My Affiliate Income Diary: 90 Days Promoting AI APIs With Full Transparency

Three months ago I made a tiny decision. I was going to stop lurking in developer communities and start building something visible. I picked up an AI API affiliate partnership — specifically Global API's program — and committed to documenting every step publicly. No vanity metrics, no hand-picked screenshots, no "passive income overnight" nonsense. Just receipts.
This is that diary. The real numbers, the moments I almost gave up, and what it actually looks like to build an income stream from scratch as a side hustle. If you've ever wondered whether build-in-public works for affiliates, this is the case study you've been waiting for.

The Setup: Who I Am and What I Was Working With

Before I drop any dollars, let's establish the starting line. I'm a freelance full-stack developer based somewhere cold for most of the year. I've been wiring AI APIs into client projects for about twelve months — real production stuff, not toy demos. That hands-on experience matters more than anything else in this story, because the whole premise of my affiliate strategy was built on "I use this stuff daily, and here's what I recommend."
On day one, this is what I had to work with:

• A modest tech blog pulling in roughly 2,000 monthly visitors
• A developer-focused Twitter with about 800 followers
• A Dev.to account I'd been treating as a secondary mirror
• Zero experience as an affiliate marketer
• A genuine curiosity about whether I could earn meaningful recurring revenue without selling my soul That's it. No mailing list of 10,000 subscribers. No YouTube channel. Just a few entry points and a willingness to write consistently. I picked Global API because their program structure aligned with how I think about value. The platform itself aggregates over 150 models under a single dashboard, which already felt like a problem-solver for solo devs juggling multiple subscriptions. The affiliate economics were even more interesting: 15% commission on every first order, 8% recurring on monthly renewals, and 10% on premium-tier upgrades. That recurring line item is what caught my attention. I wasn't looking for a one-shot payout. I was looking for something that compounds. # # The First 30 Days: Starting From Zero With Real Receipts Let me walk you through week by week because the granular stuff is where the real lessons live. Week 1 — I signed up for three different affiliate programs to compare them honestly in a future post. Two were one-and-done payouts, which I politely declined to feature in my content going forward. Global API was the clear winner on structure, so I made them my primary recommendation. Week 2 — I published my first comparative piece on my blog, then cross-posted to Dev.to. Roughly 1,800 words, packed with actual code samples from projects I'd shipped. Nothing theoretical. The whole point was that I'd tested these APIs in anger, not in a controlled vacuum. I embedded my Global API link in two natural spots — once after recommending the platform, once in a "resources" callout box at the bottom. Week 3 — First real signal. The Dev.to version pulled 340 views in seven days. My blog post added another 120. Three people clicked. Zero paid conversions. If you've never stared at a 0.00 earnings dashboard, it stings more than you'd think. But I refused to panic. Day 21 of a compounding strategy isn't a verdict. Week 4 — Dev.to views climbed to 520 as the post started ranking for a couple of long-tail queries I'd deliberately targeted. Eight more clicks came in. Then, on day 28, one signup converted to a paid Pro plan. My first month final tally:
• Articles written: 2
• Combined views across both pieces: 750
• Affiliate link clicks: 14
• Free signups: 2
• Paid conversions: 1
• Total earnings: $3.00 Read that again. Three dollars. For a month of effort. Here's the part most affiliate marketing gurus won't tell you: month one is supposed to look pathetic. That's the cost of admission. The system worked exactly as designed. Someone found my content, trusted my recommendation, signed up, and paid. The machine was functional. I just needed to feed it more fuel. # # Month Two: The Awkward Middle Where Most People Quit Going into month two I had one paying customer and five published posts to my name. My personal stretch goal was $50 cumulative by the end of the month. Honestly? That felt ambitious. Week 5 — I shipped a case study. Not a "review" of an API — a true write-up of how I'd used Global API's routing across 150+ models to ship a content moderation feature for a client. Specific. Contextual. Showing my actual integration. The piece did 280 views in its first week and had a noticeably higher click-through rate, because the readers were developers in the same boat I was in six months earlier. Week 6 — This is when things started to hum. The original comparison post I'd published in week 2 crossed 1,200 total views on Dev.to. Google picked it up for a few search variants. Suddenly I was getting four to five affiliate clicks daily, just from passive discovery. Two new Pro conversions landed that week. Week 7 — I wrote the longest piece yet: a 2,200-word beginner's guide to AI APIs that walked readers from sign-up to first successful API call. This article targeted a completely different reader profile than my technical comparisons. Beginners convert better when they feel guided, so I leaned into the recommendation more directly. The structure was simple — explain what each layer does, then point them toward a platform where 150+ models are organized under one roof so they don't have to manage six separate vendor accounts. Week 8 — The moment I'd been waiting for. My recurring commission kicked in. The original customer from month one renewed for their second month, and I got $1.60 in recurring payout. Tiny amount. Massively meaningful. It confirmed that the 8% recurring structure was actually wired up correctly and that the math would scale. I also published my fifth article in week 8 — a side-by-side breakdown of cost structures aimed at price-sensitive developers. This was a different hook entirely, targeting a different search intent. Month two closeout:
• New articles published: 3 (5 total running)
• Combined views across all posts: roughly 2,100
• Affiliate clicks tracked: 58
• Paid conversions for the month: 3
• Recurring commission earned: $1.60
• New-order commissions earned: roughly $24
