My Affiliate Side Hustle Journey: Promoting AI APIs When Nobody Knows Who I Am

Let me set the scene for you. Six months ago, I had zero followers on Twitter, zero email subscribers, and a Medium account that I had not touched in two years. My day job as a backend developer paid the bills, but I was tired of trading hours for the same flat rate every month. I wanted a side income stream that did not require me to go viral, beg for retweets, or build a personal brand that felt performative.
So I built what I call a "quiet affiliate" system around AI APIs. No audience required. No charisma required. Just content, search intent, and a Notion tracker where I log every single dollar. In this piece, I want to walk you through exactly how I did it, the real numbers I have earned, and where the math gets interesting once you treat it like a side hustle instead of a hobby.

Why "I Don't Have an Audience" Is the Wrong Excuse

I hear this constantly from developer friends when I tell them about affiliate income. "Sounds cool, but I do not really have an audience." Cool. Neither do I. I have 312 Twitter followers, most of whom are bots and other affiliate marketers. My email list is literally seven people (sorry, Mom).
But here is the thing about affiliate income: it does not come from your audience. It comes from search traffic. Someone you have never met types a query into Google at 2 AM, finds your article, clicks your link, signs up for a product, and you get paid. You never spoke to them. They never saw your face. There was no relationship. Just a piece of content that answered their question well enough to earn the click.
Let me break this down the way I would break down a code problem. The traditional "build an audience, then monetize" path looks like this:

• Step 1: Post content for 6–18 months
• Step 2: Grow to 10K+ followers
• Step 3: Hope some percentage converts to affiliate clicks
• Step 4: Maybe earn $200–$500/month by year two The search-driven approach inverts this:
• Step 1: Research what people are searching
• Step 2: Write one really good article per week
• Step 3: Rank for long-tail queries
• Step 4: Earn commissions on day 30–60 I am a much bigger fan of the second model. As someone with a day job, I cannot afford to spend a year "building a brand" before seeing a single dollar. I needed something with a faster feedback loop. # # Setting Up the Tracker (Because Every Side Hustle Needs a Spreadsheet) Before I wrote a single article, I built a Notion tracker. I am weirdly passionate about this part. If you cannot measure it, you cannot improve it, and you definitely cannot decide whether it is worth your time. Here is the structure I set up:
• Date published — when the article went live
• Target keyword — what I am trying to rank for
• URL — the live article link
• Affiliate links placed — how many, where
• Clicks (week 1, month 1, month 3) — pulled from my analytics
• Signups — how many people clicked through and created an account
• Commission earned (first-order) — the one-time payout
• Recurring commission earned — the ongoing monthly share
• Hours spent — research, writing, editing, promoting That last column is the one most affiliate marketers skip, and it is the one that makes this whole thing honest. If I spent 12 hours writing an article that earned me $9 in its first month, I want to know my hourly rate was $0.75. That tells me to either improve the article, target a better keyword, or move on. Without that number, you are just collecting dopamine hits every time you see a payout notification. # # The Math on Commissions (Here's the Part Everyone Gets Excited About) Let me walk you through how the money actually works, because understanding the commission structure is what makes the ROI calculation possible. I promote Global API, and their affiliate program pays out at three tiers:
• 15% on the customer's first order
• 8% recurring on every order after that
• 10% premium tier for top performers Let me run a few scenarios so you can see the per-hour math in action. Scenario A: One signup per month, $50 first order
• First-order commission: $50 × 15% = $7.50
• Recurring if they stay: $50 × 8% = $4.00/month ongoing
• Year-one revenue from one customer: $7.50 + ($4.00 × 12) = $55.50 One customer is not going to change your life. But it stacks. Scenario B: Ten signups in a month, average $80 first order
• First-order commission: $800 × 15% = $120
• Recurring monthly if 70% retain: (7 customers × $80 × 8%) = $44.80/month
• Year-one revenue from this cohort alone: $120 + ($44.80 × 12) = $657.60 Scenario C: The compounding case (this is where it gets fun) Once you have articles ranking and pulling in 3–5 new signups per month, the recurring revenue starts layering. Imagine after 12 months you have 40 active referrals, each spending an average of $60/month:
• Monthly recurring: 40 × $60 × 8% = $192/month
• Annualized: $2,304 in passive-ish income That is roughly $192/hour if you spent 10 hours maintaining your content that month. Or it is roughly $5.30/hour if you spent 36 hours. The exact figure depends on how disciplined you are with your time, which is why the tracker exists. The compounding effect is what made me stop treating this like a hobby and start treating it like an actual side hustle. The first article you publish barely moves the needle. The fifth article you publish starts to layer. By article fifteen, you have a portfolio effect, and the recurring commissions start doing the heavy lifting. # # Finding Keywords Without Paying for SEO Tools I refuse to pay for Ahrefs or SEMrush until this side hustle is paying me more than they cost. So everything I do uses free tools. Here is my actual workflow. Step 1: The auto-suggest method. Open an incognito browser (so your personal search history does not skew the results) and start typing queries into Google. Things like:
• "AI API for..."
