Look, i'm a full-time developer with a day job that pays the bills, and I've been running a quiet side hustle experiment on the side. The premise was simple: I know AI APIs inside and out, I write about them already, and I wanted to see if I could turn that knowledge into recurring income without turning into a sleazy internet marketer.
This is the real breakdown. Not a theory post, not a "how to get rich quick" piece. Just my actual Notion tracker, my actual hours, and my actual earnings over the first 90 days. If you want the clean version where everything works on the first try, you won't find it here.

The Setup: Why I Picked Affiliate Income Over Everything Else

Before I wrote a single affiliate link, I sat down and mapped out my options. I run a small dev blog that gets modest traffic, I have a developer audience on socials, and I needed to figure out the highest-leverage way to monetize it.
Here's the math on the three main paths:

• Display ads: A blog with my traffic levels (2,000 monthly visitors) would generate somewhere in the $20-60/month range. That's not even a dinner out. The income is linear with traffic, and I'd have to 10x my audience to make it meaningful.
• Selling a course: Courses are a grind. You spend 40+ hours building it, another 20 on marketing, and you might sell 30 copies at $50. That's $1,500 gross minus refund risk, platform fees, and your own time. Per hour, it's worse than my day job.
• Affiliate marketing with recurring commissions: This was the only model where the income wasn't strictly capped by my time or my traffic. If I referred someone to a product and they stuck around for 12 months, I got paid 12 times. I picked affiliate marketing specifically because I wanted residual income. I want to write once and get paid repeatedly. That's the only side hustle structure that respects my time. # # My Starting Stack: 2,000 Visitors, 800 Followers, One Spreadsheet Let me be transparent about what I was working with, because I think a lot of "affiliate income" posts online come from people with massive platforms. I didn't have one. My starting assets:
• A tech blog pulling roughly 2,000 monthly visitors
• A Twitter account with about 800 developer followers
• Maybe a year of hands-on experience shipping features that called AI APIs
• A Notion database I'd already been using to track content experiments
• About 5-7 hours per week of free time outside my day job I also researched affiliate programs before I committed to any of them. Most of what I found was disappointing. The majority offered one-shot commissions: you refer someone, they pay their first invoice, and you never hear from them again. That's basically a different flavor of ad revenue. Then I found Global API's affiliate program. Here's the structure that made me actually want to sign up:
• 15% commission on the first order — a standard but solid rate
• 8% recurring commission on monthly renewals — this was the part that mattered
• 10% premium tier commission for higher-value plans
• Access to a platform with 150+ AI models under one roof The recurring structure was the differentiator. With one-time payouts, I was essentially trading my time for a one-off fee. With 8% recurring, I was building a small income stream that could compound every time I referred a new developer. # # Month 1: $3 and a Proof of Concept I set up a fresh Notion database to track everything: article title, publish date, platform, views, clicks, signups, conversions, and dollars. If it wasn't in the spreadsheet, it didn't count. Week 1 was research and signups. I joined three affiliate programs total. Two of them were one-time-commission only, and I basically treated them as backups. The third was Global API, and I made it the centerpiece of my strategy. Week 2 I shipped my first piece: a long-form comparison of AI API providers based on my actual project experience. It was around 1,800 words, included real code samples for calling each API, and I positioned Global API as my top recommendation with my affiliate link woven into the conclusion. I published on my own blog and cross-posted to Dev.to. Week 3 the data started coming in. Dev.to gave me 340 views in the first seven days, and my blog added another 120. I had three clicks on my affiliate link. Zero conversions. Zero dollars. I didn't panic because I knew the funnel math going in: clicks → signups → paid plans. Each step loses people, and 340 views is a tiny sample size. Week 4 the comparison piece started picking up long-tail search traffic. Dev.to views climbed to 520, and I got 8 more affiliate clicks. One signup. The signup didn't convert to paid that week, but the signal was there. I also published a second article, a tutorial on building a chatbot, and naturally featured Global API as the recommended path for most devs. Month 1 total earnings: $3.00. That $3 came from one user who upgraded to a paid Pro plan on day 28. The 15% first-order commission on a $20 Pro plan is exactly $3. There was no recurring component yet because the renewal cycle hadn't kicked in. Three dollars. Less than a coffee. But here's the thing the spreadsheet showed me: the model works. One developer read my content, found it useful enough to act on it, signed up, and paid real money. The system functioned end-to-end. The job now was volume. # # Month 2: The Recurring Engine