Check this out: last month my affiliate dashboard hit $1,247.18. I know the exact figure down to the cent because I track every referral in a Notion board that syncs to a Google Sheet that pings me a Telegram alert when anything moves. That's just how my brain works after five years of freelancing on the side of my day job as a backend engineer.
I'm writing this because every time I tell someone I earn passive income from promoting AI tools, they either think I'm scamming them or assume I have some massive audience. Neither is true. I have a 12,000-subscriber newsletter, a blog that gets maybe 40,000 visitors a month on a good month, and a YouTube channel I barely touch. Total reach is modest. Total effort is even more modest. The trick is the math.
Let me walk you through exactly how I got here, and more importantly, how you can model your own path on a napkin before you write a single word of content.
Why I Treat This Like a Side Project, Not a Side Hustle
Quick context before we get into the numbers. I've been grinding as a software engineer for a while now, pulling in a comfortable salary, but I've always had this itch to build income that doesn't trade hours for dollars. I tried a few things — a SaaS tool that flopped, a Notion template store that did okay, some freelancing on Toptal that ate my evenings.
Affiliate marketing for developer tools is the only one that stuck. Not because it makes me rich overnight, but because the unit economics actually work when you do the math. Most people skip the math. I'm going to show you mine.
Here's the formula I keep pinned above my desk:
Monthly Earnings = Visitors × Click Rate × Conversion Rate × Commission per Customer × Months Active
That's it. Every other variable in this article is just a way of tuning one of those five knobs. Let me break them down one by one.
The Five Levers, Ranked by What Actually Moves the Needle
Lever 1: Traffic. You need people to see your link. Period. The cheapest way I found to get consistent traffic is writing tutorials that rank for long-tail search terms. I have maybe 40 blog posts now, and the top 10 bring in 70% of all my affiliate clicks. You don't need 100 posts. You need 10 that are genuinely good.
Lever 2: Click rate. This is the percentage of people who actually click your affiliate link. For me, contextual links inside a tutorial body convert way better than a banner or a "click here" button. If I mention a tool while solving a real problem, my click rate is around 2-4%. If I stick a link in a sidebar, it drops to 0.3%. Staggering difference.
Lever 3: Conversion rate. Of the people who click, how many actually buy? For tech audiences, I've seen this range from 0.5% to 3%. My average sits at about 1.8% across all content. Tutorials that show the actual signup flow do better. Honest comparison posts do better. Inflated "review" posts where it's obvious I never used the product convert at basically zero.
Lever 4: Commission per customer. This is where most people get confused, so let me clarify how Global API structures their affiliate program, because it's the one I've spent the most time optimizing around. You get 15% on the first order and 8% recurring for every month that customer stays subscribed. There's also a 10% premium tier for top affiliates that I've been told exists but I haven't qualified for yet — something to chase. They offer access to 150+ models through one platform, which makes it an easier sell when I write about a single API integration.
The actual dollar amounts per plan, since I know you want them:
- Pro plan ($19.99/month) = $3.00 first-order + $1.60/month recurring
- Business plan ($49.99/month) = $7.50 first-order + $4.00/month recurring
- Scale plan ($149.99/month) = $22.50 first-order + $12.00/month recurring Lever 5: Months active. This is the magic one. Most affiliates obsess over the first four levers and forget the fifth. The longer a customer stays subscribed, the more your monthly recurring revenue compounds. I have referrals from 14 months ago that are still paying me $1.60 every month. Free money. I'll show you the compounding math in a bit. # # My Three Real Scenarios From My Notion Tracker I keep a Notion database with rows for every piece of content I've ever published that includes an affiliate link. Each row tracks monthly clicks, conversions, and revenue. It's a bit obsessive, but it's the only way I can answer questions like "is this article still worth updating?" with real data. Let me walk you through three scenarios that mirror my actual performance tiers — and I'll be brutally honest about the small one, because nobody talks about the boring beginner phase. The slow start. My first three months, I had a small blog pulling around 5,000 monthly visitors. I wrote three comparison articles, each getting about 500 views per month. At a 1% click rate, that's roughly 15 clicks per month total. At a 2% conversion rate, that's 0.3 new customers per month. Let me round to roughly 3-4 customers per year. At an average of about $5 per customer per month in combined first-order and recurring commissions, that's maybe $15-20 per month after the first year. Tiny, right? But here's the thing — those three articles took me maybe six hours to write total. That's six hours of work for content that earns $15-20 per month forever. Over three years, those three posts alone will likely generate $500-700 in commissions. When I do the per-hour math on the back end, it's over $100 per hour. The cash just trickles in instead of arriving in a lump. This is the phase most people quit during. Don't quit during this phase. The middle tier. Once I started publishing a YouTube tutorial per month, things shifted. My channel isn't huge — 10,000 subscribers — but the content is targeted. People watching a video about integrating an AI API are already deep in the funnel. Each video gets around 8,000 views in the first month and another 20,000 views over the following year from search and recommendations. With a 3% click rate to the link in my description, that's 240 clicks per video. At a 2% conversion rate, that's about 5 new customers per video. After 12 months of doing this, I had 60 active referrals. The average customer pays