How I picked my niche after 13 years of going wide.
I ran a podcast with 75 episodes and 141 five-star reviews and still quit. Here's the one human I'm writing for now, and how AI helped me find him.
I launched a podcast in 2013. I hit 75 episodes and 141 five-star reviews. I still quit.
The Marketing Access Pass podcast. 75 episodes. 141 five-star reviews. I still quit.
The podcast that was working
It was called Marketing Access Pass. MAP for short. The idea was a roadmap to online success for people trying to build a business on the internet.
The mic was one of the best on the market at the time. I spent real money on the setup. I sat in the living room of my condo in California and put in the hours.
[IMAGE 1: Photo of me at the desk in my California living room in an Air Force hat, grey hoodie, drinking from a Marketing Access Pass mug, audio mixer on the desk and content whiteboard on the wall behind. Filename: map-living-room-setup.jpg. Alt text: “Recording Marketing Access Pass from my living room in California, around 2013-2015”]
The numbers grew. The list grew. People wrote back saying they liked it. A few even walked up to me at conferences and recognized my voice. At one point the show was sitting on iTunes next to Pat Flynn, Lewis Howes, John Lee Dumas, and Michael Stelzner. Their shows had massive audiences. Mine was in the same grid. Company I did not deserve yet.
[IMAGE 3: Screenshot of a podcast app grid showing Marketing Access Pass listed alongside School of Greatness, Smart Passive Income, Social Media Marketing, Entrepreneur on Fire, and This Old Marketing. Filename: map-itunes-grid.jpg. Alt text: “Marketing Access Pass in the iTunes grid next to Pat Flynn, Lewis Howes, and John Lee Dumas”]
It wasn’t that no one was listening. It was that I couldn’t tell anyone who I was listening for.
Why “entrepreneur” isn’t a person
My website was a pile of everything. Podcast episodes. Video tutorials. WordPress how-tos. SEO posts. Social media tips. The whole digital marketing umbrella.
At every conference, at every meetup, someone would ask the same question. “So what’s your niche? Who do you actually help?”
I didn’t have a clean answer. I was a podcaster and a web designer and an SEO guy and a WordPress guy and a video guy. I was chasing every revenue line at once. Recording, interviewing, editing, publishing, sharing. All of it me. All of it for “entrepreneurs,” which turns out isn’t a person.
[IMAGE 2: Photo of the Marketing Access Pass podcast card propped in front of the editing monitor with Adobe Audition open, post-it notes covering the desk with topic ideas. Filename: map-editing-audition.jpg. Alt text: “Editing an episode of Marketing Access Pass in Adobe Audition, with content ideas on post-its”]
“Entrepreneur” is a census category. A 24 year old dropshipper and a 58 year old plumbing company owner both check the box. They don’t listen to the same shows. They don’t have the same worries. They don’t care about the same things. Writing to both at once means writing to neither.
I knew this on paper. Every marketing book I’d read said the same thing. Pick the specific person. Write to one human. I’d told clients the same thing a hundred times.
I still ignored the rule on my own project. Mostly because picking one person felt like giving up on everyone else. I burned out before I admitted it. Not because the content failed. Because I was doing the work of five people for an audience I couldn’t describe in one sentence.
My podcast sitting next to Pat Flynn, Gary Vee, and Smart Passive Income. Going wide.
How to know your niche is specific enough
When I started over this year, I did one thing different. Before I wrote a word, I picked one human.
A career marketer, 35 to 45. Probably male, probably a dad. Sole breadwinner. Good at his job for fifteen years. Has the kid, the mortgage, the gym membership he uses twice a month. Two or three hours a day to spare. Usually at 5 AM before the house wakes up, or 10 PM after it goes to sleep.
He’s tired. Not the cartoon burnout kind. The quiet kind. Where the work still gets done but nothing feels like his anymore. He’s watched AI chew through things he used to charge for. He’s scared about what the next five years look like. He doesn’t want to quit his job tomorrow. He wants to build something that’s his on the side, before he has to.
That’s who I’m writing for now. Not because he’s the biggest market. Because he’s me five years ago. And I know exactly what I needed to hear then and didn’t have.
I used Claude to pressure test the niche before I committed. Fed it the rough idea and asked it to poke holes. Asked it to map the audience’s daily worries. Asked it to find where this person was already gathering online, which podcasts he listened to, which newsletters he opened.
The tool didn’t pick the niche for me. It made me defend mine out loud until the weak spots fell off. Every time Claude asked “why this person and not a broader group,” I had to answer. Half my answers didn’t survive.
By the end I could describe my reader in one paragraph without stumbling. I could name three problems he had that week. I could name the podcast he probably listened to on his drive.
Here’s the test. Write down the human. One person. Describe the part of their day where your content would help. If you can’t do that in one paragraph, the niche is still too wide.
The 2013 me could not have passed that test. The 2026 me barely can. That’s the bar.
The part I’m still getting wrong
I’m not claiming I’ve got it right now.
But “career marketers over 40 who want to use AI to build something of their own” is specific enough to write a post to. More than I could ever say for “entrepreneurs.”
If you want the bigger picture of what I’m doing here, the manifesto post is where this series starts. Read the manifesto.
There’s also a newsletter. If you want updates when new posts drop, the form is at the top of any page. If not, come back when you feel like it.
Come build with me,
Anthony
The build log.
New post drops, tool tests, and the occasional honest look at what isn't working. One email at a time. Unsubscribe in one click.