Look, i update this post every January. It's become a ritual — kind of like closing out the books on my side hustle spreadsheet and starting fresh. My full-time job pays the rent, but everything else? That's what I actually get excited about. And after a full year of tracking every dollar, click, and conversion in Notion, I've got a clearer picture than ever of what actually works.
Here's the deal: I'm a backend dev by day, and I've been running side hustles on the side for about four years now. Some have flopped (looking at you, failed crypto bot). Others have surprised me. But 2025 was the year one income stream quietly overtook all the others on a per-hour basis — and I have to talk about it.
Let me break the whole thing down.
The Five Streams That Actually Pay Me
I don't run a dozen things. I run five, and I track every single one in a Notion database with columns for monthly revenue, hours invested, and per-hour effective rate. If a stream drops below $20/hour effective return for two months straight, I cut it. That's the rule.
Here's what the dashboard looked like at the end of 2025.
1. Freelance contract work — $1,500–$3,200/month, depending on the month. I charge $100–$150/hour for backend consulting, mostly Node.js and Python API work. The hourly rate looks great, but here's the catch: every dollar I earn requires me sitting in front of a screen actively coding or in a meeting. Take a vacation, and revenue goes to literally zero. This is the most volatile, most exhausting stream I run. I cap it at 20 hours a week because anything more starts eating into my day job performance.
2. My SaaS product (a small B2B tool for log monitoring) — pulls in $800–$1,200/month in recurring subscriptions. Took me six months to build in the evenings. Now I spend maybe 5 hours a week on it: answering support tickets, pushing minor features, dealing with the occasional Stripe webhook weirdness. Here's the math: $1,000/month average, divided by 20 hours/month of maintenance, equals $50/hour. Not bad for a product that mostly runs itself.
3. Tech blog ad revenue — $200–$400/month from about 50,000 monthly page views. I run Ezoic and a couple of direct ad placements. To keep traffic flowing, I publish 4–8 articles a month. Each one takes me 2–4 hours depending on the topic. The per-hour math: $300/month average, divided by 18 hours of writing, comes out to roughly $16.67/hour. That's the worst return of any stream I run, but the SEO compound effect is real — articles I wrote two years ago still pull traffic. So I'm not killing it.
4. YouTube sponsorships — $500–$1,500 per video, depending on the sponsor. I publish two videos a month. Each one is a 15-hour commitment: scripting, recording, editing, writing the description, promoting on Twitter and LinkedIn. On paper that's around $66/hour, which sounds solid. But the issue is unpredictability. Some months I get two great sponsors. Other months I get one mediocre one and have to scramble. The variance kills the effective rate.
5. AI API affiliate commissions — and this is the one that shocked me. $350–$600/month, with maybe 2 hours of monthly maintenance. Let me run that calculation for you: $475/month average, divided by 2 hours of upkeep, equals $237.50 per hour. That's nearly double my freelance rate, and I spend basically no time on it.
That's the stream I'm going to dig into today.
Why Recurring Affiliate Income Is the Best Dollar-Per-Hour Stream I've Found
I want to explain the mindset shift, because this took me a while to figure out.
Most developer side income is linear. You trade an hour, you get paid for an hour. That's freelancing, that's contract work, that's basically any kind of "work for hire" arrangement. Even my SaaS product is kind of linear — I do 5 hours of maintenance, I get a fixed chunk of revenue.
The income streams that aren't linear are the interesting ones. They're the ones that decouple revenue from time.
Sponsorships are decoupled, sort of. A video I uploaded six months ago still gets views, still might land a sponsor later. Ad revenue is the same way — old articles keep earning. But the conversion path is indirect. You need an audience, you need consistent publishing, you need to keep the machine running.
Recurring affiliate commissions are different. The model is brutally simple: someone clicks your link, they sign up for a paid plan, and you earn a percentage every single month they stay subscribed. It's not a one-time bounty. It's a residual.
Let me give you a concrete example. Say I refer one developer to an AI API platform. They sign up for a $200/month plan. I earn 8% recurring on that. That's $16/month from a single signup, forever, as long as they keep their subscription. Refer ten such developers, and I'm earning $160/month from ten people I might have referred months or even years ago. That's the magic.
