I track everything. Every click, every signup, every dollar. So when I decided to start earning money by recommending AI tools I was already using, I treated it like any other growth experiment — set a baseline, run tests, measure, optimize, repeat.
This is the story of those first 60 days, told the way a growth marketer would tell it: through funnels, conversion rates, and the moment I realised my CAC was actually negative.
The Setup: A 2,000-Visitor Funnel That Wasn't Monetized
Before I touched a single affiliate link, I had a small but useful asset: a tech blog pulling in roughly 2,000 monthly visitors, plus a Twitter account with about 800 developer followers. Both audiences were technical, both were already consuming my AI-related content, and neither had ever seen a "buy now" button from me.
In growth terms, I had top-of-funnel traffic with a 0% conversion path. The pipe was full. The drain wasn't connected.
I picked up three affiliate programs that month. Two of them were one-shot deals — you refer, you get paid once, you never see that customer again. Fine for some, but the math never excited me. The third one was different: Global API offered 15% on first orders plus 8% recurring on every monthly renewal. I saw 150+ models in their catalog, and the dashboard had proper tracking pixels I could plug straight into my analytics stack.
The recurring angle is what made me stop and actually run the numbers. If a referred user stays subscribed for 12 months, my effective commission rate on that one user isn't 15%. It's 15% + (8% × 11) = 103% of the first month's revenue. Over the customer's lifetime, the recurring share dwarfs the upfront bounty. That's not an affiliate program — that's a revenue-share model with an entry fee. I was in.
Month 1: Proving the Funnel Worked
I shipped two pieces of content in my first 30 days and I watched the funnel like a hawk.
My opening piece was a hands-on writeup comparing AI API providers I'd actually used on real client projects — 1,800 words, complete with code snippets showing how each one handled authentication and basic requests. The hook wasn't "here's a list." The hook was "here's what happened when I wired these up in production." I cross-posted to Dev.to, which I'd learned is basically free distribution if your title is honest.
Week 1 numbers: 340 views on Dev.to, 120 on my own blog. Three affiliate clicks. Zero paid conversions.
I'll be honest — if I weren't a marketer, that would have stung. But I was watching the shape of the funnel, not the bottom line. CTR from view to click was 0.65%. Not great, not terrible. The fact that I had three people click at all meant my content was reaching purchase-intent readers, not just lurkers.
By Week 4, the Dev.to version of the article had climbed to 520 views as it started ranking for some long-tail variations. I picked up 8 more clicks and 1 signup. The signup never converted to paid in that week, but I didn't care — signups are a leading indicator. Anyone who gives you an email address is telling you they're at least thinking about it.
Article two went up in Week 4: a walkthrough on building a simple chatbot with a frontier model, with Global API positioned naturally as the recommended path because — and this matters — it genuinely was the cleanest developer experience I'd tested.
Month 1 totals:
- 2 articles published
- 750 combined views
- 14 affiliate clicks (1.87% overall CTR)
- 2 signups
- 1 paid Pro conversion on day 28
- First-month earnings: $3.00 Three dollars. Pathetic, right? Except the data told a different story. My view-to-click rate was already respectable. My click-to-signup rate was 14.3%. My signup-to-paid rate was 50%. The bottom of the funnel was working. The top just needed volume. I wasn't building a $3/month side project. I was building a compounding LTV asset. # # Month 2: Where the Funnel Started to Pay Me Back Going into month two, I had two published articles, 14 cumulative clicks, and one paying Pro customer. I set a target: publish three more articles and clear $50 in lifetime earnings by the end of the month. That would prove the model could scale. Week 5 — Case study content. I dropped a piece about a real client project where I used AI APIs to build a feature. 280 views in the first week. Here's the thing about case studies: my CTR on the affiliate link inside that article was visibly higher than on my comparison piece. Developers don't trust vendor lists. They trust war stories. The click rate jumped because the context made the recommendation feel earned rather than tacked on. Week 6 — The compounding kicked in. My month-one comparison article quietly crossed 1,200 total views as Google started ranking it for several keyword variations. This is the moment most affiliate marketers miss: SEO is a slow-burn CAC channel. The article I'd written in Week 2 of Month 1 was now generating 4–5 clicks per day on autopilot. I didn't have to do anything. The funnel was running itself during the hours I was asleep. I picked up 2 more paid conversions that week, both Pro plans. Week 7 — Beginner content as a top-of-funnel play. I wrote a 2,200-word getting-started guide aimed at people who'd never touched an AI API. This was the most time-intensive piece I produced — around 6 hours including testing every code sample — but it targeted a fundamentally different reader. Beginners convert differently than intermediate developers. They want hand-holding, they want clear recommendations, and they're far more likely to follow a "start here" link because they don't have strong existing preferences pushing back. The