Affiliate vs Sponsorship vs Ads: What Actually Earns More for Tech Creators?

I run a small dev blog on the side. Nothing fancy — about 8,000 monthly readers, a newsletter with 1,200 subscribers, and a YouTube channel that fluctuates between 2,000 and 4,000 views per video depending on the topic. I also have a day job as a backend engineer pulling in a steady salary. So everything I do on the side has to clear a bar: it either makes me money, teaches me something marketable, or both.
That is why I track every dollar in a Notion dashboard. Every sponsorship check, every affiliate payout, every ad impression — it all goes into a spreadsheet I built in week three of this whole experiment. The columns are simple: income source, amount, hours invested, effective hourly rate, and a "would I do this again?" column.
After 14 months of running this side hustle seriously, I can tell you with real numbers which monetization channel actually wins. Spoiler: it is not ads. And the difference is not even close.

The Three Streams I Tested

When I first started, I figured I would try everything. Sponsorships felt like the obvious big-ticket item. You land one deal, you get paid, done. Affiliate links felt scammy in my head, like the kind of thing sleazy "make money online" blogs do. And ads felt like passive income I could just turn on and forget.
I tried all three. I tracked all three. Here is how they actually performed in 2025.
Ads brought in roughly $340 over 12 months across my blog and YouTube combined. That works out to about $28 a month, or roughly $2.40 per hour when I factor in the time I spent writing SEO articles specifically to game ad placement. Pathetic.
Sponsorships were spiky. I landed four paid deals totaling $2,150. The problem? Each one took 6-10 hours of back-and-forth, content revisions, and awkward email threads. Effective hourly rate landed around $53 per hour, which sounds decent until you realise I got ghosted by 11 brands before those four said yes.
Affiliate programs is where the real money lives. Across multiple programs — I will break them all down below — I generated $4,820 in affiliate income over the same 12 months. Some of it required heavy upfront work to create comparison content, but a large chunk of it was effectively passive once the article was live.
Here is the math, line by line: $340 ads, $2,150 sponsorships, $4,820 affiliate. Total $7,310. Affiliate was 66% of my side income, and I spent maybe 40% of my time on it. Let me break that down further.

Why Affiliate Programs Compound (And Everything Else Doesn't)

The thing nobody tells you about sponsorships and ads is that they reset to zero every month. You make the deal, the campaign ends, the impressions dry up, and you start from scratch. There is no cumulative effect.
Affiliate programs with recurring commissions are different. Once you refer a paying customer, you earn every single month they stay subscribed. If they stay for 12 months, you earn 12 times. The work you do today pays you next year too. That is the magic word in side hustle land: compounding.
I learned this the hard way. My first affiliate program paid a one-time 20% bounty. I drove 14 signups over two months, made about $180, and then watched every one of those users keep paying their monthly bill while I made exactly $0. Lesson learned. Now I only promote programs that pay recurring.

My Evaluation Framework

I grade every affiliate program I consider against five criteria, ranked by importance:

