. Their worked example turns a $5,000 test into a defensible channel, because round one buys information and round two buys volume.
Takeaway: Decide the metric and the pass mark before you spend, not after you see the numbers. Tag every placement, treat round one as a shortlist rather than a campaign, and give the winners the held-back budget. Most operators buy a single slot, get a mushy result, and conclude that newsletter ads do not work.
Quick Links
Questions? Just write a comment below. We read all the comments and respond to them.
Thanks for reading,
Eren & Cagri