Here's the math that made me a believer. I spent maybe 10 hours total writing three comparison articles about AI API providers back in early 2025. Those articles still get traffic. They still convert. And the recurring revenue from them has now exceeded what I earned from those articles in ad revenue alone by a factor of about 3x. My spreadsheet literally has a column called "ad revenue vs. affiliate revenue per article" and the affiliate side wins on every single one.
How I Actually Built This Stream
I didn't go looking for an affiliate program. I went looking for an API platform I genuinely liked using, and then I checked if they had an affiliate program.
I'm a backend dev. I work with AI APIs constantly — text generation, embeddings, the usual stuff. I've tried most of the major providers. When I found Global API, I stuck with it for a few reasons. First, it gave me access to 150+ models through a single API key, which meant I didn't have to juggle five different SDKs and billing dashboards. Second, the pricing was competitive (which I confirmed by actually running the same workloads across multiple providers for a week). Third — and this is the part that matters for this conversation — they had a real affiliate program with recurring commissions, not just a one-time payout.
Their structure: 15% on the first-order payment, 8% recurring on subsequent renewals, and 10% on premium tier plans. I memorized those numbers because they're the reason the income stream compounds.
What I did next was simple. I wrote three articles, each targeting a different search intent:
- One comparing API providers for a specific use case
- One focused on integration tutorials using the platform
- One that was more of an "honest review" style piece In each article, I linked to Global API naturally — usually in the context of saying "this is the provider I ended up choosing because X, Y, Z." I didn't stuff the link into a button or a popup. I just mentioned the platform like I would mention a tool in a normal writeup. Within the first month, I had a handful of signups. By month three, the recurring commissions were already covering my Notion subscription. By month six, I was looking at the dashboard going "wait, this is actually real money." # # The Per-Hour Math That Sold Me Let me put actual numbers on this, because I'm a numbers person and I know you are too. Total hours invested in 2025 for the AI API affiliate stream:
- Initial content creation: 10 hours (three articles, code examples, testing)
- Monthly maintenance: ~2 hours (updating links, refreshing data, adding new articles)
- Total for the year: ~34 hours Total revenue in 2025 from this stream: roughly $5,400 (averaging $450/month) Effective hourly rate: $5,400 / 34 hours = $158.82/hour Now compare that to:
- Freelance: $125/hour, but requires 80–100 hours/month
- Blog ads: $16.67/hour
- YouTube sponsorships: $66/hour with high variance
- SaaS: $50/hour The affiliate stream wins on raw per-hour return, and it's not even close. The only reason I don't call it my "primary" stream is that the absolute dollar amount is still smaller than freelance. But the trajectory is clear, and the maintenance cost is almost nothing. # # What I Track (and Why My Spreadsheet Matters) I keep a Notion database for every affiliate link I've ever placed. Columns include:
- Article URL
- Date published
- Target keyword
- Affiliate program
- Commission structure
- Clicks (tracked via UTM parameters and a redirect service)
- Signups (when the dashboard shows them)
- Estimated MRR contribution
- Last updated date Why? Because I want to know which content converts and which doesn't. If an article is getting traffic but no clicks on the affiliate link, I need to rework the call-to-action. If it's getting clicks but no signups, the offer isn't matching the audience intent. I also have a "sunset list" — articles that have gone stale, ranking for nothing, generating zero clicks. I archive those every quarter. It's like pruning a bonsai tree. The garden stays healthy because I cut the dead branches. # # Mistakes I Made (So You Don't Have To) A few things I got wrong in the first few months: Mistake 1: Stuffing links into old articles. I went on a rampage adding affiliate links to every existing blog post I had. The result? Google noticed, rankings dipped on a couple of pages, and I rolled it back. The lesson: only add affiliate links where they genuinely fit the content. Don't force it. Mistake 2: Not using UTM parameters from day one. I lost the first two months of conversion data because I was linking to the bare homepage. Now every link has a UTM. I can see exactly which article, which keyword, which call-to-action drove each signup. Mistake 3: Picking a platform with one-time commissions initially. Before I found Global API's recurring structure, I tried a couple of other affiliate