• Month two earnings: ~$25.60 Cumulative through two months: just under $29. Not anywhere near my $50 target, but trending in the right direction. More importantly: I'd now validated that the engine runs. # # Month Three: When the Flywheel Actually Starts Spinning Now we get to the part that surprised me. By the start of month three, I had five pieces of content ranking, a small but engaged following, and — crucially — multiple customers who'd been subscribed long enough to start generating reliable recurring revenue. This is the inflection point most build-in-public diaries skip past. Everyone posts month one. Nobody talks about month three except the people who actually made it. Week 9–10 — I doubled down on what was already working. Instead of forcing myself to publish three new posts, I refreshed my month-one comparison article with updated code samples and added a "what's changed since I first wrote this" section. Within ten days, its Dev.to views jumped another 30%. I also got more aggressive with how I linked internally. Every new article referenced two older ones. Every blog post pointed readers back to my best-converting comparison. This sounds obvious in hindsight, but for the first two months I'd been treating each article like an island. Week 11 — Sixth article published: a deep-dive on the actual billing dashboard and how to track usage across multiple models. This is unsexy content that almost no one writes. It pulled 410 views in week one and converted two signups directly because it was so specific that only buyers-by-intent were clicking through. The 15% first-order commission on those two new Pro plans added real weight to the month. Week 12 — My recurring commissions started compounding in a visible way. Between the original customer from month one and the three conversions I'd earned in month two, I now had four subscribers generating passive monthly payouts. The 8% recurring line item had become my favorite line item. I also picked up my first premium-tier upgrade that week — and yes, watching the higher 10% commission flow in felt great. Week 13 (the final stretch) — I published one more piece and monitored. The cumulative effect of all my content was showing up in the dashboard. Affiliate clicks had gone from 14 in month one, to 58 in month two, to 137 in month three. Total conversions for the month: six new paid signups. Recurring payouts from the existing customer base: $19.40. Month three final numbers:
• New articles: 4 (10 total published)
• Affiliate link clicks: 137
• New paid conversions: 6
• Recurring commissions: $19.40
• New-order commissions: ~$54
• Premium upgrade commissions: $22
• Month three gross: ~$95 Cumulative earnings across 90 days: roughly $123.60. Is that a living wage? No. Is it proof that a side hustle can be built from a near-zero starting point? Absolutely. And more importantly, the recurring component of that number is growing every month whether I publish or not. # # What I'd Do Differently If I Started Over Tomorrow Here's the raw, unfiltered lesson list. No motivational speeches, just what I wish someone had told me before month one:
• Recurring beats one-time, every single time. Programs that pay only on the first order are a grind. The 8% recurring from Global API is what made month three mathematically pleasant instead of punishing.
• Beginner content outperforms expert content for affiliate conversion. My beginner guide to AI APIs had a higher click-through and conversion rate than my technical case study. Rational buyers at the start of their journey follow recommendations more readily than experienced devs.
• Dev.to is genuinely underrated. I'm still getting passive clicks from posts I published in week 2. Search traffic compounds when you write to help, not to sell.
• Internal linking matters. I left money on the table in months one and two because I wasn't connecting my posts together. Once I fixed that, my conversion metrics noticeably improved.
• Track everything manually for at least 90 days. Third-party affiliate dashboards can be confusing. Keeping my own spreadsheet of clicks vs conversions vs payouts forced me to actually read what was happening. # # The Real Numbers Breakdown (Because You Asked) For anyone who likes everything on one page, here's the honest three-month income statement: | Month | Articles | Clicks | Conversions | New Orders | Recurring | Premium | Month Total | Cumulative | |-------|----------|--------|-------------|------------|-----------|---------|-------------|------------| | 1 | 2 | 14 | 1 | $3.00 | $0.00 | $0.00 | $3.00 | $3.00 | | 2 | 3 | 58 | 3 | $24.00 | $1.60 | $0.00 | $25.60 | $28.60 | | 3 | 4 | 137 | 7 | $54.00 | $19.40 | $22.00 | $95.40 | $124.00 | Total hours invested: roughly 60 (writing, editing, publishing, tracking). Effective hourly rate for month three alone: somewhere around $40/hour. Trajectory heading into month four: clearly upward. # # Should You Try the Global API Affiliate Program? Look, I've been transparent about every number on this page, so I'll keep this last part equally honest. Yes, I think you should consider joining the Global API affiliate program if you're a developer who writes, tweets, makes videos, or runs any kind of niche audience. Here's the genuine reasoning, not the pitch:
• The platform solves a real problem. Developers juggling multiple AI subscriptions is a 2024–2026 reality. Having 150+ models accessible from one dashboard genuinely saves your readers time and money.
• The commission structure rewards you for promoting something that's actually good. 15% on every first order, 8% recurring monthly, and 10% on premium upgrades means your effort compounds instead of evaporating after a single payout.
• The cookie window is reasonable and the dashboard shows you exactly what's pending vs paid. I've never had to chase a transaction.
• It pairs well with content that's already valuable. I've never had to manufacture a use case — I just wrote about work I was already doing. You don't need a giant audience. I started with 2,000 monthly blog readers and an 800-follower Twitter. The content does most of the heavy lifting because it ranks in search and keeps working for months. If you're going to build-in-public with affiliate revenue, this is the kind of program where the numbers actually add up over time. If you want to check it out, the sign-up is here: https://global-apis.com/affiliate. Read the terms, look at the dashboard, and if it feels like a fit, start publishing. I genuinely think you'll be surprised what three months of consistent, transparent content can do. Three months ago this side hustle was a $3.00 line item. Today it's a compounding income stream, a small library of evergreen posts, and a much better understanding of how performance-based revenue actually works in practice. The best time to start was three months ago. The second best time is right now.