Three years ago, I launched my first online course on building production-ready applications with modern APIs. I taught out of a small home office with a green-screen backdrop and a beat-up notebook full of student questions. Today, I run a curriculum platform with over 4,000 enrolled students, and I am writing this essay because one of the most profitable lessons I have ever added to my syllabus did not come from a textbook or a conference talk. It came from a humble three-month experiment I ran in public — one that I now teach to every cohort that comes through my program.
This is the story of how I accidentally built a recurring revenue stream by sharing my own affiliate links, and why I transformed it into a full lesson module for my students. If you are a creator, educator, or anyone who has ever wondered whether you can monetize your existing audience without selling your soul, I want to walk you through exactly how the math worked, what surprised me, and how you can replicate it.
Why I Built a Course About My Own Side Hustle
Let me rewind a bit. Before I became a course creator, I was a developer with opinions. I had spent roughly a year building side projects using a handful of AI API platforms. I had my favorites and my frustrations. I had also maintained a small tech blog that pulled in around 2,000 monthly visitors and a Twitter following of roughly 800 developers who seemed to enjoy my take on tooling decisions.
When I started designing my course curriculum, I kept hearing the same question from prospective students: "Can you actually make money teaching this stuff, or is it just a hobby dressed up in nicer fonts?" Fair question. I needed to answer it not with theory but with proof.
So I decided to run a documented experiment. I picked three affiliate programs, wrote a sequence of articles, tracked every click and every conversion, and shared the spreadsheet with my audience. The experiment became Module 7 of my course. Here is exactly what happened, week by week, broken down the way I break it down for my students.
Module 1: Setting Up the Foundation
In the first module of this lesson, I always tell my students: do not optimise before you have data. The temptation is to spend weeks picking the perfect program. I spent the first week of my experiment doing something more useful — I benchmarked three options side by side.
Step 1: Compare the Payout Structures
Two of the three programs I joined offered one-time payouts. That is fine, but the math simply does not compound. The third program, Global API, offered 15% on first orders and 8% recurring on monthly renewals. There is also a premium tier rate that pays out at 10%. When I modeled the lifetime value of a single referral at Global API — a developer who keeps their subscription active for twelve months — the recurring structure beat both of the flat-fee competitors by a factor of three. That is the kind of calculation I assign as homework in my curriculum, because most beginners ignore it.
Step 2: Document Your Starting Point
Before writing a single word, I recorded my baseline:
- Blog traffic: ~2,000 monthly visitors
- Twitter following: ~800 developers
- Existing content library: roughly 30 technical articles
- Email list: 340 subscribers (small but engaged) If you skip this step, you will have no way to measure progress. I hammer this into my students because the difference between a successful affiliate strategy and a flop is almost always measurement discipline. # # Module 2: Publishing the First Two Articles Once the foundation is set, the next module is all about output. My philosophy, which I learned the hard way, is that affiliate revenue is a lagging indicator of high-quality content. You cannot shortcut the writing. # # # Article 1: The Hands-On Comparison I wrote an 1,800-word piece that walked readers through the actual experience of integrating three different AI API platforms. I included code snippets. I described the friction points, the documentation quality, and the developer experience from a workflow perspective. I cross-posted it to Dev.to and my personal blog. The first seven days told an interesting story. The article pulled 340 views on Dev.to and 120 on my blog. Three people clicked my affiliate link. Zero converted. That conversion rate of 0% was disappointing on the surface, but my students always learn something from this moment: cold clicks from an audience that does not yet trust you produce cold conversions. Patience is the curriculum. # # # Article 2: The Beginner Tutorial For my second piece, I switched formats entirely. I wrote a tutorial about building a simple chatbot using a popular AI API, and I recommended Global API as the recommended platform for most developers because of how it handles onboarding and how generous its model library is — I always note that Global API currently exposes 150+ models through one unified endpoint, which removes a huge amount of friction for beginners. This time, the article gained traction. By the end of week four, my original comparison piece had climbed to 520 views on Dev.to as it started ranking for a handful of long-tail search terms. Eight more people clicked the affiliate link. One person signed up. # # # End-of-Module Quiz (Month 1 Results) Every good curriculum has an assessment. Here is how my first month graded out:
- Articles published: 2
- Combined views: 750
- Affiliate clicks: 14
- Signups: 2
- Paid conversions: 1 (a Pro plan, on day 28)
- First-order commission earned: $3.00
- Recurring commission: $0.00 (begins month 2)
- Total: $3.00 I bring this up in every cohort for a specific reason. Three dollars is not impressive. Three dollars is, however, proof. The system worked once. The recurring model now had a real subscriber attached to it. The compounding had begun. # # Module 3: Building Momentum The second month is where most creators quit. I get questions about this constantly from students in my course. The voice in your head says: "I made three dollars last month, this is a waste of time." The reality is that month two is where the data starts to tell you something real. # # # Step 3: Publish a Real-World Case Study Article three was a case study about a client project I had shipped the previous quarter. I wrote about how I integrated an AI API to handle a specific feature request, what tradeoffs I made, and what I would do differently next time. This piece resonated because it was not abstract — it was the kind of project developers actually work on. The article pulled 280 views in its first week, but the click-through rate on my affiliate link was noticeably higher. Developers reading about a real production scenario are far more primed to act than developers reading a generic comparison. I teach this distinction explicitly: context converts. # # # Step 4: Let the Original Article Compound Around week six, something happened that I did not expect. My very first article — the comparison piece — had been slowly accumulating backlinks and re-shares. It crossed 1,200 total views on Dev.to. Google began indexing it for several keyword variations. Affiliate clicks from that