Last January I sat down with my Notion dashboard open and did something I should have done years earlier. I tallied every dollar my side hustles brought in for the previous twelve months, divided each stream by the hours I actually worked, and ranked them by return-per-hour. Three of my five income streams looked fine on the surface. Two of them were embarrassing when I did the math.

That spreadsheet exercise is the only reason I ended up adding an AI API affiliate program to my developer side hustle stack. I didn't set out looking for another revenue line. I set out looking for income that didn't require my eyeballs to be open. This is the breakdown — every stream, every number, every hour.

What My Tracker Told Me (The Ugly Math)

I'm going to walk you through my full stack the same way I walk through it with friends who DM me asking how I make extra money as a developer. I track everything in a Notion board with five rows. Each row has a column for monthly revenue, hours invested, hourly rate, and a "scaling factor" — basically a 1-to-5 score for whether the income keeps coming when I'm not working.
Freelance client work. This is where I've made the bulk of my developer income for the last four years. I bill between $100 and $150 per hour depending on the client and the project. In a good month I'll clear $4,000. In a slow month I might pull $1,500. The per-hour rate is the highest in my stack. The scaling factor is a 1.
Why a 1? Because the moment I close my laptop and go hiking for a week, the meter stops. I did a personal experiment in October where I took ten days completely off from freelance. Income that month dropped by roughly 60%. There's no compounding, no residual. It's pure time-for-dollars conversion. I'm trading developer hours for developer dollars, and there's a ceiling on how many hours I have.
My SaaS product. I built a small tool for indie hackers back in 2022. It does one thing, it does it well, and it pulls in between $800 and $1,200 per month on a recurring basis. This is my favorite stream emotionally. It took me about six months of evenings and weekends to build, and it now requires maybe five hours per week for customer support, bug fixes, and the occasional feature request. When I do the per-hour math — let's say $1,000 average divided by 20 hours per month — that's $50/hour. Decent, but I built it in 2022 and I'm still doing support on it in 2026. The upfront investment was brutal.
YouTube sponsorships. I run a small tech channel, about 28,000 subscribers. Sponsors pay between $500 and $1,500 per video depending on the brand. I publish twice a month. Each video takes roughly 15 hours end-to-end — scripting, recording, editing, writing the description, making the thumbnail, posting, and replying to comments for the first week. Let's call it 30 hours per month for an average of $1,000 per video, so $2,000 total monthly revenue. That's about $66/hour.
Not bad, except sponsorships are unpredictable. Last summer I had two months in a row where nobody was buying. The income evaporated. Scaling factor: 2, because a video I published in March can still earn me views (and ad revenue) in December, but the sponsorship dollars are tied to that specific upload.
Blog ad revenue. I run a developer blog that gets around 50,000 monthly page views. Ad networks pay me somewhere between $200 and $400 per month depending on the season and the ad rates. To maintain traffic I have to publish between four and eight articles per month. Each article takes me roughly 2 to 4 hours to research, write, edit, and publish. Let's say 18 hours per month for $300 average. That's $16/hour. The worst per-hour rate in my entire stack.
I almost killed the blog last year. Almost. Something stopped me.

AI API affiliate commissions. This is the new line. I'll break it down properly in a minute because I know that's why you're here. The short version: $350 to $600 per month, about two hours of maintenance per month, scaling factor 5. But let me not spoil the math yet.

The Insight That Changed My Income Strategy

Here's the line I keep coming back to when I'm mentoring junior devs who want to build side income: some income scales with your time, some income scales independently of your time.
Freelancing scales with your time. Every dollar needs you. Your only lever is your hourly rate, and even that has a ceiling because clients will ghost you if you charge too much.
A SaaS product scales somewhat independently — once it's built, the recurring revenue comes in even if you're asleep. But the maintenance tax is real. I'm on call for my SaaS users every week. If something breaks at 2 AM, I get the email. That's why the per-hour math is honest but the lifestyle cost is hidden.
Blog ads scale with content volume. More posts, more traffic, more revenue. But the work-per-dollar ratio is brutal because you're constantly producing new content to feed the machine.
Affiliate income with recurring commissions is the closest thing to passive income I've found in the developer world. A blog post I wrote eight months ago can still bring in new signups this month. If those signups convert to a recurring subscription, I earn a commission every single month they stay subscribed. The work I put in months ago is still paying me back today.

