Today, I’m celebrating 100,000 subscribers to Marketing Ideas. 🥳
Thank you for being here. I’m blown away.
The first issue went out in August 2023, to a waitlist of about 700 people. Three years later, this is easily the best career decision I’ve ever made.
See that red circle? I’ll get back to it.
I’ve already told you how I grow this newsletter (at 18K and again at 40K). But the question I get asked even more is different: how do you write one?
So here’s my playbook for running a Substack with 100k subscribers 👇
⚙️ There is no “research” phase
People are always surprised when I tell them this:
When I sit down to write an article, the research is already done. It happened over the previous 2-3 years, one line at a time.
Every marketing idea I know is a note in Obsidian (my note-taking app). When I spot a great example → I add a line. When I think of one in the shower → I add a line. After years of that, the note is most of an article before I’ve “written” anything.
There are thousands of notes (marketing ideas in progress). So each article technically takes YEARS to write, if you count all the time I spend adding thoughts.
How a note becomes an article: I wait for a “why now”. Usually a company just did that exact marketing idea in the wild, so I can open with a hot example. Sometimes it’s a question: someone at a conference asked what I’d do starting from scratch, and that became the ZERO article. When the same question keeps coming back at the weekly live Q&A, that’s a signal too.
🧽 I stole my production model from SpongeBob
I vividly remember the day I learned that a single SpongeBob episode takes almost a year to make.
My first reaction, as a curious kid, was panic math. I had just watched a 20-episode season. Did that take them 20 years??
Of course, as an adult I get it now: they produce many episodes at the same time.
That’s exactly how this newsletter works.
It’s not like I pick one article and write it start to finish. At any given moment, I’m actively writing 10+ articles in parallel.
Most of the writing happens in the back of my mind: the structure, my main point, which examples to include, in what order, etc. Then I come back from a walk or from the shower and a write it all in one burst.
And some weeks I can’t write a damn word. That’s fine: with 10+ articles in progress, something is always close to ready.
🌀 I pay an intern $150/month to doomscroll for me
So where do all these ideas even come from?
I used to doomscroll every platform myself, hunting for cool campaigns. I’m SO thankful I don’t have to anymore. Because now:
I’m luckily positioned as “The Marketing Ideas Guy”, so people just send ideas my way.
Marketers send me campaigns their own team executed, hoping a mention will drive traffic their way (it does). When the idea is actually cool, I write about it.
I pay a $150/month intern whose job is literally doomscrolling. She uses social media like a normal person and sends me the cool campaigns she runs into.
AI helps me track what’s going viral too, but I still prefer the human taste.
The other source is my own diet: I subscribe to hundreds of newsletters, 30+ of them paid, and consume an unhealthy amount of marketing campaigns.
🛑 I was going to stop after 6 months
Six months in, barely anyone was reading this newsletter, and I felt stupid with every email I sent.
I was this close 👌 to stopping.
Around that time, I won a charity raffle Noah Kagan (AppSumo’s founder) was running. He recorded a full review of the newsletter, and at the end, he said: “Do this for 10 years - you’ll have some great success out of this”.
Woah 🤯
That was mindblowing to me: on a 10-year plan, month 6 is super early. I never seriously thought about quitting again.
(Your heroes run charity raffles like it occasionally. I met my Substack role model Lenny Rachitsky through one. Use them!)
🌍 Most of my job is generalizing
I once stood in line for 2 hours for a branded hat that says “thinking” on it. The article that came out of it became a playbook any brand can run: What to do if your marketing sucks.
That’s my actual job: when a company does something cool, the article is never just “look what they did”. The work is figuring out how YOU could steal it too, in a different industry, at a different scale.
If only the original company can run the play → I skip it.
More importantly, if I can’t deliver the promise myself, the article becomes an interview: When I knew nothing about AI SEO, I interviewed the people who cracked it - and that became the famous AI chatbots article. Same for the micro-influencer army playbook.
And if I had to sum up my writing philosophy in one word: pragmatism. Before writing, I define the promise: what will you be able to do after reading this?
I recommend giving everything you write online the same job. Pragmatism is also why I avoid ideas that boil down to “pour a million bucks on this”.