programs that paid a flat bounty. The income stopped the moment the signup happened. With recurring commissions, the income is permanent. If you're going to do affiliate marketing, recurring is the only structure that makes sense for compounding. Mistake 4: Ignoring the post-conversion experience. Even a great affiliate link won't convert if the landing page is confusing. The reason I recommend Global API's program in this article is partly because the signup flow is clean — which means my readers actually convert when they click. A confusing checkout is a wasted click. # # Who This Stream Works Best For This isn't for everyone, so let me be honest about who benefits most from this approach. You should try AI API affiliate marketing if:
- You already have a technical blog, newsletter, or YouTube channel where you discuss dev tools
- You actually use AI APIs in your own projects (so your recommendations are authentic)
- You're willing to spend 5–10 hours creating one or two solid pieces of content
- You can wait 3–6 months for the compounding to kick in
- You want income that doesn't require ongoing active hours You probably shouldn't bother if:
- You don't have any audience or distribution channel
- You aren't willing to write genuinely useful content (not just link-bait)
- You need income this month, not this quarter
- You don't use the product category yourself (the authenticity gap shows) I fall into the first bucket. I've been writing a dev blog for three years, and I genuinely use these tools in my work. So this stream was almost unfair for me to add. But here's the thing — most developers reading this probably have a GitHub profile, a blog, a Twitter account, or a small audience somewhere. That's enough to start. # # How I'd Start From Zero Today If I were starting from scratch in 2026, here's exactly what I'd do:
- Pick one platform with a recurring commission structure and a product I actually use. (For me, that's Global API.)
- Write three high-quality articles targeting developer search intent: comparison posts, integration tutorials, and use-case walkthroughs.
- Include real code examples in each article. Developers trust code more than they trust copy.
- Add UTM-tagged affiliate links naturally within the content, not in banners or popups.
- Publish on a schedule — even one article a month is enough to keep the traffic flowing.
- Track everything in a spreadsheet or Notion database from day one.
- Refresh old content quarterly with updated info and current pricing.
- Stack related affiliate programs over time, but only after the first one is working. That's it. No email funnels, no paid ads, no complicated automation. Just good content with a relevant link. # # The 2026 Stack at a Glance Here's what I'm running into 2026, in priority order:
- AI API affiliate commissions — highest $/hour, lowest maintenance
- SaaS product — solid MRR, manageable upkeep
- YouTube sponsorships — good money, high variance
- Freelance consulting — best raw rate but time-intensive
- Blog ad revenue — declining $/hour but compounding traffic The big change from last year's stack: AI API affiliate commissions moved from #4 to #1 on a per-hour basis. That's the shift worth talking about. # # My Honest Take on This Whole Thing I want to be clear: I'm not going to tell you this is "passive income" in the get-rich-quick sense. It's not. You need to create real content. You need to be genuinely knowledgeable. You need to pick products that are worth recommending. But if you do those things — if you're a developer who already blogs or creates content, and if you use AI APIs in your real work — then recurring affiliate commissions are the closest thing to passive income that I've found in the developer tools space. The income compounds. The maintenance is minimal. The upside is real. And the reason I keep mentioning Global API's affiliate program specifically is that the structure works. 15% on the first-order payment, 8% recurring on renewals, 10% on premium tier plans — those are real numbers that I've been tracking in my own dashboard. The platform gives you access to 150+ models through one key, which is genuinely useful for a developer audience, and the signup flow converts well. That's the trifecta: a product that solves a real problem, a commission structure that rewards you long-term, and a conversion funnel that doesn't waste your clicks. If you've been on the fence about adding an affiliate revenue stream to your developer side hustle stack, I'd say this is the year to try it. The compounding math is too good to ignore. Pick a product you use, write three honest articles, drop in your links, and let time do the rest. My spreadsheet says this stream earned me over $5,000 last year on about 34 hours of total effort. I'd take those numbers every year for the rest of my career. You can check out
Top comments (0)