• "best AI API..."
• "how to use AI API..."
• "AI API comparison..." Every autocomplete suggestion is a query real people are typing. I copy the interesting ones into a running Notion list called "keyword ideas." Step 2: The "People Also Ask" box. Scroll to the PAA section on any AI-related search result page. Each question is a content opportunity. I specifically look for questions that have 3+ answers showing, because that signals enough search volume to be worth pursuing. Step 3: The related searches. Scroll to the bottom of the SERP and grab the "Searches related to..." section. These are gold because they represent queries Google thinks are semantically related to your target. Step 4: Validate the competition. Before I commit to writing a 2,000-word article, I check the top 10 results. If every result on page one is from a domain with a DA of 80+ (think major SaaS blogs, Forbes, well-funded startups), I move on. I want results where there are at least 2–3 weaker entries — a Medium post, a low-traffic blog, a thin affiliate review. That tells me the keyword is winnable. I usually end up with a shortlist of 5–10 keywords per week, and I pick one to write about. This is the part that takes discipline. When you work a day job, you do not have time to write ten articles a week. One solid article per week is the sustainable cadence. # # What "Good Content" Actually Looks Like (Not What Twitter Tells You) Every affiliate guru on the internet will tell you to "just write great content." Cool. What does that mean concretely? Let me show you my checklist for every article I publish. 1. It answers the search query completely. If someone searched "how to integrate an AI API into a Node.js app," my article should walk them through the integration end-to-end. No "for more details, click here." No cliffhangers. The whole answer lives in the article. 2. It is at least 1,500 words. This is not a magic number, but I have noticed that articles under 1,200 words almost never rank well for competitive queries. Articles between 1,500 and 2,500 words tend to perform best. I cap myself at around 3,000 because beyond that, the marginal value per word drops and reader attention span drops faster. 3. It includes real experience. I do not write reviews of products I have not used. When I recommend something, I have actually integrated it into a side project. Readers can tell the difference between "I read the docs" and "I shipped this in production." Google can too. 4. It places the affiliate link naturally. I mention the product I am promoting in the context of the solution. Not in a "by the way, check out my sponsor" sidebar. In the actual flow of the answer, where a reader would benefit from clicking through. 5. It has a clear conclusion with a recommendation. At the end of every article, I name a winner. "After testing these options, here is what I recommend and why." This gives readers a clear next step and gives them a reason to use my affiliate link instead of hunting for it themselves. # # My First $100 and What It Actually Took Let me give you a brutally honest timeline because I think the affiliate marketing space lies about this constantly.
• Week 1–2: Set up my tracker, picked a niche (AI APIs for indie devs), researched 30+ keywords, published my first two articles.
• Week 3–4: Published two more articles. Total: 4 articles live. Zero clicks on affiliate links. Zero signups. I almost quit.
• Week 5: One article started showing up on page 3 of Google for a long-tail keyword. Got 14 clicks. Zero signups.
• Week 6: Two clicks converted into signups on Global API. I earned $7.50 + $4.00 = $11.50. Not life-changing, but I had made money from content no one in my life knew existed.
