I'm building a local newsletter. And I'm doing it in public
It's been on my mind for too long not to give it a shot
And of course I'm doing a newsletter about the newsletter. So meta!
The messy in-between. The first subscribers. The failed tests.
PUMPED FOR THIS
I've doubled my ad spend on weekends for my local newsletter. I decided to remove it. Here's the data:
Previous weekends: 134, 66 and 65 new subs
Weekends with double budgets: 54 and 42 new subs
The highest ROI is definitely testing out new ads
Ad fatigue is real!
It feels like a bad habit, but I check my subs count on my local newsletter almost every day.
I've noticed my sub acquisition goes up on weekend, especially Sundays.
Doubling my ad spend budget on Sundays for the next month.
Let's see if acquisition doubles too ๐
It feels like a bad habit, but I check my subs count on my local newsletter almost every day.
I've noticed my sub acquisition goes up on weekend, especially Sundays.
Doubling my ad spend budget on Sundays for the next month.
Let's see if acquisition doubles too ๐
March was a good month for my local newsletter
Doubled my subs by just testing different ad formats
The "Apple note" format ad has been killing it with $0.14 CPL so far
I just released a free 3 hour course on Claude Code for complete beginners.
I assume zero experience with Claude Code, and take you through the following topics:
> Skills - including building a Gmail email triage skill live
> MCPs - live exercise connecting to Notion, parallelizing research with a dozen subagents, and writing async to a Notion db
> Plugins and subagents (with @every's Compound Engineering)
> git and Github
> Web app development with authentication (login): including a live exercise building a job application tracker
> and much more!
The course is here on YouTube: https://t.co/IkRY4hNAiN
On Thursday at 8 PM EST, I'll have a free group call inside my Skool server (https://t.co/M4e29ID8gI). You can ask me any questions you have about Claude Code/Codex/agentic coding, whether for economics research or otherwise, and I'll stay on until all questions are answered.
You can also ask me questions async by posting in the Skool server, and I'll record the call and distribute it on YouTube in the next few weeks.
Additionally, inside the Skool server in the classroom section, you'll find a slide deck with 51 slides corresponding to the material in the course on YouTube.
If you're not testing out multiple Meta ads you're losing
Reduced my paid sub acquisition to 10 cents from 60. I had an evergreen ad running, less than 20 cents/sub initially but it increased with time. Ad fatigue is real
Not too bad!
Local newsletter folks
How transparent are you with sponsorship rates?
Do you share every type of offer pricing ahead, or you work on a budget basis and make a proposal?
Launched a Dinner Club this week with a simple @youformdotcom in my local newsletter now at 1.2k people
Got fully booked in less than 24 hours.
I wanted to get a bit more fancy, so I shipped a custom page ridiculously copying @MikeyPesto style.
Still not perfect, but ok!
Sunday 8AM. Sharp.
Got off my bed. Got into my running shoes, and joined a local run in Ancona.
A local newsletter is not just writing. It's bounding. It's community.
And I would not ever absolutely ran 10km by myself!
photo of Francesco Flores
Gathering feedbacks about organizing a dinner club for my local newsletter
18/20 people would like to join in. Also received a bunch of positive emails.
The signal is good
But man, Italy taxation makes it very expensive for me to sell seats. I may need to find a sponsor
Don't beat yourself too much over perfection.
@github sent me this product update. There's a clear undesired repetition. Probably it could have been avoided with an extra pair of eyes
but at the end of the day...who cares?
What's important is to SHIP