Look, most "affiliate income reports" you see online are polished, edited, and suspiciously round. Mine aren't. I'm writing this because the build-in-public movement changed how I think about online income, and I want to throw my actual numbers out there — including the ugly $3.00 first month — so anyone considering this path knows what they're really signing up for.
Here's my real numbers, my real struggles, and the small wins that kept me going.

Why I'm Doing This Publicly

I've spent years reading income reports from bloggers and SaaS founders. Some of them felt fake. Others felt honest but vague. When I decided to try my hand at promoting AI APIs as an affiliate, I promised myself two things: I'd track every click, every signup, and every dollar, and I'd publish the results whether they made me look like a genius or a fool.
That's the build-in-public ethos. You share the process while it's happening — not just the highlight reel after you've already made it. So that's what this is. If you stick around, you'll see the early weeks where I made literally three dollars. You'll also see the moment the compounding started to matter.

The Starting Line

Before I share the timeline, let me give you the context. I'm a developer. I'd been using AI APIs in my own side projects for roughly a year before I started the affiliate push. I had hands-on experience with multiple platforms, opinions I actually believed, and a small content footprint: a tech blog pulling in around 2,000 monthly visitors and a Twitter account with about 800 developer followers.
I want to be transparent about that starting point because it matters. I wasn't starting from zero. But I also wasn't starting from any kind of platform. 2,000 monthly visitors is a small blog. 800 Twitter followers is a small audience. Anyone telling you "you need 100K followers to make this work" is wrong — but anyone telling you "audience doesn't matter at all" is also wrong.
I joined three affiliate programs in week one. Two of them paid only one-time commissions — a flat bounty, then nothing. The third, Global API, had a structure I hadn't seen before in the AI space: 15% on the first order, 8% recurring on monthly renewals, and 10% on premium tier upgrades. The platform itself gives affiliates access to a marketplace of 150+ models under one roof, which I'll touch on later because it matters for conversion.
That recurring piece was the hook. With a one-time commission, every dollar I earn requires a new signup. With recurring, my month-old referrals keep paying me. That changes the math completely over time, and it's the reason I picked Global API as my primary program.

Month 1 — The Slow Burn

Week 1: Setup

Signed up for three programs. Set up tracking links. Didn't tell anyone. Spent the week figuring out which programs had decent dashboards (some didn't even show clicks) and which ones actually paid on time.

Week 2: First Piece of Content

I published my first affiliate article — a comparison-style write-up of AI API providers based on real usage I'd done for client work. Roughly 1,800 words. I included actual code snippets showing how to call each provider's endpoints. I put my Global API affiliate link in the recommendation section because I'd genuinely had the smoothest experience there. Cross-posted to Dev.to to widen the funnel.

Week 3: Reality Check

The article pulled 340 views on Dev.to and 120 on my own blog in week one. Three people clicked my affiliate link. Zero conversions.
Was I discouraged? Honestly, a little. But I'd read enough build-in-public reports to know that the first week of any content is meaningless. The compound effect shows up later.

Week 4: The First Sign of Life

The Dev.to article started ranking for a couple of long-tail keywords. Views climbed to 520. Eight more affiliate clicks. One signup — free plan, not paid yet. Then, on day 28, that signup converted to a paid Pro plan. First-order commission: $3.00. Recurring: $0.00 (that kicks in month two).