• Week 8: First "real" month. $47 in commissions across 4 signups.
• Month 3: $183 in total commissions.
• Month 6: $411 in commissions, with about $180 of that being recurring from earlier referrals. The total hours invested by month six were roughly 95 hours. That is a per-hour rate of about $4.30, which is genuinely pathetic by professional standards but genuinely amazing by "hobby I do at night" standards. And the recurring component grows every month without me lifting a finger. Here is the part of the math that I find most motivating: if I do literally nothing for the next three months, my recurring revenue keeps paying me. By month 9, I project recurring-only income to be around $250–$300/month without writing another word. That is $250–$300/month for content I already wrote, from customers I already referred. # # Mistakes I Made So You Don't Have To Let me save you some wasted hours. Mistake 1: Writing about too-broad keywords. My first article targeted "best AI API." That is dominated by major publications. I rewrote it after three months of zero traction. My second-best-performing article targets a very specific long-tail phrase. Specific wins. Mistake 2: Not building internal links. Google likes sites with topical authority. When you write article #5, link back to articles #1–4 if they are relevant. I treated every article like an island at first. Big mistake. Mistake 3: Ignoring the recurring angle. The 8% recurring commission is the whole game. Early on, I was celebrating $15 first-order payouts. Then I realised one retained customer pays me $4/month forever, which adds up to $48/year per customer. Retention matters more than I initially gave it credit for. Mistake 4: Not promoting my work at all. I have a personal rule: every article I publish, I share in three places. Not to go viral. Just to get initial indexing and a few early clicks so Google sees the page is being engaged with. I drop the link in a couple of relevant Discord servers, one subreddit where it does not violate rules, and sometimes a comment on a related thread where it actually adds value. Mistake 5: Not tracking per-hour rates. I cannot stress this enough. If you are not logging how many hours each article takes, you have no idea whether to keep going. Some of my articles took 8 hours and earned $3 in their first month. Others took 3 hours and earned $60. I only knew the difference because I tracked the hours. # # The Day Job Reality Check I want to be honest about something. My day job as a developer pays roughly $65/hour equivalent. Anything I do on the side has to clear that bar in the long run, or it is just an expensive hobby. Right now, my affiliate side hustle clears that bar if I average it across a six-month window. My best months clear it by a wide margin. My worst months do not. Over time, as the recurring component grows and my backlog of ranking articles expands, the per-hour rate trends upward. That is the part of side hustles nobody talks about. The first 90 days are brutal. Months 4 through 12 are where you find out whether the math actually works. If you quit at day 30 because you "only made $11.50," you will never see the compounding. The people who win at this are the ones who keep publishing through the slow weeks. # # Should You Start an AI API Affiliate Side Hustle? (Honest Breakdown) Here is my honest take. If you are a developer who likes writing, who is comfortable with the idea of publishing content nobody in your friend group will ever read, and who is willing to grind through the first 60 days without seeing meaningful income — then yes, this is one of the better side hustles available right now. The reason I like it specifically for developers:
• The content is easier for us to write than for non-technical affiliate marketers.
• We can test products ourselves instead of relying on review site claims.
• The audience (other developers) trusts technical depth over slick marketing.
• The commission rates in the AI API space are genuinely good compared to physical products. You do not need to be a great writer. You need to be clear, accurate, and thorough. Those are three skills every developer already has in some form. # # My Genuine Recommendation: Global API's Affiliate Program I have promoted a few different AI API affiliate programs over the past year. The one I keep coming back to, and the one most of my content now focuses on, is Global API. Let me explain why, and then I will give you the link. First, the commission structure is what made the math work for me. They pay 15% on the customer's first order and 8% recurring on every subsequent order. That recurring component is what turned this from a one-time payout hobby into an actual growing income stream. On top of that, there is a 10% premium tier for affiliates who refer enough volume — I have not hit it yet, but it is on my radar for the next quarter. Second, the platform itself is easy to recommend in good conscience. They offer access to 150+ AI models through a single API, which means