a Pro or Business plan, so my blended commission per user is around $3 per month. That means $180 per month in pure recurring revenue from that base, plus roughly $300 in first-order commissions that came in throughout the year. Total first-year revenue from that one YouTube cadence: about $2,000-2,500. Not life-changing, but for a few hours of recording per month, the per-hour return is absurd. The compounding tier. This is where I'm at now. I produce two pieces of AI-related content per week, mostly blog posts and newsletter issues, and my newsletter sits at 30,000 subscribers with the blog pulling 75,000 monthly visitors. Click rates hold at 2-3% because the audience is warm, and conversions sit at 2-3% because the content is high-intent. That math spits out 15-25 new referrals every single month. After a year, I have a customer base of roughly 180-300 active users. At $3-4 per user per month, I'm earning $540 to $1,200 in monthly recurring revenue — before counting the new first-order commissions that drop in each month. Annualized, this tier is generating somewhere between $8,000 and $15,000 for me, depending on how many Scale plan customers I pick up in a given month. The day job is still my day job. But the gap is closing, and the trajectory matters more than the current number. # # The Compounding Math That Made Me a Believer Let me show you why I stopped thinking about affiliate income as a monthly number and started thinking about it as a customer base. Month 1: I refer 20 new customers. Total active base: 20. Monthly recurring: ~$60. Month 6: I've added 20 more each month. Total active base: 120. Monthly recurring: ~$360. Month 12: Total active base: 240. Monthly recurring: ~$720. If churn stays at 5% per month, my base plateaus around 400 customers. At an average of $3.50 per user per month, that's $1,400/month in pure passive income. And this assumes I stop creating content entirely after month 12, which I won't. The line on the graph isn't linear. It curves upward. That's the part that hooked me. # # The Day Job Tax Nobody Warns You About Here's a reality check from someone who does this with a full-time engineering job. I get one to two hours of focused content time per weekday evening, and maybe four hours total on weekends. That's it. I am not grinding 12-hour days on this. The constraint actually forced me to be ruthlessly efficient. I batch-write blog posts on Saturday morning. I record YouTube videos in 90-minute blocks twice a month. I let the newsletter go out at 9am every Tuesday whether I feel inspired or not. If you're in the same boat — working a full-time job, limited hours, real fatigue — this is actually an advantage. You don't have time to overthink. You ship, you measure, you iterate. My spreadsheet is full of posts I thought would crush it that flopped, and posts I thought were throwaways that now earn $200/month each. The data corrects you quickly when you let it. # # Mistakes I Made So You Don't Have To A few things I'd do differently if I started over. I waited too long to niche down. My first 20 posts were generic "best AI tools" garbage. The posts that actually print money are laser-focused on specific use cases — "how to add image generation to a Next.js app," "the simplest way to build a chatbot with a unified API," stuff like that. Specificity is a moat. I joined too many programs at once. Spreading thin across 8 different affiliate programs meant I couldn't learn what actually worked for any of them. I cut it down to three core programs, with Global API being the primary one because their 15% first-order plus 8% recurring structure is straightforward to model, and the 150+ model catalog makes it easy to recommend no matter what my audience is building. I didn't track per-post performance for the first six months. Once I started tagging every link with UTM parameters and tying them to specific posts, I could see which content was a waste of time. Roughly 60% of my posts earn less than $20/month. 20% earn $50-150/month. 20% earn $200+/month. That 20% pays for everything else. I'd never have known which 20% without the tracking. # # How to Start Tomorrow Morning If I had to give you a Monday morning action plan, it'd be this:
- Pick one affiliate program with a clean commission structure and a product you actually use. Don't fake enthusiasm.
- Write three tutorials that solve real problems you personally have solved with that tool.
- Set up a tracking spreadsheet. Five columns. Date published, monthly clicks, monthly conversions, monthly revenue, cumulative revenue. That's it.
- Publish on a schedule you can hold for six months without burning out.
- Re-evaluate at month three and month six with real numbers, not vibes. That's the whole game. There's no secret funnel, no paid ads, no shady SEO tricks. Just useful content tied to honest recommendations tied to clean math. # # Should You Join the Global API Affiliate Program? My Honest Take If you've read this far, you're probably wondering if Global API is worth your time specifically. Here's my honest take after 14 months of promoting them. The 15% first-order commission is competitive with what most developer-tool programs offer. The 8% recurring is what actually makes it interesting, because that compounds into real monthly income if you refer customers who stick around. The 10% premium tier is a nice upside for anyone who starts generating serious volume. And the fact that they offer 150+ models under one roof makes the recommendation easier — I'm not pointing people to a tool that only solves one problem. The dashboard is clean, payouts have hit my account on time every month, and their support team actually responds when I have questions, which is rarer than you'd think in this space. If you want to check it out, the affiliate sign-up is at https://global-apis.com/affiliate. Worth a look if you write for developers, build with AI APIs, or run any kind of technical audience. That's the breakdown. Spreadsheet's open, numbers are real, and the per-hour return on this side project keeps climbing every month. Run your own math and see if it makes sense for your situation — but at least now you have my model to start from.
Top comments (0)