💩 “Well, this is shit”
That’s my first thought, 100% of the time, when I re-read one of my own articles a few days before it goes out. I’m not kidding.
Then comes the rewrite, my least favorite part of writing this newsletter. I delete sections, find better examples, ask marketer friends what’s missing. I’ve gotten that slap from my past self every single week, for 3 years.
If your own first drafts read bad to you, mine still do too, 100,000 subscribers in. You’re not doing anything wrong.
Once the rewrite is done, the polishing begins: I send myself the preview email over and over. Literally dozens of times, per article.
I simulate the article catching me in different moments: busy at work, waking up, actually having time to read. I open it on every device and email client I can.
The changes are usually tiny. Split a paragraph, cut 5 words, move an image up, swap an emoji, ALL CAPS a phrase. I’m obsessed with getting every word to sit exactly where it should. In the year-1 article I wrote that 90% of my writing time goes to deleting words, and that is still very true.
Done is better than perfect? Sometimes. But if it’s your baby, make it perfect.
🤖 AI never writes a word here
I’m allergic to AI-generated text, and so are you.
I do use AI daily, just never for creation. It’s my research assistant and my harshest editor. My real prompts look like: “Here’s my upcoming article. What did I miss? Is there a better example?” and “What can I remove? What’s repetitive or breaks the reading flow?”
🖼️ Art is 100% human
A single image in this newsletter takes 2-3 weeks to make.
I work with 4-5 pixel-art designers: for the simple stuff (blog covers), and for the complex illustrations and animations.
Sure, AI could generate all of it.
But I just love the art.
🧪 I spend way too long on the first words you see
The subject line is probably the most important sentence I write every week. All 100,000 of you see it in your inbox. But…only some of you open the email behind it.
So I treat those 5-10 words accordingly: some weeks the subject line gets more rewrites and work than the article under it.
People assume the writing is the hard part. Nope, it’s nailing the subject line.
One email, by effort, actually looks like this:
You can probably tell: I’m obsessed with every detail of this thing. 😅
Let me know in the comments if you want my full guide to writing email subject lines that hook people.
🌲 My articles cannot expire
The payoff of writing an article over years: it also stays good for years. I can see exactly which articles make free readers upgrade to paid, and it’s surprisingly consistent. Here’s the all-time top 10:
🌱 If I had to start from ZERO today, here’s exactly what I’d do
💰 How to send LinkedIn messages that get billion-dollar CEOs to reply
⛔️ The dark art of Reddit marketing (without getting banned)
🪖 How to recruit an army of micro-influencers (for almost $0)
New here? Start with any of these.
📦 It’s never about the newsletter
My last habit is the one I care about most: I’m always looking for ways to add value beyond the articles themselves. The goal is to make this the best marketing resource on Earth, and to make getting a paid subscription a complete no-brainer.
2,000+ paid subscribers apparently agree 🫶
This is my actual to-do list of benefits:
The censored lines are the perks that aren’t out yet (and there are a lot of them).
The next perk, probably the BIGGEST in this newsletter’s history, goes live next month 🎊
🎁 One more thing…
Writing this newsletter is the best career decision I’ve ever made. I’m my own boss here, with no briefs and no approval rounds, and my coworkers are 100,000 marketers who send me their best ideas. Two years ago I wrote that running a newsletter feels like playing an addictive video game, and it still does.
If you’ve been thinking about starting your own newsletter: start your “ideas” folder today. The written articles can come years later, like mine did.
And let me do for you what Noah did for me. I can’t review what you’re building the way he reviewed my newsletter, but the important part of his feedback fits anyone, so take it from me: Keep it up, do this for 10 years. I think you’ll have some great success out of it.
Send me the link when you’re live.
See you next week ✌️
Tom
P.S. Noah, if you’re reading this: 7 years to go ❤️
אין אין אין עליך! ❤️
what a freaking stellar post Tom.
the "make it in parallel" was such a massive learning for me a few years ago and it really works with my energy. there's such an additional pressure from trying to wrap up something to completion when you can just have a bunch of little seeds you're watering at once.