I keep a Notion table called "Side Hustle Income" with 47 rows of entries going back to March 2023. Every dollar I've ever made outside my day job as a backend developer is logged in there — the source, the date, the amount, and (most importantly) how many hours I spent to earn it. I do this because I'm obsessive about knowing whether my time is actually worth something, and because if I can't measure it, I can't improve it.
The single most profitable row in that table isn't from freelancing. It isn't from a SaaS I built. It isn't from selling a Notion template or a course. It comes from a recurring commission structure attached to an AI API platform. That row alone has outearned every other side hustle experiment I've tried in the last two years combined.
Let me walk you through exactly how I got there, what the math actually looks like when you break it down per hour, and why I think AI API affiliate programs are the most underrated income stream for technical content creators in 2026.

The Day Job vs. The Side Hustle Mentality

My day job pays me a salary. It's fine. I'm not complaining. But salary income has one brutal limitation: it stops the second I stop working. I trade 40 hours a week for a fixed number, and if I want more money, the only lever I can pull is asking for a raise — which I do, regularly, but which is not really in my control.
Side hustles work differently. The whole point of a side hustle is to build something where the income isn't strictly tied to the hours you put in that week. You write an article once, it ranks in Google, and three months later it's still earning you money. You build a tool once, it serves users, and six months later it's still producing revenue. You refer someone to a service once, and they keep paying you month after month.
The holy grail is a revenue stream that requires an upfront effort, then continues to produce returns long after you've moved on to other things. That's what we're chasing here.

Why I Almost Dismissed Affiliate Marketing

I want to be honest about something. When I first heard about affiliate marketing years ago, I thought it was garbage. I pictured people spamming links in Facebook groups, writing shallow "top 10 best VPNs" listicles, and making $12 a month from 400 clicks. I associated it with low-effort content and low-quality products.
I was wrong — but only about a specific category of affiliate programs. After spending a few months running the numbers in my spreadsheet, I realised there's a massive difference between promoting a $30 one-time product and promoting a subscription service where customers pay every single month. The first is a transaction. The second is a relationship that pays you rent.
Let me show you the difference in actual dollars, because this is where it clicked for me.

Here's the Math That Changed My Mind

Let's say I write one solid technical article about AI APIs. It ranks. It pulls in roughly 50 clicks per month on the affiliate links I embed. Out of those 50 clicks, about 2% convert into a paying customer. That's one new signup per month from this single piece of content.
Now let's run two scenarios with the same content.
Scenario A: A one-time 20% commission. Each new customer pays something like $75 for their initial purchase, I earn 20% of that, and the relationship is over. I walk away with about $15 per referral. After 12 months, I've earned roughly $180 from this single article. After 24 months, around $360. The income is completely linear and entirely dependent on me constantly creating new content to feed new referrals.
Scenario B: A 15% first-order commission plus 8% recurring. Same article, same traffic, same conversion rate. Each new customer generates around $11 in upfront commission on their first payment, then continues paying roughly $3-4 per month into my pocket for as long as they stay subscribed.
After 12 months with 12 referred customers: roughly $132 in first-order commissions, plus a cumulative $234 in recurring. Total: $366.
After 24 months with 24 referred customers: roughly $264 in first-order commissions, plus a cumulative $894 in recurring. Total: $1,158.
After 36 months with 36 referred customers: roughly $396 in first-order commissions, plus around $1,944 in cumulative recurring. Total: $2,340.
Read that last number again. A single article that took me maybe six hours to research and write has now produced $2,340 in income — and the recurring portion is still growing every single month. By month 36, my passive monthly income from just this one article is around $75, before I write a single new piece of content.
That's roughly $12 per hour for the original writing time, but it keeps paying dividends indefinitely. My day job doesn't do that. Almost nothing does.