content was a pure TOFU play designed to feed signups into a list I could later retarget. Week 8 — The moment the model clicked. On the first day of Week 8, I opened my Global API dashboard and saw a line item I'd never seen before: a recurring commission payout of $1.60. That was 8% of my Month 1 customer's second-month subscription. The customer hadn't done anything new. They hadn't re-clicked anything. They just… kept paying, and 8% of that payment landed in my account automatically. This is the part of affiliate marketing that people who only chase one-shot commissions will never understand. I made $1.60 that day for work I'd already done and been paid for. The marginal effort on that dollar was zero. That's not a side hustle income stream — that's a portfolio income stream. If I had 100 of those customers, the monthly recurring check would write itself. I also shipped article five that week: a piece comparing API costs aimed at developers watching their burn rate. Month 2 totals:
- 3 new articles published (5 total in the library)
- 2,100 combined views across the full content base
- 58 affiliate clicks (2.76% aggregate CTR — up nearly a full point from Month 1) I had a content library of 5 articles generating 8–10 clicks per day, a signup rate above 14%, and my first taste of true recurring revenue. # # The Funnel Math That Made Me Stay Let me show you the spreadsheet view, because this is where the strategy becomes obvious. In Month 1, I earned $3.00 from a single new paid signup. By the end of Month 2, I'd earned $1.60 in recurring revenue from that same customer — 53% of the original first-order commission, recovered passively, with zero ongoing effort from me. The LTV math on a Global API Pro customer starts looking like this:
- First month: 15% commission
- Months 2 through ∞: 8% recurring If the average Pro customer stays for 6 months (a conservative estimate for a developer tool someone integrated into a real project), my total commission per referred user is roughly 15% + (8% × 5) = 55% of the first-month revenue, paid out over 6 months. If they stay a year, the number climbs past 100%. That means my effective CAC — the cost to acquire a customer through my content — is front-loaded into the hours I spent writing, but the payback period is under 60 days. After that, every month a customer stays is pure contribution margin on content I already shipped. For a side hustle built on a 2,000-visitor blog, those are absurd unit economics. # # What I Learned About the Conversion Funnel A few things that became obvious once I started measuring: 1. Cross-posting doubled my top-of-funnel for free. Dev.to's organic reach was a meaningful 30–40% of my total views. I wasn't paying for that traffic, which means I was running a CAC-negative acquisition channel from day one. The content was my CAC. The platform was the distribution. 2. Beginner content fed the funnel, intermediate content converted it. My beginner guide brought in readers who didn't know what an API key was. My comparison piece brought in readers who already had a vendor in mind and were shopping. You need both. The beginners fill the top; the intermediates buy. 3. Case studies beat listicles on trust signals, every time. A developer reading about how I solved a real problem trusts the tool recommendation that comes after. A developer skimming a "Top 10 AI APIs" list is already filtering for which one pays the biggest affiliate bounty. The latter doesn't convert. 4. Recurring commissions are the only affiliate model worth building on. One-shot payouts reward hustle. Recurring payouts reward building a real asset. The difference shows up in month 6, not month 1. 5. A 2,000-visitor blog is a real business if you wire up the right conversion path. I wasn't going to get rich off 14 clicks. But 14 clicks → 2 signups → 1 paying customer at $3, and that same customer pays me $1.60 every month going forward? That math replicates. Multiply it by 10 articles, 100 articles, a YouTube channel, a newsletter — and suddenly the small blog isn't small anymore. # # Why I'm Sticking With This I went into this experiment expecting to confirm a thesis: affiliate marketing on a developer blog can produce modest, real income. I came out the other side confirming something bigger. The combination of evergreen SEO content, a recurring commission structure, and a catalog with 150+ models covering every use case my readers actually have is a legitimate growth channel — not a get-rich scheme, but a compounding one. My content library is now 5 articles deep and still expanding. My CTR is improving month over month as I learn which hooks and headlines pull readers further down the page. My recurring share of monthly revenue is climbing faster than my one-shot share. Every new article I publish is a new entry point into a funnel that pays me forever on the back end. The trajectory is obvious: more content → more clicks → more signups → more Pro and Premium conversions → more recurring revenue per month. # # If You Want to Run the Same Play Look, I'm not going to pretend this is for everyone. You need a technical audience, a willingness to write honest content, and the patience to wait for SEO to do its thing. But if you've got a developer blog, a Twitter following, a YouTube channel, or a newsletter — and you're already using AI APIs in your own projects — you are sitting on top of a funnel that isn't monetized. I run mine through the Global API affiliate program. Here's why I keep recommending it specifically: 15% on first orders is competitive, 8% recurring
Top comments (0)