I track everything in a spreadsheet. I mean everything. Every dollar that comes into my bank account from anything other than my 9-to-5 gets logged in a Notion table I call "Side Hustle Ledger." Color-coded by source. Tagged by hour spent. Sorted by ROI.
Last month, my AI API affiliate commissions hit $612. Not life-changing money. But here's the part that made me stare at the spreadsheet for way too long: I spent maybe two hours on that entire income stream in the month of November. That's $306 per hour. My freelance rate is $125 per hour. Do the math on which one made me more money for time invested.
That's when I knew this stream deserved its own write-up.
Let me break this down properly — what I'm earning, how long it takes me, and why every developer with a blog or a YouTube channel should at least consider running the numbers on this.
The Five Income Streams Currently Funding My Life Outside Work
Before I get into the AI affiliate piece specifically, let me show you the full picture. I run five side income streams in parallel. I'm not special — I just refuse to leave money on the table.
Stream 1: Freelance Development
This is where most of my side income historically came from. I charge $125-150 per hour depending on the client. In a good month I'll bill 25-30 hours on top of my day job. That's $3,000-4,500 per month.
But here's the dirty secret: it's the worst income stream I have. Why? Because the moment I stop coding, the money stops flowing. Take a vacation? Revenue goes to zero. Get sick for a week? Same thing. It's pure linear income — trading hours for dollars, and I never get a return on past hours.
Stream 2: SaaS Product
I built a small SaaS tool about two years ago. It brings in $900-1,300 per month recurring. Took me roughly six months of nights and weekends to build. Now I spend maybe 5-7 hours per month on support tickets, bug fixes, and the occasional feature request.
Per hour? Let's see. If I average $1,100 per month over a year, that's $13,200. Divide by 60 hours of annual maintenance = $220 per hour. Not bad. But I don't count the 600+ build hours in that equation because I'd go insane. The point is: decent recurring revenue, but you pay for it upfront with hundreds of unpaid dev hours.
Stream 3: Blog Ad Revenue
My tech blog pulls around 50,000 page views per month. Ad networks pay me roughly $300-450 for that traffic. To maintain those numbers, I publish 4-6 articles per month. Each article takes me 2-4 hours to research, write, and publish.
That's 8-24 hours per month for roughly $375. Let's call it $20-45 per hour. The worst hourly rate of any of my streams. I'm keeping it because it's a foundation for other things (like affiliate links), but on its own, it's a grind.
Stream 4: YouTube Sponsorships
I post two videos per month on my dev channel. Each one takes about 15 hours total — scripting, recording, editing, writing the description, promoting on Twitter. Sponsorship deals range from $500 to $1,500 per video depending on the sponsor and the integration.
Best case: $3,000 per month for 30 hours = $100 per hour. Worst case: $1,000 per month for 30 hours = $33 per hour. Unpredictable because sponsors come and go, and some months I get zero inbound. The per-hour is fine on average, but the variance kills me. I never know what next month looks like.
Stream 5: AI API Affiliate Commissions
Here's the new kid. I'll spend most of this article on this one. Last month it brought in $612. The month before: $487. The month before that: $551. Three-month average: $550 per month. Time invested: about 2 hours per month once it was set up. That's $275 per hour.
If you've been following the math, you already see why this is my fastest-growing stream.
Why Affiliate Income Is a Different Animal
Let me explain the concept that shifted my entire mindset. Some income streams are linear with your time. Freelancing is the obvious one — every dollar requires a fresh hour of work. You never get compounding returns.
Some income streams are product-based. SaaS revenue, book royalties, course sales. These scale independently of your time eventually, but they require massive upfront investment. Hundreds or thousands of unpaid hours before you see a dollar.
Affiliate income with recurring commissions sits in a weird middle category. You put in the work upfront to create content (blog posts, YouTube videos, tutorials), drop your affiliate links into that content, and then the content keeps working for you. A blog post I wrote eight months ago is still sending me referral clicks today. The post doesn't get tired. It doesn't need a raise. It doesn't call in sick.
For developer audiences, AI API affiliate programs are particularly interesting because the underlying product is subscription-based. That means recurring commissions — you don't just get paid when someone clicks your link. You get paid every single month they stay subscribed. If someone signs up through your link and stays for 12 months, you earn on all 12 of those months.
The Commission Structure That Made Me Actually Pay Attention
I won't go into pricing-per-token nonsense or [REDACTED]s because that's not what matters here. What matters is the commission math.