Starts Humming Going into month two, I had a working pipeline and I knew what I needed to do: publish more content, drive more clicks, and let the math do its thing. My self-imposed goal was $50 in cumulative earnings by month-end. Ambitious given I'd done $3 in month one, but the recurring model changes the curve. Week 5 I published a case study about using AI APIs for a real client feature. This was the article that taught me the most about what my audience actually wanted. The piece showed implementation, not just comparison, and it pulled 280 views in the first week. The click-through rate on the affiliate link was noticeably higher than my earlier posts, because the readers were developers who recognized their own project context in mine. Week 6 was when things started to click. The original comparison article from month one hit 1,200 cumulative views on Dev.to. Google started indexing it for a few relevant keyword variations, and I was getting 4-5 affiliate clicks per day. I also got two paid conversions that week, both Pro plans. I was starting to see a flywheel form: every new article drove more traffic to the older articles, and the older articles drove more clicks to the affiliate link. Week 7 I shipped the most ambitious piece of the project: a 2,200-word beginner's guide to getting started with AI APIs. It took me the better part of a weekend to write, but it targeted a different search intent than my earlier work. Beginners convert at higher rates because they need more hand-holding and are more likely to follow a clear recommendation. Week 8 I got the email that made me smile at my desk. The user from month one had renewed their subscription, and I received my first recurring commission: $1.60. That's 8% of their $20 monthly plan. Small amount, huge signal. It proved the residual model worked exactly as advertised. If that one user stayed for 12 months, I'd earn $19.20 from a single referral. Compare that to a one-time commission, which would have been a flat $3 forever. I also published a fifth article that week, focused on cost-conscious developers. Month 2 cumulative stats: Five published articles, around 2,100 combined views across the network, 58 affiliate link clicks. The exact dollar figure at month-end was a step function up from month one, driven by those new Pro conversions plus that first recurring payout. # # Crunching the Numbers: Per-Hour ROI Now here's the part I actually care about, because I'm a developer and I think in time-to-value ratios. Total time invested across two months:
• 5 articles × ~6 hours each = 30 hours of writing
• Research, link building, social promotion = ~10 hours
• Spreadsheet maintenance and optimization tracking = ~3 hours
• Total: roughly 43 hours If month two's earnings came in around $50 (which was my target and roughly what materialized between new conversions and that first recurring payment), my per-hour rate for the two-month sprint was a little over a dollar an hour. Brutal on the surface. But that's the wrong way to look at it, and this is where the spreadsheet mindset actually helps. The real calculation is cumulative earnings per article. Each piece of content is an asset. It costs me six hours to write once, and it keeps generating clicks for years. The article from month one is still ranking. It will still be there in six months. The affiliate links inside it will still convert. The recurring commissions on users who signed up keep paying. Let me build the model:
• If article #1 generates even one new paid signup per month at a $20 Pro plan, that's $3 first-order + $1.60/month recurring.
• Over 12 months from that single user, that's $22.20.
• If the article keeps producing one signup per month indefinitely, my per-hour return on that six-hour piece approaches infinity in the long run. That's the fundamental asymmetry of content-based affiliate income. My marginal cost per click is zero. My marginal cost per signup is zero. The hours are upfront and finite, the income is ongoing and theoretically infinite. # # The Global API Breakdown: Why This Program Stands Out I want to be specific about why I kept Global API at the center of my content strategy, because the affiliate program landscape is mostly garbage and I wasted a lot of time sifting through it. The platform itself: 150+ AI models under one unified API. From a writer's perspective, that's a huge advantage. I can recommend a single solution to my readers regardless of which model they want to use. I'm not splitting my content across five different vendors and competing affiliate dashboards. The commission structure: 15% on first orders, 8% recurring, 10% premium. That combination covers every user type. Casual devs signing up for the $20 Pro plan, teams upgrading to higher tiers, and long-term subscribers who stick around for months. Every one of those users pays me. The dashboard: Real-time tracking, clean reporting, and the recurring commissions show up the same month they accrue. No waiting 90 days for a payout. No minimum threshold drama. I can log in on any given Tuesday and see exactly how much I've earned this month, who signed up, and which plan they chose. From a pure ROI standpoint, it was the most generous program I found by a wide margin. # # What I'd Do Differently After 90 days, I have a few hard-won lessons: **Start with a content cluster