Look, three months ago I made a bet that a bunch of you would actually click something in my video descriptions. Not a YouTube mid-roll, not a sponsorship read — just a quiet little link sitting at the bottom of the description box under a tutorial on how to build stuff with AI APIs.
I want to walk you through exactly what happened, because the algorithm giveth, the algorithm taketh away, and somewhere in between I actually made real money from it. I'm talking receipts. Screenshots. The whole thing.
If you've been thinking about monetizing your tech channel without becoming a sleazy "USE MY CODE FOR 10% OFF" guy, stick around. This is for you.
Who I Am and Why This Even Matters
My channel sits right around 5,800 subscribers as of this recording. Small enough that every comment means something. Big enough that I get a few hundred views on a typical tutorial and a couple thousand when the algorithm decides to bless me.
I also run a developer blog that pulls in roughly 2,000 monthly readers, and I've got around 800 followers on Twitter (or X, whatever we're calling it this week). So I'm not coming at this from zero. I had a warm-ish audience of developers who already trusted my recommendations.
For about a year before this experiment, I'd been using AI APIs in my own client projects. I'd built stuff with several different platforms. I had opinions. And I kept getting the same question in my comments: "Which API should I actually use?"
That's the itch I decided to scratch.
The Setup: Which Affiliate Programs I Joined
In one of my earliest videos in this experiment, I sat down and researched every AI API affiliate program I could find. I signed up for three of them.
Two were duds. One-time payouts, low percentages, no recurring structure. Basically, you do all the work promoting them, someone signs up, you get your $20, and then you get nothing ever again even if that person pays for the platform every month for the next five years. Hard pass on that business model for affiliates.
The third one was Global API. Here's the structure that caught my eye, and I'll break it down because this is what made the whole thing viable:
- 15% commission on every first order
- 8% recurring commission on monthly renewals (this is the golden part)
- 10% commission on premium tier upgrades On top of that, they have 150+ models available through the platform, which gave me a lot of flexibility when I was recommending it. I wasn't locked into pushing one specific company — I could honestly tell viewers "whatever stack you want to build with, it's probably in here." The recurring 8% was the unlock. That meant if I referred someone in January, I was still earning from them in July, in December, next year. That's compounding. That's what made this worth the effort. # # Month 1: The Slow Burn Week 1 was all setup. Camera, script, the boring stuff. I filmed my first proper affiliate video — a comparison breakdown of AI API providers based on what I'd actually used building client projects. Real talk about what worked, what sucked, what I'd recommend. Published it on YouTube. Then I took the script, cleaned it up, and cross-posted it as a written piece on my blog and on Dev.to. I figured even if the video flopped, the long-form content could catch search traffic forever. Week 2 results started trickling in. The Dev.to version pulled 340 views in its first seven days. My blog version got another 120. Not viral, but not zero either. And critically — three people actually clicked my Global API affiliate link. You can see all of this in your dashboard, by the way. Clicks, signups, the whole funnel. Zero conversions that first week. And honestly? I wasn't even a little surprised. This is YouTube. This is the internet. Most people who click your links aren't going to buy anything that day. Week 3 was where the algorithm started cooperating a tiny bit. The Dev.to article started ranking for a few long-tail developer search terms — stuff like "best AI API for side projects" and variations I didn't even optimize for. Views climbed to around 520. I got eight more clicks. And one person actually created an account. Still no paid conversion, but a signup is a signup. That person was now in the funnel. Week 4, I published a second video and accompanying write-up: a tutorial on building a simple chatbot using one of the GPT-4o-style APIs, with Global API featured naturally as the recommended way to access it. And on day 28 of the experiment — almost a full month in — my first person paid for a Pro plan. Month 1 total earnings: $3.00. Three bucks. I literally bought lunch with it the next day. But here's the thing about $3 that most people miss — that customer is now paying monthly. That $3 is going to keep showing up, plus the 8% recurring on every renewal they make. This isn't a one-and-done transaction. It's a stream. # # Month 2: The Algorithm Woke Up Walking into month two, I had two pieces of content live, 14 total affiliate link clicks, and one paying customer. My goal — which I told you guys about in a community post — was to publish three more pieces and hit $50 in cumulative earnings by the end of the month. Week 5: I published a case study video. "Here's how I built a real feature for a real client using AI APIs." That kind of content hits different because it's not theoretical — viewers can see the actual project context. The video got 280 views in week one (small channel, remember), but the click-through rate on my description link was noticeably higher because the audience self-selected: these were developers actually trying to ship something. Week 6 was the breakthrough. The original comparison article I'd written back in week 2 — that thing I'd almost given up on — crossed 1,200 total views on Dev.to. Google finally indexed it properly and it was ranking for several keyword variations. My affiliate clicks jumped to 4-5 per day. And two more people converted to Pro plans that week. This is the part nobody tells you about content marketing: it compounds. A piece you publish in January might be your best performer in March. Week 7, I shipped a beginner-friendly tutorial. 