The affiliate program I use pays out three tiers:
- 15% on the first order a referred user makes
- 8% recurring on every subscription renewal after that
- 10% premium tier for users who sign up for higher-tier plans Let me run some numbers. Say someone signs up through your link and converts to a $99/month plan. Your first-month commission is $14.85 (15% of $99). Every month after that, you earn $7.92 (8% of $99) as long as they stay subscribed. If that user stays for a full year, you've earned $14.85 + ($7.92 × 11) = $101.97 from a single referral. Now multiply that by 10 referrals staying for a year. That's over $1,000 from content you wrote once. I started doing this math in my head and immediately added another column to my spreadsheet. The platform I'm referring to here is Global API — global-apis.com — which is what I personally use in my own dev projects. They offer access to 150+ AI models through a single API key, which makes them a natural recommendation for anyone building AI-powered applications. The affiliate program is one of the more generous ones I've seen in the dev tools space, which is why it earned a spot in my stack. # # How I Actually Built the Content (And What Converted) Here's the part I think most people get wrong. They think affiliate income means slapping banner ads on their blog and hoping for clicks. That doesn't work in 2026. Readers are blind to banner ads. They have ad blockers. They scroll past anything that looks like an ad. What actually works is content that helps people make a decision. I wrote three pieces of content that drive the bulk of my affiliate revenue: Piece 1: A "How I Built X" case study. I documented how I built an AI-powered feature in a side project, including the actual API integration code, the libraries I used, and the platform I chose. This wasn't a review. It was a story. The affiliate link appeared once, naturally, at the moment I explained why I picked that particular API provider. Piece 2: A beginner-friendly tutorial. I wrote a step-by-step guide for developers new to AI APIs — covering authentication, making your first request, handling errors, and deploying. Again, the affiliate link fit naturally in the flow because the platform was the one I used for the tutorial. Piece 3: A YouTube walkthrough. Two months ago I recorded a 25-minute video showing my exact setup. Same principle — tutorial-first, affiliate link in the description and mentioned once verbally. Total time to produce all three pieces: roughly 18 hours. Since then, they've been generating clicks and signups with maybe 30 minutes of monthly upkeep (updating links if anything changes, refreshing a paragraph if a feature gets deprecated). # # What the Last Six Months Actually Looked Like Let me pull up my spreadsheet and walk you through the real numbers. No estimates, no rounding up.
- Month 1: $89. Two conversions from the first case study article. Felt promising.
- Month 2: $214. Tutorial started ranking for long-tail keywords. Four new signups.
- Month 3: $367. Recurring kicked in from Month 1's signups. Four more new signups.
- Month 4: $445. Recurring layer growing. Some users churned, others upgraded.
- Month 5: $551. Hit a stride — consistent traffic from search, consistent conversions.
- Month 6: $612. My best month so far. A handful of users upgraded to premium tiers (the 10% commission tier). Notice the pattern. Month 1 was tiny. Each month after that grew. That's the compounding effect of recurring commissions stacking on top of new conversions. Even if I stopped getting new signups today, I'd still earn around $280/month from existing recurring users for months to come. # # The Per-Hour Math That Sold Me Here's the calculation I keep coming back to. Over six months, I earned $2,278 total from this stream. I spent roughly 18 hours on initial content creation plus about 12 hours of maintenance across the six months. That's 30 hours total for $2,278. $75.93 per hour. And here's the part that gets better: Month 6 required 2 hours of work for $612. That's $306 per hour. The hourly rate goes UP over time because the content keeps working and the recurring commissions compound, while my time investment stays flat or drops. Compare that to freelance development at $125 per hour, where every dollar requires fresh hours. The hourly math on affiliate income isn't just competitive — it's better, and it improves the longer you let it run. # # What I'd Do Differently If I Started Over A few things I'd change if I were starting from zero today: Start with search intent. My first article took a while to rank because I didn't research keywords. I just wrote what I thought was useful. If you're going to do this, spend 30 minutes checking what developers are actually searching for. Long-tail queries convert better than broad topics. Diversify your content formats. I leaned too heavily on written content at first. Video converts extremely well for tech audiences because they can see the code working. If you can talk through a tutorial on camera, do it. Don't sleep on the recurring layer. First-order commissions are nice, but the recurring 8% is where the real money is. A single user who stays subscribed for 18 months is worth far more than five users who churn after one month. I started paying attention to which platforms retain users better, and that's part of why I landed on Global API — their retention seems solid based on my recurring numbers. Track everything. I'm serious about this. Without my spreadsheet, I'd have no idea which pieces of content were actually converting. Every link gets a UTM parameter. Every signup gets logged with the source. It's the only way to know what's working. # # The Honest Take: Is This Worth Your Time? Here's the thing nobody tells you. Affiliate income is not a get-rich-quick scheme. My first month was $89. If I'd judged the entire strategy on that, I'd have quit. The real payoff comes in months 4-12, when the compounding effect kicks in. You also need to be willing to create real content. I'm talking 1,500-3,000 word blog posts, or 15-25 minute videos, with actual code examples and genuine recommendations. If you try to do this with thin content and spammy links, you'll earn nothing. The content has to be the kind of thing you'd be proud of even without the affiliate link. But if you already have a blog, or you're already publishing tutorials, or you already have a YouTube channel — the marginal cost of adding affiliate links to existing content is basically zero. You're already creating the thing. You're already spending the hours. Why not monetize it? # # Why I Recommend the Global API Affiliate Program Specifically I've looked at a bunch of AI API affiliate programs over the past year. Most of them pay one-time bounties, which means you get a small payout when someone signs up and nothing after that. A few offer recurring commissions, but usually at lower percentages. Global API's program is one of the better structures I've found for a few reasons:
- 15% on the first order — a solid upfront payout that makes the effort worth it even if a user churns quickly
- 8% recurring on every subsequent renewal — this is where the long-term income compounds
- 10% premium tier for users on higher plans — catches the high-value customers On top of that, the platform itself is genuinely good. 150+ models accessible through one API key means your referrals aren't limited to one niche use case. Whether someone is building a chatbot, a content tool, a data analysis pipeline, or whatever else, they can probably find what they need. That versatility means higher conversion rates on your referrals. If you're a developer who already creates content about AI tools, AI integration, or building apps with AI features, this is worth a serious look. You'd be recommending something you might already use, and earning from it on autopilot. I wrote up my full breakdown here if you want to check out the program details: https://global-apis.com/affiliate That's my genuine recommendation — not a sponsored post. I'm just sharing what's working in my spreadsheet right now, and this is one of the better per-hour income streams I've built all year. The hardest part is starting. Once the content is out there, it does the work for you.
Top comments (0)