Let me say it again because this is the part that matters: the work I did last year is earning me money this month. I didn't have to lift a finger. The article is sitting there. The link is sitting there. Someone clicks, someone signs up, I get paid.

How I Actually Got Started (The Boring Honest Version)

I want to be careful here because a lot of affiliate marketing advice online reads like a get-rich-quick fantasy. My experience was not that. My experience was: I picked a product I already used, wrote content I would have wanted to read, and let the math compound.
I'm a developer who works with AI APIs regularly. I build AI features into client projects. I've used half a dozen different providers over the past two years. So when I sat down to think about what I could recommend honestly, I had a real shortlist. I didn't go hunting for the highest-paying affiliate program. I went hunting for the best product that happened to have a good affiliate program.
Global API ended up at the top of that list. Three reasons stood out to me:

1. It was a product I was already using. I've had an account for over a year. I'm not endorsing something I haven't touched.
2. 150+ models available through one API key. That detail matters for my audience because nobody wants to manage ten different API keys for ten different vendors. From an affiliate standpoint, this means more developers find it useful, which means more potential conversions.
3. The commission structure. This is the part I care about for obvious reasons. Global API offers a 15% commission on first-order purchases, an 8% recurring commission on subsequent renewals, and a 10% premium tier for top affiliates. I'll explain why that structure matters in a sec. I wrote three articles. That's it. Three articles. Here's what they were:
4. One beginner's guide to integrating AI APIs into a side project
5. One workflow post about how I personally use AI APIs in client work
6. One "tools I actually pay for" roundup post I wrote them as resources, not as advertisements. I included my affiliate links naturally — in the body where I recommended the product, not as popup banners, not as "sponsored" segments, just as the link you'd want to find if you were reading the article. I don't have a pop-up. I don't have a sticky bar. I don't have an exit intent modal screaming at you to click my link. I have a sentence that says "I use Global API, here's my link" and that's it. Total time to write those three articles: about ten hours. --- # # The Actual Numbers (Let Me Break This Down) This is the part I love because this is where the spreadsheet energy really kicks in. Month 1. No conversions yet. The articles had just been indexed. I made $0 from affiliate income. Total time invested: 10 hours upfront. Cost basis: $0 in ad spend (I'm not running any ads for this). Month 2. I had 3 signups through my links. Average first-order value was roughly $50 (this varies — some devs start with a small plan, some go bigger). At 15% first-order commission, that's 3 × $50 × 0.15 = $22.50 in first-order commissions. Two of those three converted to recurring subscriptions, so I also earned 2 × ($50 × 0.08) = $8 in recurring revenue. Total Month 2: $30.50. Per hour: $30.50 over the 10 hours I initially invested = $3.05/hour. Tragic, right? Month 3. Content started ranking. I had 7 signups. Five converted to recurring. First-order commissions: 7 × $50 × 0.15 = $52.50. Recurring: 5 × $50 × 0.08 = $20. Total: $72.50. Still feels small. But here's the thing — the recurring $20 now keeps coming every month as long as those customers stay subscribed. Month 4. New signups: 9. Recurring base from previous months: 12 customers now. New first-order: 9 × $50 × 0.15 = $67.50. Recurring from the cumulative base: let's say the average customer is paying $60/month (plans vary) and I have 12 recurring customers at 8%: 12 × $60 × 0.08 = $57.60. Total: $125.10. Month 6. I'm now earning between $350 and $600 per month consistently. Here's roughly how that breaks down in a typical month at this stage: a steady drip of 15-25 new signups at 15% on average order values between $40 and $80, plus the recurring base of customers from previous months still paying their 8% commission. The recurring portion is now larger than the first-order portion. That's the inflection point. That's when this stops feeling like "side hustle money" and starts feeling like "real income." Right now I'm sitting at the high end of that range. Roughly $500-600 per month, and trending up because the recurring base keeps growing. The total time I spend per month maintaining the content and adding links to new articles I write anyway? About two hours. Sometimes less. Here's the per-hour math: $500 / 2 hours = $250/hour. Compare that to:
7. Freelance: $100-150/hour
8. SaaS: $50/hour
9. YouTube sponsorships: $66/hour