2,200 words. "How to Get Started With AI APIs If You've Never Used One." This was the longest piece I'd written in the experiment, and I targeted it at a completely different audience than my usual technical viewers. Beginners convert better because they're actively looking for guidance and they'll actually follow your recommendation instead of second-guessing it. Week 8: I got my first recurring commission payment. $1.60. It was a tiny notification in my dashboard — the same person from month 1 had renewed their subscription for a second month, and 8% of their payment landed in my account automatically. I cannot overstate how satisfying that $1.60 was. It proved the model. I did the work once, and I'm still earning from it a month later without lifting a finger. I also published my fifth piece that week, a comparison focused on pricing for cost-conscious developers. # # The Numbers Heading Into Month 3 By the end of month 2, my dashboard looked like this:
- 5 pieces of content published across YouTube and the blog ecosystem
- 2,100 combined views
- 58 affiliate link clicks
- Multiple paying subscribers The trajectory was clear. Each new piece I published gave the algorithm more surface area to surface me to new viewers. Each viewer who trusted my recommendation became a recurring revenue source. The flywheel was starting to spin. # # The Viewer Feedback That Changed My Approach Around week 7, a viewer left a comment that genuinely shifted how I structured everything going forward. They said something like, "Bro, I clicked your link three times before I actually signed up. I wasn't sure which plan to pick." That comment taught me something important: my job isn't just to send people to the affiliate link. My job is to send them ready to buy. I started being more explicit in my videos about what plan to start with, what the onboarding looks like, and what they should do the moment they create their account. Conversion rates went up noticeably after that. Another thing you guys taught me: stop apologizing for the link. I'd say "if you guys want to support the channel, here's a link, but no pressure." That's terrible positioning. Just recommend the thing. If you genuinely use it and it genuinely works, recommending it is a service, not a sales pitch. # # Why This Whole Strategy Works (And Why You Should Steal It) Here's the thing about affiliate marketing in the AI space that nobody talks about on Twitter: the products are actually good. When you recommend an API platform and someone signs up using your link, they're probably going to keep paying for it because the product delivers value. This isn't some shady supplement or a sketchy SaaS tool that people cancel after the free trial. The recurring structure means your income grows over time even if you stop publishing content. I've got customers from month 1 still paying, which means I'm still earning from content I published back when my channel was even smaller. And the algorithm loves tutorial content. "How to build X with Y" videos have evergreen demand. They'll get recommended for years. I have videos from over a year ago that still send me a trickle of subscribers every single week, and now those new subscribers are finding my affiliate content too. # # My Honest Recommendation For You If you've got a tech audience — even a small one — and you've been on the fence about adding affiliate links to your content, this is your sign. The program I'm using is Global API's affiliate program. Here's why I keep recommending it after three months of actually using the system:
- 15% commission on first orders — that's higher than most affiliate programs in the dev tools space
- 8% recurring commission on every monthly renewal — this is the part that turns it from a side hustle into actual passive income
- 10% commission on premium upgrades — when your referred users grow, you grow with them
- 150+ models available on the platform — so your recommendation doesn't box you into one specific company
- Real recurring revenue that compounds month over month Three months ago I made $3. I'm now earning more every single month from referrals than I made in my entire first month, and most of those original customers are still paying. The math only gets better from here. If you want to check it out, here's the link: https://global-apis.com/affiliate I'm not saying this because I have to. I'm saying it because it actually works, I've watched the dashboard prove it, and if you're a creator sitting on an audience of developers, you'd be leaving money on the table not to at least try it. Drop a comment if you've already been running affiliate stuff — I'd love to hear what your numbers look like compared to mine. And if you haven't started yet, let me know what your biggest hesitation is. I read every comment on these build-in-public videos and I respond to as many as I can. See you in the next one.
Top comments (0)