He Tried Print-on-Demand Shirts, a Blog, 2,000 YouTube Videos — Then Hit $35K/Month in Year Four

A few days ago I was digging through my saved articles and came across one about compounding. One line stopped me: too many people pour their time and energy into things that never grow exponentially.

My first thought went to the system I maintain every day for auto-writing my WeChat account. But instead of just sitting with that thought, I went looking for a harder, real example — a guy named Jack Friks, who spent four real years learning that same lesson the hard way.

It took Jack four years to get his revenue to roughly $35,000 a month (about 50,000 CAD). Those four years weren't idle. He designed print-on-demand T-shirts, ran an affiliate marketing blog, put out more than 2,000 YouTube videos, and traded crypto.

Each one sounded like a legitimate side hustle, and he genuinely worked at all of them. But four years in, there was nothing left in the account. He later said he cried on a walk with his fiancée during that stretch — still grinding to hit $3,000 a month.

What actually turned things around was a small tool he was forced to build for himself. He had to post the same content to Twitter, Instagram, LinkedIn, and TikTok every day, and manually copy-pasting across platforms was breaking him. So he wrote a scheduling tool for his own use, shared it online, turned it into a product within a month, and now that tool alone brings in $35,000 a month.

2,000 videos didn't turn his life around. A small tool that fixed his own annoyance did. The difference isn't as simple as "content doesn't work, products do."

When I put the 2,000 videos next to the scheduling tool, the dividing line turns out to be something every programmer already knows: is this a one-off script, or a service that keeps running on its own.

Whether a video gets seen depends on whether the platform's algorithm decides to push it that day. Want traffic tomorrow? Shoot another video, roll the algorithm dice again — no different from re-running a script by hand each time; if you don't run it, nothing happens. Sell one T-shirt, get paid for that one shirt. Affiliate commissions only show up when traffic shows up, and drop to zero the moment traffic stops. All of it requires you to personally show up again before you get paid again.

The scheduling tool is different. He built it to solve his own problem first. Every other person who subscribes keeps paying as long as they're still juggling multiple platforms — the annoyance never goes away, so he keeps getting paid without having to talk any single customer into renewing every single month. That's a standing service: build it once, ship it, and after that it keeps running without him showing up each time.

This isn't really "content doesn't work, products do." A tutorial that stays ranked near the top of search results is a standing asset too. Conversely, a subscription service that has to re-serve some fresh, ad-hoc request from the client every month is still a script somebody has to babysit. What actually decides whether something compounds is whether the first delivery leaves something behind that keeps working for you.

What I'm maintaining right now works the same way: every day, pulling material from my knowledge base, drafting, making images, pushing a draft — all of it real time spent. I'm not doing this for today's read count. I'm betting on whether this pipeline can eventually keep turning without me having to think up a new topic from scratch every single time.

Whether it's actually working yet, I honestly can't say — whether a WeChat post gets pushed into the recommendation feed is itself a variable nobody fully controls, building the system doesn't guarantee it. But at least this time I know which bet I'm actually making: not whether today's post does well, but whether this thing is a standing service instead of a one-off script.

Next time someone tells you they're grinding away at a side hustle, you can ask one quiet question: once you finish this thing, do you have to show up and start over next time? The answer tells you whether they're building a system that runs on its own, or manually re-running a script that never finishes.