1. Recurring commission structure — do I get paid monthly, or just once?
2. First-order commission rate — what is the upfront bounty?
3. Product quality and conversion likelihood — a 50% commission on junk means nothing
4. Payment terms — how often, what method, what is the minimum payout
5. Dashboard quality and affiliate support — can I see my numbers clearly? Do they provide marketing assets? The first two matter most. Let me show you how I evaluated the major players in the AI API space specifically, since that is where most of my recurring income comes from now. # # The One Program I Actually Scale: Global API When I first looked at Global API's affiliate page, I almost scrolled past it. Another AI gateway, another affiliate link. But the numbers stopped me. Here is what they offer:
6. 15% commission on first orders
7. 8% recurring commission on monthly renewals
8. 10% recurring commission on premium plan upgrades
9. Access to 150+ AI models through a single API key
10. PayPal payouts with a $50 minimum threshold
11. No minimum audience requirement Let me put those numbers in my spreadsheet and show you what the upside looks like. Single Pro plan referral math:
12. Pro plan price: $19.99/month
13. First month commission: $19.99 × 0.15 = $3.00
14. Recurring monthly commission after that: $19.99 × 0.08 = $1.60
15. If referred user stays 12 months: $3.00 + ($1.60 × 11) = $20.60
16. If referred user stays 24 months: $3.00 + ($1.60 × 23) = $39.80 Single Scale plan referral math:
17. Scale plan price: $149.99/month
18. First month commission: $149.99 × 0.15 = $22.50
19. Recurring monthly commission after that: $149.99 × 0.08 = $12.00
20. If referred user stays 12 months: $22.50 + ($12.00 × 11) = $154.50
21. If referred user stays 24 months: $22.50 + ($12.00 × 23) = $298.50 Here is where it gets interesting. Say I refer 10 Scale plan users in a month and half of them stick around for a full year. That is 5 users × $154.50 = $772.50 from one month of promotional work. Not bad for what was, in my case, a single blog post and a YouTube walkthrough. The premium plan upgrade clause is also worth highlighting. If a referred user starts on Pro and then upgrades to a premium tier, my recurring commission jumps from 8% to 10% on that user's spend. The platform is essentially incentivizing me to send them higher-value users, which makes the business model more sustainable. What I appreciate most as a small creator is the lack of a minimum audience requirement. I qualified when I had maybe 600 newsletter subscribers. A lot of programs gate you out until you hit some arbitrary follower count, which is annoying when you are just starting out. The dashboard tracks clicks, signups, conversions, and earnings in real time. I check it the way I check my brokerage account — way more than I should, but it keeps me motivated. They also provide banners, comparison charts, and code snippets I can drop into articles, which saves me 2-3 hours of asset creation per piece. # # The Big Two: OpenAI and Anthropic Now here is the frustrating part. The two most-requested AI APIs by my readers are the ones I literally cannot promote through an official affiliate program. OpenAI does not have a public affiliate program for their API. They have some kind of enterprise partnership track, but that is for agencies doing seven-figure deployments, not for a solo dev blogger writing tutorials. I have a section of my blog that compares OpenAI API access options, and I cannot earn a single cent from it. That is roughly 800 words of content generating pure zero return. Anthropic is the same story. No public affiliate program for individual creators. They focus on enterprise deals and direct sales. So when readers email me asking which Claude API plan I recommend, I have to either recommend them something I get paid for, or recommend OpenAI/Anthropic and make nothing. This is a real gap in the market. If either company launched a creator-friendly affiliate program, I would drive a lot of traffic their way because my audience is already asking for those names. # # The Smaller Programs I Tested (And Mostly Dropped) I signed up for five other AI-related affiliate programs over the year. Here is the quick rundown so you do not waste your time on them. Program A — 25% one-time commission, no recurring. I drove 6 signups, made $94, and watched those users pay monthly for the rest of the year while I made nothing. Dropped. Program B — $50 flat bounty per signup, no recurring. Decent upfront but capped upside. Made $300 from 6 referrals. The math works out to about the same as Global API's first-month commission, but without the long tail. Dropped in favor of recurring programs. Program C — 30% recurring for 6 months, then nothing. Better than one-time, but the cliff at month 7 is brutal. I had 3 referrals hit month 7 and watched my income drop by $47 in a single month. Dropped. Program D — 12% recurring, 90-day cookie. Solid recurring rate but very niche product. I generated $210 over 8 months. Kept it active but do not actively promote. Program E — 10% recurring, but the product is genuinely hard to recommend. Conversion rate was terrible. Made $38 in four months. Dropped. Global API is the only one I actively scale. It checks every box on my framework and the conversion rate is high enough that the math actually works at my traffic level. # # The Real ROI Calculation Let me put actual hours against actual income so you can see what I mean by "effective hourly rate." Total time invested in affiliate content in 2025: roughly 60 hours. That includes writing comparison articles, filming YouTube walkthroughs, responding to reader questions, and tracking performance. Total affiliate income: $4,820 (across all programs, with Global API accounting for about $3,100 of that) Effective hourly rate: $4,820 / 60 = $80.33/hour Compare that to my sponsorship hourly rate of $53, or my ad income hourly rate of $2.40. The affiliate channel is roughly 1.5x sponsorships and 33x ads. The spread is wild. And here is the kicker: the hourly rate is going UP over time as my recurring users accumulate. My December 2025 affiliate income was $612. My January 2026 was $648, and I did almost nothing new that month. That is the compounding effect in action. # # What I Would Tell Someone Starting Today If you are a developer or technical content creator reading this and thinking about adding affiliate income to your side hustle, here is the order of operations I would recommend. First, pick ONE affiliate program with a strong recurring commission structure and learn it deeply. Do not spread yourself thin across ten programs. I made that mistake early on. Second, build comparison or tutorial content around that product. Do not just slap a banner in your sidebar and hope. The income comes from genuine, helpful content that ranks in search or gets shared in communities. Third, track everything in a spreadsheet. If you cannot measure it, you cannot optimize it. My Notion dashboard is the single biggest reason I have stuck with this — seeing the numbers grow month over month is addictive in a productive way. Fourth, be patient with recurring programs. The first 3 months will look small. Month 6 starts to feel real. By month 12, you will have a baseline income stream that grows even when you take a week off. # # The Final Word (And A Genuine Recommendation) I do not shill things I do not use. The Global API affiliate program is the one I promote most actively because it is also the one I would recommend to a friend who asked me which AI API gateway to use. The recurring 8% commission means my incentives are aligned with the platform's long-term success — I want my referred users to stick around, which means I only refer people I think will actually benefit from the service. The numbers are straightforward: 15% on the first order, 8% recurring on monthly renewals, 10% recurring on premium upgrades, payouts via PayPal once you hit $50, and access to 150+ AI models through one API key. If you have any kind of technical audience — a blog, a newsletter, a Discord, a YouTube channel — this is the kind of program that fits naturally into content you are probably already creating. I added up my projected 2026 Global API income at current growth rate and it should clear $5,400 in passive recurring commissions alone. For content I largely already wrote. That is the power of picking the right program once and letting it compound. If you want to check it out for yourself, the affiliate program is at https://global-apis.com/affiliate. Sign up, grab your link, drop it into a piece of content, and start tracking the numbers. Your future self will thank you in month 12 when you open your dashboard and see income you barely lifted a finger for that month.