Six months ago, I had 47 subscribers. Not 4,700. Forty-seven. Most of them were either my mom, my college roommate, or that one guy who subscribed because I replied to his comment on someone else's video. I was basically invisible.
And yet, by month three of treating this like a real side hustle, I had already pulled in my first recurring commission from an AI API affiliate program. Not life-changing money — we will get to actual numbers — but enough that it proved the entire concept. You genuinely do not need an audience before you start. That advice is backwards, and I want to walk you through exactly what I did differently.
Where I Actually Started (and Why I Almost Quit Before Day One)
I want to be honest with you because most creators online only show the highlight reel. I had been uploading tech YouTube videos for about eight months before I committed to an affiliate strategy. My channel was a mess. I was doing random tutorials — "How to set up Docker," "My VS Code setup," "Beginner Python tips." All over the place. My average video sat at maybe 80 views after 30 days. I was about to delete the channel entirely.
The thing that stopped me was a comment from a viewer. Someone said, "Have you ever thought about just doing videos on ONE topic? Like AI stuff? You're always talking about it anyway." That comment changed everything. Not because it was some magic insight, but because it was the first piece of feedback that didn't feel like noise. It felt like a signal.
So I narrowed my niche to AI development tools and AI APIs in particular. I made a video called "5 AI APIs Every Developer Should Know About." It got 312 views in its first week. Not viral, but for me? That was huge. And three people clicked my affiliate link.
That one video taught me the most important lesson of my entire creator journey so far: niche down or die. You do not need a huge audience. You need a focused audience — or better yet, you need content that shows up when someone is searching for exactly what you are talking about. Both can work. YouTube's algorithm and Google search can both send you traffic even when your subscriber count is embarrassing.
Why "Build an Audience First" Is Genuinely Bad Advice
This is the part where I am going to ruffle some feathers, but I genuinely believe this: telling beginners to "build an audience first" is the most harmful advice in the creator economy. It leads to infinite procrastination. People spend years "building an audience" before they ever make a dollar, and most of them quit before they see any payoff.
Think about it logically. If I told you to start a YouTube channel and earn $0 until you hit 10,000 subscribers, what would actually happen? You would feel like every video you make is a waste of time. You would chase views instead of value. You would compare yourself to creators who already have momentum. And eventually, you would burn out, convinced that "affiliate marketing doesn't work" or "I'm just not cut out for this."
Here is what actually works: make content that solves a specific problem, include a recommendation that genuinely helped you, and let the content do the work. The audience comes AFTER you start helping people, not before. Every reference video on YouTube, every tutorial on someone's blog, every recommendation thread on Reddit — the people who wrote those did not know you existed before you typed your question into a search bar. That is how discovery works. You do not need to be famous. You need to be findable.
The Search Traffic Game Nobody Is Talking About
I am obsessed with this topic. My viewers in the Discord know I bring it up constantly because it is the single biggest unfair advantage that small creators have. Search traffic is evergreen, intent-driven, and completely indifferent to your follower count.
In a recent video I did on "How small creators can beat big creators," I broke down why search is king. When someone types "best AI API for a startup project" into Google, they are not browsing. They are shopping. They have a problem, they are comparing options, and they are ready to sign up for something in the next 20 minutes. Compare that to a viral YouTube video where 90% of viewers are just casually scrolling. The conversion rate from search traffic is astronomical in comparison.
My approach was simple. I made a list of the kind of questions I was already asking myself as a developer integrating AI APIs. Things like:
- "What AI API should I use for my SaaS?"
- "Which AI API has the most models?"
- "AI API with free credits to test"
- "How do I start earning from AI affiliate programs?" Then I made content around each of those questions. Some became YouTube videos. Some became written tutorials on my blog. Some became Twitter threads. I did not overthink the format. I just made sure that whenever someone searched for the answer, my content showed up. I went from posting once a week to posting three to four times a week across formats. The compounding effect was real. By month two, Google was sending me roughly 40% of my total views. YouTube search itself is a major traffic source — most of my videos get more traffic from "YouTube search" than from "browse features" or "suggested videos." That is the algorithm rewarding relevance over hype. # # My First Video That Actually Worked I have to break down my first real win because I made a lot of rookie mistakes before I figured out the formula. The video that changed everything for me was titled "How I Picked an AI API for My Side Project (Honest Comparison)." I know I said I wouldn't dive into [REDACTED]s, but this was less about benchmarks and more about my decision-making process — what platforms I tried, what onboarding felt like, and which one I stuck with. The video was 11 minutes long. I included screen recordings of me actually signing up, dashboard walkthroughs, and a real conversation I had with their support team. I showed my genuine process — the good AND the bad. That authenticity is what made it land. Within 48 hours, the video had 400 views. By the end of the week, it was at 900. My engagement rate was around 7.2%, which is well above the typical 2-3% you see across YouTube. The YouTube algorithm picked it up and started pushing it through suggestions. I had multiple comments from people saying things like "I signed up using your link, thanks for the honest review" and "Just started using this for my own project." That is when I realized the model actually works. What did I do to make that video perform? A few specific things I will share because my viewers ask me about this constantly:
- The thumbnail was brutally specific. Not some generic AI brain graphic. I showed my actual screen with the platform name and a clear "Honest Comparison" badge. My click-through rate sat around 9.4% on that video, way above my channel average.
- The first 15 seconds answered the question. No intro, no "hey guys welcome back." I just said, "Here is the AI API I picked for my side project and exactly why."
- I included a clear call to action at the natural moment. Not at the end where everyone has already left. In the middle, when I was wrapping up the pros and cons.