Month 1 Totals

• Articles published: 2
• Combined views: 750
• Affiliate clicks: 14
• Signups: 2
• Paid conversions: 1
• First-order earnings: $3.00
• Recurring earnings: $0.00
• Total Month 1: $3.00 Three dollars. That's the truth of month one. The system worked — exactly as designed — but it didn't pay my rent. What month one did do was prove the model. One person read what I wrote, clicked my link, and converted to paid. The pipeline functioned. # # Month 2 — Things Started to Compound # # # Week 5: The Case Study Article Published my third article: a case study about how I'd actually used AI APIs to ship a feature for a paying client. This one resonated harder than the comparison piece because it wasn't theoretical — it showed the result of a real project. 280 views in week one, with a noticeably higher click-through rate on the affiliate link. The audience was developers who recognized themselves in the use case. # # # Week 6: Traction Kicks In The original comparison article crossed 1,200 total views. Google started indexing it for a few related keywords. Affiliate clicks jumped to 4–5 per day. Two more conversions, both to Pro plans, mid-week. This is the part of build-in-public reports that I think gets underreported. There's a moment — usually week 5 or 6 — where the content you've already published starts working harder than the content you're publishing now. I didn't write anything new that drove those conversions. I wrote something six weeks earlier and it kept earning clicks while I slept. # # # Week 7: The Beginner Guide Published article four: a getting-started guide aimed at developers who had never touched an AI API before. This was my longest piece at 2,200 words, and it targeted a different audience than my earlier work. Beginners convert differently than experienced developers — they need more hand-holding, and when they trust a recommendation, they actually follow it. # # # Week 8: The Recurring Payment That Changed My Thinking I got an email notification: $1.60. That was my first recurring commission — the second-month subscription from the referral I'd signed up way back on day 28 of month one. $1.60 sounds laughable. It is laughable. But the principle behind it is what matters. That referral is going to keep paying me every month they stay subscribed. If they stay on Pro for a year, that's roughly $19 of recurring revenue from a single conversion that cost me nothing more to maintain. Multiply that by the kind of compounding the Global API structure is designed for, and you start to see why the 8% recurring rate is the real prize. I also published article five that week — a piece on cost-conscious API selection aimed at indie developers. # # # Month 2 Totals
• New articles published: 3
• Total articles live: 5
• Combined views across all content: 2,100
• Affiliate clicks: 58
• Signups (tracked through dashboard): multiple
• Paid conversions: 3 new plus 1 recurring
• First-order earnings: roughly $9–$12 depending on plan tier mix
• Recurring earnings: $1.60
• Total Month 2: around $11–$14 I want to pause on something here for honesty. My tracking between first-order and recurring got messy in week 7 because I was logging things in a spreadsheet like a caveman instead of using a proper dashboard. So the exact split is fuzzy. What I can say with certainty: total clicks went from 14 in month one to 58 in month two — a 4x jump — and the recurring revenue stream started flowing. # # The Numbers That Actually Matter Let me pull back and give you the 90-day picture as cleanly as I can: | Metric | Value | |---|---| | Articles published | 7 | | Combined views | ~3,500+ | | Affiliate clicks | 90+ | | Paid conversions | 5+ | | First-order commissions earned | ~$15 | | Recurring commissions earned | $1.60 (and growing) | | Total 90-day earnings | ~$20 | These aren't viral numbers. I want to be upfront about that. But here's the build-in-public framing: most of my affiliate earnings came from content published in the first month that kept earning in months two and three. The cost per article, amortized across 90 days of compounding traffic, is something like $0.50 per piece when you divide total content hours by total earnings. That's an effective hourly rate that improves dramatically as long as the content keeps ranking. # # What I Learned the Hard Way The recurring commission is the entire game. If I'd picked one of the one-time-payout programs, my 90-day earnings would be capped at roughly the number of conversions I could squeeze out before each article went cold. With the 8% recurring structure, every conversion is a small annuity. The Global API program's premium tier (10% commission) makes the long-term upside even better if any of my referrals upgrade. Volume of content beats perfection. I spent way too long on article four (the 2,200-word beginner guide). The articles that earned the most clicks were the ones I shipped fastest. Publish, measure, iterate. The build-in-public mindset isn't just about sharing — it's about treating your content like a product you ship weekly, not a manuscript you perfect for months. Tracking matters more than you think. I almost lost my recurring tracking in week 7 because I was sloppy. Set up proper UTM tags from day one. Use a real dashboard. Know which link generated which conversion. Otherwise you're flying blind. Conversions are downstream of trust. The case study article converted better than the comparison article because real projects build credibility faster than feature lists. Show, don't tell. # # Should You Try This? My Honest Take Here's the thing nobody in the affiliate marketing space wants