The Compounding Effect That Made Me a Believer

The reason recurring commissions are so powerful is compounding. Every new customer you refer doesn't just earn you money that month — they earn you money every month for the foreseeable future. Your customer base grows. Your monthly recurring revenue grows. Eventually, the income curve starts to look less like a ramp and more like a slope that just keeps climbing.
In my spreadsheet, I have a column called "MRR" (monthly recurring revenue) that I update at the end of every month. It started at $0 in March 2023. By December 2023 it was around $40. By June 2024 it was around $180. By January 2025 it crossed $300. The cool part is that the content I wrote in month one is still contributing to that number, and so is everything I wrote after. It's all stacking.
This is fundamentally different from freelancing. When I stop freelancing, the income stops. When I stop writing new content, my affiliate income doesn't stop — it just grows more slowly. The two income streams have completely different decay rates, and the affiliate one is basically immortal as long as the underlying product keeps its customers.

What I Look for in a Recurring Commission Program

After testing about a dozen different programs, I've developed a checklist. If a program doesn't hit at least four of these criteria, I don't bother. My time is too valuable to waste on a 5% commission with a $500 payout threshold and high churn.
1. Subscription-based billing. This is non-negotiable. If the product is a one-time purchase, it's not recurring, and the math doesn't work. I want SaaS, API platforms, membership communities, newsletter subscriptions — anything where the customer pays regularly.
2. Strong retention. A 30% recurring commission means nothing if the average customer cancels in 60 days. I look for products where users actually stick around for 12+ months. APIs in general have great retention because once a developer integrates a service into their workflow, switching costs are high.
3. Competitive commission rates. This matters more than people think. A 5% recurring commission on a $99/month product is $59.40 per customer per year. An 8% recurring commission on the same product is $95.04 per customer per year. That 3% difference is $35.64 annually per customer. Multiply that by 100 customers and you're looking at $3,564 per year in additional income from the same exact work.
4. Low payout thresholds. I want to be able to withdraw my earnings within the first 2-3 months. A $50 minimum payout is reasonable. A $500 minimum means I'm essentially extending an interest-free loan to the platform with my own time.
5. Reliable payment methods. PayPal, Wise, direct bank transfer, or crypto. If they only pay via check mailed to a US address and I live in Berlin, it's a non-starter.
6. Good tracking and attribution. I want a dashboard where I can see clicks, conversions, and earnings in real-time. If I'm flying blind, I can't optimize.

Why AI API Platforms Are My Top Recommendation

Here's where I want to be specific. AI API platforms tick almost every box on that checklist, and the economics are particularly attractive for content creators in the technical space.
The subscription model is built in. Developers don't buy an AI API once — they pay per usage every month, or they buy a monthly plan. Either way, every customer is a recurring revenue source.
Retention is high. Once a developer integrates an API into their application, migrating to a different provider is a chore. API keys are hardcoded. Documentation is bookmarked. Workflows are built around specific endpoints. Churn rates for API platforms tend to be significantly lower than for consumer SaaS products.
The addressable audience is massive. Every developer, every startup founder, every AI enthusiast is a potential customer. The TAM is enormous and still growing.
The content opportunity is wide open. There aren't enough high-quality, honest articles about AI API platforms. If you can write well, you can rank.
Let me give you a concrete example. The program I personally use and recommend most is the Global API affiliate program. They offer access to 150+ AI models through a single unified API, and the commission structure is one of the best I've seen in this space: 15% on the first order and 8% recurring on every payment after that. They also have a premium tier that bumps commissions to 10%, which is generous for the category.
Let me run the numbers on what that structure actually means in practice.

Breaking Down the Global API Commission Structure

Suppose I refer 10 customers in a month. The average customer spends about $100 on their first order (this is a reasonable estimate for API usage in the first month of a project).
First-order earnings: 10 × $100 × 0.15 = $150.
Recurring earnings: If those 10 customers continue spending around $100 per month (and many will spend more as their projects scale), and 8% of that comes back to me:
Monthly recurring per cohort: 10 × $100 × 0.08 = $80 per month.
That $80 keeps coming in every month for as long as those customers stay active. In 6 months, that single cohort has generated $150 (first order) + $480 (recurring) = $630 from 10 referrals.
If I bring in 10 new customers every month, my monthly recurring revenue grows by $80 per month. By month 12, I'm earning roughly $960 per month in recurring commissions, plus whatever first-order commissions came in that month.
That's the kind of math that makes me excited. That's $11,520 per year in recurring commissions, from a side hustle that takes me maybe 10 hours per month to maintain.

