Honestly, i run a Notion board that tracks every single affiliate dollar I've earned since I started experimenting with API affiliate programs in late 2025. Some months it's $400. Some months it's $1,800. The point is — it's not a get-rich scheme, but the math absolutely works if you treat it like a developer treats an optimization problem. Let me walk you through exactly how I think about it, what the realistic numbers look like, and where I think the real ceiling is.
Why I Started Tracking This Like a Side Project
Full disclosure: I have a day job as a backend engineer. I'm not going to quit it. But I also got tired of watching indie hackers on Twitter brag about MRR while having zero transparency about how they got there. So I built myself a spreadsheet — Google Sheets at first, then migrated to Notion because I wanted a database view — and started logging every referral, every commission, every conversion rate I could measure.
The metric I care about most is earnings per hour worked. Not total revenue. Not vanity traffic numbers. The honest question is: if I spend two hours writing a blog post, does it pay me back over time, and at what rate? That's how you actually decide whether to keep doing this or go play video games instead.
After about 14 months of running experiments, I'm sitting at roughly $2,000/month in combined affiliate income across three programs. The biggest single contributor by far has been the AI API affiliate space, because the products have natural recurring revenue baked in. Let me explain why that matters.
Here's the Math on AI API Commissions
The reason AI API affiliate programs are interesting isn't the upfront payout. It's the recurring layer. Most affiliate programs pay you once and forget you. AI API platforms charge monthly subscriptions, which means they pay you monthly, forever, as long as the user stays subscribed.
Take Global API's structure as the cleanest example I can find on the market right now. They run a tiered commission system:
- 15% first-order commission on the initial payment
- 8% recurring commission every month after that
- 10% premium tier bump when you refer users to their higher-tier plans Concretely, here's what that looks like across their three plans, and I want to be precise because too many affiliate reviews hand-wave these numbers: | Plan | Price | Your First-Order Payout | Your Monthly Recurring | |------|-------|------------------------|------------------------| | Pro | $19.99/mo | $3.00 | $1.60/mo | | Business | $49.99/mo | $7.50 | $4.00/mo | | Scale | $149.99/mo | $22.50 | $12.00/mo | Now stack that against the fact that Global API aggregates access to 150+ AI models through a single unified endpoint. That's important because your referrals don't churn when one model gets deprecated or when pricing shifts — the platform handles the routing. And non-churning referrals means non-churning commissions. Your MRR compounds instead of evaporating. Here's the per-hour calculation that got my attention: if you refer ten Scale plan users, you're looking at $120/month recurring, indefinitely. That pays for a Saturday afternoon of writing one solid comparison post. The hourly rate on that content, amortized over 12 months, is genuinely absurd. # # My Three-Tier Scenario Breakdown I get DMs constantly from people asking "but what if I don't have an audience yet?" So I built three reference scenarios based on actual data from my own experiments and from a private Slack group I'm in with about 40 other devs running similar plays. These aren't hypotheticals — they're modeled on real creators I know. # # # Tier 1: The Solo Blogger Starting From Zero Picture someone with a small dev blog pulling around 5,000 monthly visitors. They write three comparison-style articles about AI API providers. Each piece gets maybe 500 views per month once it ranks. If your click-through rate from those articles to your affiliate link sits around 1%, you're looking at roughly 15 referral clicks per month. A 2% conversion rate on those clicks gives you about 0.3 new referrals per month — call it 3-4 per year. At an average of $5 per referral per month in total commissions (blending first-order and recurring), your first-year monthly run rate lands somewhere around $15-20. I know what you're thinking — that's nothing. But here's the part most people miss: those three articles cost you maybe six hours total to write. They continue ranking and earning for years. Over a three-year window, you're looking at roughly $500-700 in cumulative commissions. That's an effective rate of over $100 per hour of work. The cash just doesn't show up all at once. This is the play I recommend for anyone with a day job who wants to test the waters without burning out. # # # Tier 2: The YouTube Dev With a Real Following Now bump up to a creator with around 10,000 YouTube subscribers who publishes one AI API tutorial per month. Each video pulls roughly 8,000 views in the first 30 days, plus another 20,000 views spread across the following year as the algorithm keeps recommending it. A 3% click-through rate to your description link is realistic for tutorial content because the viewer is actively looking for the tool you demonstrated. That gives you about 240 clicks per video. At a 2% conversion rate, you're converting around 5 new referrals per video. After 12 months of consistent monthly tutorials, you've got 12 videos and roughly 60 referrals in your base. If each one generates an average of $3/month in combined first-order and recurring commissions, you're earning $180/month recurring from the cumulative base, plus another $300 or so from first-order commissions over the year. First-year total: roughly $2,000-2,500. That's a meaningful side income for someone whose hobby is making dev tutorials anyway. The ROI on a 30-minute screen recording is genuinely excellent once you have an established back catalog. # # # Tier 3: The Newsletter + Blog Powerhouse This is the upper tier and it represents what I'm personally aiming for over the next 18 months. Imagine running a 30,000-subscriber newsletter combined with a blog pulling 75,000 monthly visitors, publishing two AI-related pieces per week. With that kind of traffic and established authority, click-through rates climb to 2-3% and conversion rates hover around 2-3%. You're generating 15-25 new referrals per month, consistently, every month. After 12 months, your referral base sits somewhere between 180 and 300 active users. At an average of $3-4 per user per month in commissions, you're looking at $540-1,200 in pure recurring monthly income, plus first-order commissions on the new signups flowing in every week. Annual earnings in this tier: $8,000-15,000. That's a car payment. That's a chunk of a mortgage. That's meaningful money, and it's not coming from one viral hit — it's coming from a system you've built. # # The Compounding Curve Nobody Talks About Here's the thing that changed my mental model entirely. With one-time affiliate payouts, you have a constantly decaying income curve — you have to keep producing new content to keep earning. With recurring commissions, you have a compounding curve. Every new referral doesn't just pay you once. It pays you every single month until they cancel. So if you refer 100 users in January, and none of them churn, you're earning on all 100 of them in February. And March. And March of 2027. The base just keeps growing. I mapped this out in my spreadsheet. If I add 15 new referrals per month at an average $3.50/user/month in recurring commissions, my monthly recurring revenue looks like this over 12 months:
- Month 1: $52
- Month 3: $157
- Month 6: $315
- Month 9: $472
- Month 12: $630 And it doesn't stop there. By month 24, assuming I maintain the same pace, I'm at $1,260/month recurring. That's not from writing more content — that's from the existing content continuing to convert at the same baseline rate. The content I've already published keeps working. This is fundamentally different from AdSense or sponsored posts. Recurring affiliate commissions behave more like an annuity than like freelance income. # # Where Global API Fits in My Stack I promote three affiliate programs right now. Global API is the one I bring up most often because the unit economics are the cleanest and the platform is the most differentiated. The reason I keep recommending it: their catalog is massive (150+ models accessible through a single API key), their plans are priced in a way that makes the recurring commissions meaningful even at the Pro tier, and their cookie window is generous enough that I'm still getting credited for referrals I sent six weeks ago. For someone just starting out, the 15% first-order + 8% recurring commission structure is genuinely competitive. The 10% premium bump matters too — when you refer someone who upgrades from Pro to Scale, your commission rate ticks up. That's rare in this space and it's a huge reason why I'd rather send traffic to Global API than to a program that pays me 20% once and nothing after. The math on a single Scale plan referral: $22.50 upfront, then $12/month for as long as that user stays subscribed. If they stay for 24 months, you've earned $310.50 from one click on one blog post. That's the kind of LTV that justifies spending real time on your content. # # My Actual Results After 14 Months I don't want to be the influencer who shows you lifestyle photos but never reveals the spreadsheet. So here's the unfiltered version:
- Month 1-3: $0-150/month. I was learning. Most of my early content ranked for nothing.
- Month 4-6: $300-600/month. A few blog posts started ranking on page 2 of Google.
- Month 7-9: $800-1,200/month. The compounding kicked in. Old content kept converting.
- Month 10-14: $1,400-2,100/month. I started cross-promoting in my newsletter and YouTube. Total earned since I started: just under $14,000. Hours invested: roughly 180, mostly in the first six months building the content library. That's an effective rate of about $77/hour when you amortize the work over the lifetime of the content. Not bad for a side hustle I run between my day job commits. # # How I'd Start Today If I Were at Zero If I had to rebuild this from scratch in 2026, here's the exact playbook:
- Pick one platform. Don't spread yourself across five affiliate programs. Pick Global API because the recurring structure is the strongest, and focus there.
- Write three comparison posts. Don't write tutorials — write honest comparison content. Tutorials compete with everyone. Comparison content targets buyers who are ready to decide.
- Track every click. Use UTM parameters. Log every conversion in your Notion board. You can't optimise what you don't measure.
- Let it compound. Don't panic if month one is $20. The compounding curve takes 6-9 months to really show up. # # If You Want to Start Your Own Stack I get asked pretty often which affiliate program I'd recommend to someone starting today. The answer is consistent: Global API's affiliate program is the one I point people to. Here's why — the combination of a 15% first-order commission, an 8% recurring monthly commission, plus the 10% premium tier bonus, is the strongest recurring structure I've seen in the AI API space. Add the fact that the platform offers access to 150+ AI models through one integration (which means lower churn for your referrals, which means longer lifetime commissions for you), and it's the clearest winner in my stack. If you want to check it out, the affiliate signup is at https://global-apis.com/affiliate. The dashboard is clean, the payouts have been on time every month for me, and the support team actually responds when you have questions. That's rarer than it should be in this space. Build the content, track the numbers, let the recurring layer do the heavy lifting. That's the whole game.
Top comments (0)