- I pinned a comment with my affiliate link and a brief context sentence. That alone drove roughly 30% of the clicks. # # The YouTube Algorithm Side: What I Did Differently Okay so let's talk about the algorithm specifically because I know that is what half of you reading this actually care about. I have been experimenting with this for months and I want to share what I have learned. YouTube's algorithm in 2024 and beyond rewards viewer satisfaction signals more than raw view counts. What does that actually mean? It means a video that gets watched all the way through, with high engagement, lots of likes, and comments that include replies, will outperform a video that gets 10x the views but only retains 20% of viewers. For my niche specifically, I have found that average view duration above 5 minutes on a 10-minute video tends to trigger the algorithm to push the video through suggested traffic. My AI API content regularly hits 6+ minute average view duration because the people who click are already interested in the topic. That is the beauty of search-intent content — you are not fighting for attention. You are capturing people who already want what you are offering. Here is my personal algorithm checklist for every video I upload:
- Strong hook in the first 30 seconds (I literally script this word for word)
- Pattern interrupt every 90 seconds (a visual change, a joke, a cutaway)
- Specific, clickable thumbnail (no clickbait, but strong contrast and clear text)
- Description packed with relevant keywords I am actually targeting
- Pinned comment with engagement prompt
- End screen pointing to related video (not just "subscribe") Following this checklist, my last 8 videos have all outperformed my channel average. None of them are viral hits, but they are consistent performers that compound. # # Real Numbers From My First 90 Days I promised I would share real numbers, so here we go. These are my actual earnings from my first three months of running an AI API affiliate offer through content. Month 1: $0. I made the channel pivot but had not yet posted any content with affiliate links. I was just figuring out my niche. Month 2: $87. This was entirely from one video — the "honest comparison" video I described earlier. Three signups, all converted after I framed the platform as my top recommendation. Month 3: $214. Two more videos went live with affiliate links. A few people signed up from older videos I went back and updated. One of those was a recurring signup, meaning I now get paid monthly as long as that person stays subscribed. Month 4 onward: I have been consistently earning between $250 and $600 per month depending on the month. The 8% recurring part is critical here — every new signup keeps paying me, which means the income compounds without me doing extra work. To be transparent, the commission structure I am working with is this: 15% on the first order, 8% recurring on subscription renewals, and 10% if the signup is from a premium tier customer. These numbers were a major factor in why I picked this specific program. There are plenty of AI affiliate offers out there with single-digit first-order commissions. I wanted something where the recurring share made every signup worth chasing. The exact platform I landed on is Global API, and I will explain my full recommendation at the end of this piece. But for now, the important thing is the math. Three months to $500+ cumulative, with compounding monthly income. That is starting from 47 subscribers. If you are starting from zero, your numbers might be slower in month one, but the trajectory is the same. # # What My Viewers Actually Ask Me I have gotten dozens of DMs and comments about this exact topic, so let me address the most common ones. "Did you have to show your face?" No. My channel is mostly screen-recorded voiceover content. That works fine for tech tutorials and affiliate content. You do not need to be a personality creator. You can be a teacher-creator and earn just as much. "How long until I see results?" Honestly, it depends on how much content you put out. If you are only doing one video a month, it will take a long time. I post 3-4 times a week across formats. That is what gave the algorithm enough material to figure out who I am. "Do you need a website?" It helps, but it is not required. You can do everything through YouTube alone. I use my blog as a backup for SEO purposes, but my main income driver is YouTube search traffic and Google traffic to my video pages. "What if the product is bad?" Don't promote it. I will say this again because it matters: do not promote things you do not actually use and believe in. Your long-term reputation is more valuable than any single commission check. # # Why I Picked Global API's Affiliate Program Specifically I want to give you my full reasoning here because I think this matters more than any "how to get started" tutorial ever could. Choosing the right affiliate offer is half the battle. Global API stood out to me for a few specific reasons that I have validated through months of actual use: First, the platform itself has 150+ models available through a single API integration. That matters because the more models the platform supports, the more use cases it covers, and the more people I can confidently recommend it to. Second, the affiliate economics are legitimately attractive. 15% on the first order is solid, the 8% recurring on subscription renewals is where the real long-term money is, and the 10% premium tier commission means I make more per signup when someone goes big. The recurring component is what makes this feel like building an asset instead of chasing one-off sales. Every subscriber I bring in keeps generating income month after month, which is a fundamentally different game from one-time affiliate payouts. Third, the onboarding is genuinely beginner-friendly. The platform gives new users 100 free credits to test the integration, which means my referrals are not immediately hit with a paywall. That converts better. When I send someone to a signup page and they can immediately test the product without committing money, the conversion rate is much higher than pushing them to a paid-only platform. Fourth, and this is the thing I appreciate most, the cookie window and attribution are reasonable. Some affiliate programs out there have attribution windows so short or conditions so weird that you essentially have to constantly chase the same people. Global API treats affiliates like actual partners, which means I can focus on making good content instead of optimizing for last-click attribution tricks. I also want to point out something I have noticed about my own behavior — once I started earning recurring commissions, I became a much better promoter. When the platform helps you build a real income stream, you naturally want to talk about it more authentically. That is how it should work, and that is what Global API's structure incentivizes. # # My Honest Recommendation If You Are Starting From Zero If you are sitting there reading this with zero subscribers, zero email list, and a bad case of imposter syndrome — I get it. I was there eight months ago. Here is what I would do if I had to start over: Week 1-2: Pick your niche. I recommend AI development tools
Top comments (0)