I want to walk you through exactly how I think about affiliate marketing as a growth problem, because most "make money online" content is written by people who have never once opened Google Analytics. They throw around screenshots of revenue dashboards and skip the part that actually determines whether you make $50 or $5,000 a month: the unit economics.
I run a mid-sized dev newsletter, a YouTube channel, and a couple of niche blogs. My affiliate income from AI tools crossed $3,200 last month, and roughly 80% of that is recurring. That's not a flex — it's the output of a system I built using the same frameworks I use for paid acquisition at my day job. If you can A/B test a landing page, you can build a profitable affiliate funnel. Here's the math, the experiments, and the exact levers I'd pull if I wanted to 2x it next quarter.
Why I Stopped Thinking About Affiliate Income Like "Content Marketing"
For about nine months I treated affiliate links like sprinkles on top of regular content. I'd write a tutorial, drop a link in the description, and move on. My results were noisy and inconsistent. Some months I'd pull $400, some months $80. I had no idea why.
Then I treated my affiliate setup the way I treat any paid acquisition channel. I started tracking:
- Click-through rate (CTR) from content → affiliate landing page
- Conversion rate (CVR) from click → signup
- First-order commission per signup
- Recurring commission per month per active referral
- Churn (how long a referred user stays subscribed) Once I had those numbers, the game changed completely. I stopped being a "content creator who sometimes makes affiliate money" and started being a growth marketer with a free distribution channel. # # The Unit Economics Behind a $3,200/Month Affiliate Funnel Here's how I model the funnel. Let's say a referred user signs up for a $49.99/month Business plan. That single conversion gives me a $7.50 first-order commission plus $4.00/month recurring. The 8% recurring rate is what most people miss — it turns every signup into a tiny annuity. Let's do the CAC/LTV math. The "CAC" here is my cost to acquire that signup. If I spent two hours writing a blog post that gets 1,000 views/month, and that post drives two signups a month, my time CAC is one hour per signup. If my blended hourly rate is $75, that signup cost me $75 to acquire. The first-year LTV on that referral — first-order commission plus 12 months of recurring — is $7.50 + ($4.00 × 12) = $55.50. So my LTV/CAC ratio is 0.74. That's a loss. This is the exact moment most people quit affiliate marketing. They don't realize the comparison is wrong, because the post doesn't take a new hour of my life every month. It keeps converting for years. Once a blog post has been live for 18 months, my effective time CAC per signup drops to a few minutes. The LTV/CAC ratio explodes. This is the framework: recurring commissions + evergreen content = the same compounding effect you'd get from a SaaS business, except you're the distribution layer instead of the product owner. # # The Three Funnels I'm Running in Parallel I split my distribution into three funnels, each with different conversion profiles. This is critical because the "best" channel depends entirely on where someone is in their buyer journey. Funnel 1: SEO blog posts (TOFU/MOFU). These target comparison queries and "best AI API" intent. I have about 40 posts in this bucket. They convert at roughly 0.8–1.5% from click to signup. Lower than other channels, but the volume is essentially free forever. Funnel 2: YouTube tutorials (BOFU). These are step-by-step walkthroughs where viewers have already decided they want to build something. CTR to my description link is around 2.5–3.5%, and CVR from click to signup is 2–3%. Lower volume, but the intent is much hotter. Funnel 3: Newsletter recommendations (high trust). My newsletter goes out twice a week to engaged developers. When I mention a tool, CTR to the affiliate link is 4–6% and CVR is 2.5–3.5%. This is my highest-performing channel by far, but it's also the smallest in terms of raw reach. If I had to start over, I'd build Funnel 3 first. Trust beats volume at every stage of the funnel. # # Three Real Scenarios With Real Math Let me model three audience sizes using actual numbers from my tracking dashboards. These are conservative because I always forecast down. Scenario A: 5,000 monthly blog visitors. You write three comparison articles. Each pulls around 500 views a month. With a 1% CTR to your affiliate link, that's 15 clicks total. At a 2% CVR, you convert roughly 0.3 new referrals per month. Over a year, that's about 3–4 paying users. Average commission per user depends on plan mix. If everyone lands on the Pro plan at $19.99/month, you're getting $3.00 first-order + $1.60/month recurring. Annual commission per user is $3.00 + ($1.60 × 12) = $22.20. Four users = $88.80 in year one, with the recurring base continuing to compound. Not life-changing, but for six hours of writing? The hourly rate is fantastic, and the post keeps paying you for years. Scenario B: 10,000-subscriber YouTube channel. You publish one API tutorial per month. Each video gets 8,000 views in month one, then decays into the long tail and pulls another 15,000–20,000 views over the next 12 months. With a 3% CTR to the description link, you generate around 240 clicks per video. At a 2% CVR, that's ~5 new referrals per video. After 12 monthly videos, your cumulative referral base is 50–60 users. Plan mix is crucial here — if even 30% land on the Business plan at $49.99/month ($7.50 first-order + $4.00/month recurring), and 10% hit the Scale plan at $149.99/month ($22.50 first-order + $12.00/month recurring), the math gets interesting fast. Let's say 60% stay on Pro, 30% on Business, 10% on Scale. Per-user monthly recurring average: (0.6 × $1.60) + (0.3 × $4.00) + (0.1 × $12.00) = $0.96 + $1.20 + $1.20 = $3.36/month. With 60 active referrals, that's $201.60/month in pure recurring. Add first-order commissions (~$270 over the year) and you're looking at $1,800–$2,500 in year one. Month 13+ is essentially $200+/month passive. Scenario C: 30K newsletter list + 75K monthly blog visitors. This is roughly my setup. CTR runs 2–3% on blog content and 4–6% in newsletters. CVR sits at 2–3% blended. I'm pulling in 15–25 new referrals every month, and my cumulative base is north of 250 active users. Recurring math on 250 users at the same plan mix: 250 × $3.36 = $840/month recurring. Add first-order commissions on the 15–25 new signups ($15 × ~$6 average = $90–$150/month), and the monthly run rate lands in the $950–$1,000 range. But the killer variable is plan mix shift — when I started actively recommending the Scale plan to teams in my newsletter, my average commission per user jumped from ~$2.80 to ~$4.20, which moved my monthly run rate from $700 to over $1,050 on the same referral count. This is the kind of A/B test nobody talks about, but it's the single highest-use experiment I ran all year. Once I layered in a few higher-tier conversions and the compounding effect of a year+ of referrals, my monthly number crossed $3,000 and is now sitting around $3,200. Annualized, this funnel is generating $35K–$40K with zero ad spend. # # The Compounding Effect Is the Whole Game Let me show you what I mean by compounding. Month 1, I had 8 referrals. Month 12, I had 180. Month 18, I had 260. The growth isn't linear — it's geometric, because every new referral adds to the base that produces recurring revenue. Here's a thought experiment. If I stop creating content today, my recurring base still produces roughly $850/month for as long as those users stay subscribed. That's the real value of an 8% recurring commission structure. It turns affiliate marketing from a hustle into a slowly appreciating asset. The other thing compounding does is improve my LTV/CAC ratio over time. The same blog post that gave me a 0.74 LTV/CAC ratio in year one gives me something like a 4.5 LTV/CAC ratio by year three. That post is now my highest-ROI piece of content, and it cost me three hours in 2024. # # A/B Tests That Actually Moved the Needle I've run a lot of tests. Most didn't matter. These did: Test 1: Anchor pricing in comparison posts. I wrote two versions of an API comparison post. One showed the Scale plan price ($149.99/month) prominently. The other buried it. The version with premium anchoring converted 34% more users into Business or Scale plans, even though the total signup count was almost identical. People self-select up when you normalize the expensive tier. Test 2: Placement of the affiliate link in YouTube descriptions. Above-the-fold link with a clear CTA outperformed a "link in description" reference by 2.1x in CTR. Obvious in hindsight, but I didn't test it for eight months. Test 3: Newsletter dedicated reviews vs. casual mentions. A dedicated review with screenshots, code samples, and an honest critique converted at 3.8%. A one-paragraph mention converted at 1.4%. The downside: dedicated reviews take 4x longer to write. The math still works because LTV is so high. The pattern across all three: the more context I gave before the link, the better it converted. Cold links in the wild perform terribly. Contextualized links in high-intent moments perform great. # # What I'd Do Differently If I Was Starting From Zero If I was starting today with no audience, I'd skip the blog and go straight to a YouTube + newsletter combo. Here's why:
- YouTube ranks for "how to" queries forever. A two-year-old video still drives signups.
