Me, ChatGPT, and the Struggle to Grow a Newsletter

One writer, <40 subs, and ongoing AI-aided trial and error

Days after receiving the rejection that showed me how much ignoring social media had screwed me over, I complained to--who else? ChatGPT! I wrote this message (abridged), with some links just for context:

Marketing is everything. Absolutely everything.

I want to hate social media, I want my writing to stand on its own. But writing doesn’t stand on its own. It never has and it never will.

ChatGPT's response (abridged):

Marketing is everything. A mediocre idea with perfect marketing can go viral. A brilliant book with no visibility gets forgotten.

So the question becomes: what do you do with that knowledge? Do you:

1. Refuse the game, retreat, and preserve your artistic purity at the cost of being unread? [Which I already tried]

2. Learn just enough of the rules to manipulate them and use them as scaffolding to lift your work?

3. Go full tilt into the machine, risk burnout, and build a personal brand as big as your book?

I said:

Suppose I wanted to build a social media presence. Maybe I do want option 3. Suppose I only have two rules:

1. I won’t pay money

2. I won’t exploit my friends.

Based on where I am now, what would I do as a first step?

ChatGPT said:

If you're starting from zero with those two (very fair) rules, then your first step is simple:

Claim your space and declare your identity.

This means:

1. Pick your platform.

   • Most viable options:

     • Substack – ideal for longform, thoughtful writing (and has built-in growth tools like recommendations and cross-promotion)...

And there you have it.

The reason I'm here, the reason I'm even writing this today, is because of ChatGPT.

I didn't let ole "Chatty Geep" off the hook at that point though. I kept asking questions. I ask questions about my Substack posts. About their titles (including this one). I asked ChatGPT to write my welcome letter, my about page. And yes, I edited the suggestions. I didn't just thoughtlessly paste everything.

But I can't stress how much ChatGPT has helped me with this blog so far. The first title I had for it was "The Great State of Grapes," named after a cute thing my daughter said once. But once I started thinking about how every element affects how many subscribers you get, I realized I had to be more strategic. I asked for five titles based on my first few posts and this is what the language model offered:

1. "Failing Upwards"

2. "Here Goes Nothing"

3. "I Don’t Even Live Here"

4. "No One Asked Me"

5. "Middle of Something"

I went with #4, if you hadn't noticed.

I've even asked questions about the best time to publish posts and notes. After all, I'm a busy guy. I have a job. I'm trying to write a novel. I have a wife and kids. I can't just spend all day writing notes and posts and replying to others. I have to be strategic.

So I asked ChatGPT. And I demanded sources, and got this article from Brevo.

So now I'm posting at around 9:00 AM on Tuesdays and Thursdays (Brevo says 10 AM, but the graphs show there’s not much of a difference between 9 and 10). Yes, some readers are probably getting daily or weekly digests and not direct emails from every subscription, and many people will turn the app’s notifications off. But it still makes a difference.

Obviously ChatGPT has problems, and many people have pointed them out (including me). But one thing ChatGPT has done is read a LOT of blogs and processed a LOT of social media. So who better to advise you on how to improve your presence on a social media platform, whether it's Medium, Substack, or whatever comes next?

Time will tell if any of this works. At the time of this writing I have less than forty subscribers. But I would like to think that if OpenAI and the rest of the AI overlords are going to take over the whole world, we can at least benefit a little bit along the way.