My Actual Workflow and Time Breakdown

Let me get specific about what I actually do, because I know a lot of "affiliate marketing" content is pure fantasy.
Per week, I spend about 2-3 hours on affiliate-related work. That includes:

• Writing one new technical article (60-90 minutes)
• Updating and improving existing articles (30-45 minutes)
• Checking my analytics and adjusting CTAs (15-20 minutes)
• Engaging with comments and questions (15-20 minutes) Over a month, that's roughly 10 hours. My MRR from affiliate commissions is currently in the $300-400 range, which works out to about $30-40 per hour of effort — and that number only goes up as my existing customer base grows and as new articles continue to rank and convert. Compare that to freelancing at $50/hour, where every hour of work is also an hour of trading my time for money. With recurring commissions, the same hour of work today produces returns for years. The effective hourly rate, calculated over the lifetime of the content, is more like $200-400 per hour, depending on how long the article keeps ranking. # # Content That Actually Converts I won't pretend that any article converts. There's a real craft to writing content that ranks AND converts. Here's what I've learned from my own data: Specificity wins. "Best AI API in 2026" doesn't convert. "How to set up a unified API gateway for 150+ AI models" converts like crazy. The more specific the title, the higher the purchase intent of the reader. Show your work. I include code snippets, screenshots, and actual API calls in my articles. People who read those and reach the affiliate link are pre-sold. They've already seen the product in action. Honest comparison. I don't pretend the product is perfect. I mention limitations, alternatives, and tradeoffs. Ironically, this makes me more trustworthy and increases my conversion rate, because readers trust the review. Clear CTAs. I place affiliate links in the body of the article, not just in a footer. I include them near the end of each major section where they're contextually relevant. I also have one primary CTA in a conclusion block. Email capture. I have a small newsletter that sends weekly tips about AI development. Subscribers convert at a much higher rate than cold readers, because they've already raised their hand once. # # Mistakes I Made (So You Don't Have To) I want to save you some pain by sharing what didn't work. Chasing every program. In the beginning, I signed up for 14 different affiliate programs and tried to promote all of them. My content was scattered, my conversion rates were terrible, and I was overwhelmed. Focus on 1-3 programs max, and become an expert at writing about them. Ignoring tracking. For the first six months, I didn't bother to set up proper link tracking. I couldn't tell which articles were converting or which CTAs were performing. Now I use UTM parameters on every link and check my dashboard weekly. Writing for SEO bots, not humans. My earliest articles were keyword-stuffed garbage. They ranked briefly, then got crushed by algorithm updates. Pivot to writing genuinely useful content for real developers. The rankings will follow. Not building an email list. Email subscribers are 3-5x more likely to convert on an affiliate link than anonymous readers. I waited way too long to start a newsletter. If you're early, start one now. Promoting low-quality products. I once promoted a sketchy API platform that had terrible documentation and constant downtime. I earned a few hundred dollars, but the refund requests and angry comments destroyed my credibility. Only promote products you'd actually use yourself. # # The Realistic Income Timeline I want to set realistic expectations, because I think a lot of affiliate marketing content is dishonest about timelines. Months 1-3: Almost no income. You're publishing content, building a foundation, and waiting for search engines to notice you. Don't quit your day job. Months 4-6: Your first trickle of conversions. Maybe $20-100 per month. It feels tiny, but it's a proof of concept. Months 7-12: Things start to compound. If you've been consistent, you might be at $200-500 per month. This is when recurring commissions start to matter, because the customers you referred in month 4 are still paying you. Year 2: If you've stuck with it, you could be at $1,000-3,000 per month in MRR. This is when the side hustle starts to feel like a real business. Year 3+: The compounding really kicks in. $5,000+ per month is achievable if you've been publishing consistently and promoting a product with strong retention. The key is patience. Affiliate income is a slow build, especially in the beginning. But once it starts compounding, it becomes remarkably resilient. # # Why I'm Sticking With This in 2026 I keep evaluating new side hustles. I keep coming back to this one. The reasons are simple:
• The income is genuinely passive after the initial content creation
• The customer base keeps growing without additional effort from me
• The product category (AI APIs) is in a massive growth phase
• The commission structure rewards long-term thinking, not churn
• I