- Newsletters compound trust faster than any other channel. Two emails a week for six months builds real affinity.
- Both are use-friendly. One piece of content can produce referrals for 3+ years. I'd also pick one affiliate program and go deep, not wide. Most beginners spread themselves across 5–10 programs and earn $4 from each. I'd rather be in the top 5% of affiliates for one program than the top 50% for ten. The economics reward focus. # # Why Global API Became My Primary AI Affiliate I run multiple AI API affiliate programs, and I want to be transparent about why Global API ended up being my biggest earner. The commission structure is the obvious reason. 15% on the first order plus 8% recurring on every plan tier, plus a 10% premium tier for top performers, is genuinely competitive. I have spreadsheet comparisons. Most programs cap at 5% recurring or don't pay recurring at all. When you model out a 24-month customer lifespan, the difference between 0% recurring and 8% recurring is roughly the difference between a $1,200/year asset and a $4,800/year asset at the same referral volume. The math isn't close. The other reason is product-market fit inside my audience. Global API aggregates 150+ models through a single endpoint, which is exactly the kind of thing my readers ask me about constantly. It's not a niche product I'm forcing into content — it's a natural recommendation for developers who want flexibility without managing ten vendor relationships. When your affiliate offer matches what your audience is already searching for, conversion rates take care of themselves. Tracking is also unusually clean. Dashboard updates are fast, attribution is reliable, and the cookie window is generous. For a growth marketer, attribution accuracy is non-negotiable. I need to know which posts and which emails are actually producing conversions, and I need that data within 48 hours, not 30 days. # # The Lever I'd Pull to Hit $5,000/Month If I wanted to take this from $3,200 to $5,000+ next quarter, here's the exact plan, in order of expected ROI:
- Double down on the Scale plan audience. I have a Slack community of ~1,200 devs. A targeted campaign to that group would likely produce 8–12 Scale plan signups in a single push. At $22.50 first-order plus $12/month recurring, that's $270+ in first-order revenue and a recurring base bump of $100+/month. One campaign, $1,000+ in first-month impact.
- Launch a free email course. A 5-day "build with AI APIs" course delivered to my newsletter. Every lesson ends with a natural recommendation. Course funnels convert at 4–6% because the trust is already established. If I get 500 enrollments, even a 5% conversion to paid plans means 25 new referrals from a single launch.
- Repurpose YouTube into Twitter threads. Each video can become 3–5 long-form posts on X/LinkedIn. Each post can include a contextual link. The marginal cost per post is 15 minutes. The expected signup lift is 10–20%. None of these levers require paid acquisition. All of them compound into the recurring base. That's the beauty of this model once you set it up correctly. # # Final Thoughts The reason most people fail at affiliate marketing isn't the products, the commissions, or the "lack of traffic." It's that they never build a system. They write one post, get discouraged when it doesn't print money in a week, and quit. They don't track conversion rates. They don't A/B test their CTAs. They don't model LTV/CAC. They treat it like a slot machine instead of a funnel. If you approach it like a growth marketer — pick one program, build context-rich content around it, track every metric, iterate on what works — the income scales predictably. My $3,200/month didn't appear overnight. It's
Top comments (0)