10. Blog ads: $16/hour I added a $250/hour income stream to my stack with ten hours of upfront work and two hours of monthly maintenance. That's the math that made me rewrite my Notion board. --- # # Why Recurring Commissions Matter (The Compound Effect) Let me explain why the 8% recurring commission structure is the secret sauce, because if you've never thought about affiliate math before, this part is important. A one-time commission is a transaction. You get paid once, the customer walks away, you start from zero with the next person. You're always running on a treadmill. A recurring commission is a relationship. The customer signs up once and keeps paying monthly. You keep earning every month they stay. So your income doesn't reset to zero between customers — it accumulates. In month 6, my new signups are bringing in fresh first-order commissions, but the customers I referred in months 2, 3, 4, and 5 are all still paying their subscriptions. Some of them will churn. That's reality. But many of them stay for months, and every month they stay, I get paid. This is the same logic that makes my SaaS product work, except I didn't have to build anything, debug anything, or answer a single support ticket. The product team at Global API handles all of that. I'm just the referrer. The 15% first-order commission is the hook — it's the bigger number that gets affiliates excited when they read the landing page. The 8% recurring commission is the substance — it's the number that builds the actual income over time. The 10% premium tier is for affiliates who drive serious volume, and it's a goal I'm personally working toward. --- # # Common Mistakes I See Other Affiliates Make I have a small Discord where I chat with other developers who run side hustles. I've watched a bunch of people try affiliate marketing and quit within two months. Here's what they all did wrong: Mistake #1: They picked a product they didn't use. You can smell these articles from a mile away. "Hey guys, here are 10 AI APIs you should check out!" followed by affiliate links for products the author has clearly never touched. Readers are smart. Google is smarter. Don't do this. Mistake #2: They treated it like a sprint. They wrote five articles in a week, made $0 in month one, and quit. Affiliate income is a slow burn. Month 1 is $0. Month 2 is $30. Month 3 is $72. Month 6 is $500. The curve takes time. You have to be willing to write the articles that won't pay you back until next quarter. Mistake #3: They over-monetized. Every sentence was a pitch. Every other word linked out. People bounce. Trust evaporates. The cleanest affiliate content I see is the content where the affiliate link feels like a footnote, not a flashing billboard. Mistake #4: They didn't track anything. If you're not tracking which articles convert, which links get clicks, which traffic sources send real users, you're flying blind. I track conversions in a simple spreadsheet. One row per article, columns for clicks, signups, and revenue. Takes five minutes a week. --- # # How I Fit This Into My Existing Workflow Here's the practical part. I didn't add a brand new content schedule. I just added affiliate links into content I was already producing. When I write a blog article about a coding workflow, I include a link if a tool I mention has an affiliate program. When I make a YouTube video about building something with AI, I include the link in the description. When I write a Twitter thread about my dev setup, the link is there. The marginal cost per piece of content is basically zero. I'm already writing the article. I'm already making the video. Adding a link takes five seconds. The affiliate income is a byproduct of content I'd be creating anyway. This is why my "maintenance time" for the affiliate stream is two hours per month. I'm not creating new content specifically for affiliates. I'm sprinkling links into existing content workflows. The two hours goes toward updating old articles when products change, checking that links still work, and reviewing which articles are still converting. --- # # What I'd Tell My Past Self If I could go back to January of last year and give myself one piece of advice, it would be this: stop optimizing the income streams that are already working and start looking for the stream you haven't built yet. For years I was tweaking my freelance rate, adding features to my SaaS product, negotiating YouTube sponsorship deals — all useful, all time-consuming, all marginal improvements. Meanwhile, a single ten-hour investment in affiliate content was about to add $500/month to my income. The opportunity cost of not doing it earlier is probably $6,000-7,000 in revenue I left on the table. The irony is that the affiliate income doesn't replace any of my other streams. It sits on top of them. My freelance income is the same. My SaaS is the same. My YouTube is the same. I just added another line in the spreadsheet. --- # # Should You Add This to Your Stack? (Honest Assessment) Let me be straight with you. Affiliate income is not for everyone. If you don't have an audience — no blog, no YouTube, no Twitter following, no newsletter — affiliate marketing is going to be a slog because you have no traffic to send to your links. Build an audience first, then add affiliate links. But if you already create content as a developer — tutorials, documentation, technical blog posts, You