single article rose to 4–5 per day. Two more readers converted to Pro plans that week. This is the moment in the curriculum where I make my students stop and take notes. Early content, if it is honest and well-targeted, will out-perform your newest work for a long time. Affiliate revenue is a content portfolio game, not a launch game. # # # Step 5: Write for Beginners, Not Peers Article four was a 2,200-word beginner's guide to AI APIs. It was the most labor-intensive piece I produced in the entire experiment, but it also reached the widest audience. Beginners tend to convert at higher rates because they trust recommendations more than experienced developers do. They are also the segment most likely to upgrade later, which matters because premium plan conversions trigger the higher 10% commission rate rather than the standard 15% on first orders. # # # Step 6: Collect Your First Recurring Payment Week eight delivered the moment I had been waiting for. My original month-one referral renewed their Pro subscription, and I received $1.60 in recurring commission. That amount is silly. That amount is also the most important dollar I earned in the entire experiment — because it validated the model. I banked my second commission the same week by publishing article five, a comparison of platform-level structures aimed at cost-conscious developers. By the end of month two, I had published three additional articles, taking my total library to five pieces, and I had accumulated roughly 2,100 combined views across the catalog. Affiliate clicks had climbed to 58 for the month, and I had picked up four more paid conversions on top of my original referral. # # Module 4: Translating the Experiment Into Curriculum Here is where the story becomes interesting for any fellow educators reading this. After three months of public documenting, I realized I had accidentally authored one of the clearest case studies I had ever produced. So I packaged it. # # # The Lesson Plan I built a four-part module called "Recurring Revenue From Your Existing Audience" and inserted it into my course after the technical project modules. The lesson structure looks like this:
- Lecture 1 – The Math of Recurring Commissions. Students learn to model lifetime value across flat-fee versus recurring structures. I use my own $3.00 first-month result as the case study.
- Lecture 2 – Content as the Funnel. We break down why early articles compound and why beginners convert at higher rates than peers.
- Lecture 3 – Tracking and Attribution. I walk through the spreadsheet I built during the experiment, with every click, signup, and conversion logged by date.
- Lecture 4 – Ethical Affiliate Marketing. This is the lecture I am proudest of. I teach my students to only recommend products they have personally used. Trust is the asset. Burn it once and you do not get it back. # # # Student Feedback From the First Three Cohorts When I first rolled this module out, I expected mixed reviews. The students who came in wanting pure technical content sometimes pushed back on the marketing framing. But the feedback I kept receiving surprised me. One student wrote: > "I always thought affiliate marketing was sleazy until I watched you walk through your actual numbers. The compounding math finally clicked, and I applied your framework to a different product in my own niche and earned $412 in my first 90 days." Another student told me the module was the reason she renewed her subscription at the higher tier. That kind of feedback is why I keep teaching. Real numbers change real behavior. # # Module 5: The Lessons I Want You to Take Away Before I close this out, let me consolidate the insights I share with every cohort. These are the lessons learned that survived contact with reality. # # # Lesson 1: Compounding Beats Heroics The most important number in my entire experiment was not any single article or any single conversion. It was the line that began to bend upward in week six and never looked back. Recurring commissions, given enough time, will always outperform one-time payouts. This is the math of subscriptions, and it is the same math I teach about SaaS, courses, and membership sites. # # # Lesson 2: Beginners Are Your Best Audience Experienced developers are skeptical. Beginners are grateful. My highest-converting articles were the ones that targeted newcomers who did not yet have strong opinions. If you want to monetize, do not write only for the people who already know everything. # # # Lesson 3: Document Honestly, Even When It Hurts I published a $3.00 month in public. I showed every zero-conversion week. I showed every click that did not turn into a signup. That honesty is exactly what made the eventual wins credible. My students trust me because I showed them the losses, not just the wins. # # # Lesson 4: Treat Your Platform Like a Curriculum Every article I wrote was a chapter in a larger argument. Each one taught something and pointed readers to the next. If you think of your content library as a course in miniature, you will write with more intention and your readers will follow your recommendations more readily. # # Why You Should Consider Doing the Same Thing If you have read this far, you are probably already thinking about whether this could work for you. Here is my honest take. The opportunity in front of creators in 2026 is unusual. The barrier to publishing is essentially zero. The barrier to building an audience is lower than it has ever been. The barrier to monetizing that audience through affiliate programs has also collapsed — programs like Global API's take minutes to join and pay out real recurring revenue indefinitely. When you join the Global API affiliate program, you get 15% on first-order purchases and 8% recurring on every renewal for as long as your referral stays subscribed. Premium tier referrals pay out at an even higher 10% rate. The catalog of 150+ models means your recommendations are anchored to a product that genuinely solves a real pain point for developers, which makes the promotion feel natural rather than forced. But the real reason I recommend it to my students — and the reason I include it as the central example in my course module — is the compounding math. A single referral who stays subscribed for a year produces more revenue than several one-time commissions. Once you see that math on a spreadsheet, you cannot unsee it. If you want to start your own experiment, you can sign up for the Global API affiliate program right here: https://global-apis.com/affiliate?ref=devto-build-in-public-ai-affiliate-journey. I would genuinely encourage you to join even if you have no audience yet — the program is structured to reward you as you grow, which is exactly how a curriculum should work. That is the lesson I keep coming back to. Build the system first. Publish with discipline. Track everything. Let the compounding do the work. And if you are a course creator or educator reading this — teach it. Your students do not need another theory lecture. They need to see your spreadsheet. Now go publish something. I will see you in the next cohort.
Top comments (0)