Richard Yu's $32.7M Digital Products System
Richard Yu says he made $32.7 million selling other people's digital products via AI-built affiliate storefronts and faceless AI videos. The process is well known, but the income math is hypothetical and he never addresses Instagram's suppression of AI link-farming content.
THE OPPORTUNITY
Richard Yu is pitching an "AI product arbitrage" model: promoting existing digital products (ebooks, courses, supplements) through affiliate links. He is using AI to build a website in under 60 seconds, and creating AI-generated social content to drive traffic without showing your face. The core point is earning 50–90%+ commissions on other people's products by sending traffic from Instagram or similar platforms to your affiliate storefront.
HOW IT'S EXECUTED
- You start by browsing affiliate marketplaces like Allen.store or Digistore24 to find digital products in health, wealth, or relationships niches. These pay commissions ranging from $15 to several hundred dollars per sale.
- You grab your unique affiliate link, then use Wix's AI website builder to create a storefront in under a minute — paste the affiliate link, add product images and descriptions, and you're live.
- For traffic, you create short-form content (7–20 second videos) using AI avatar tools like HeyGen.
- The process: find high-performing content in your niche on Instagram.
- Extract the script using Allen AI, modify it slightly (using his M&M- model and modify approach), generate an AI avatar reading that script, and post it daily
- When viewers click your bio link and purchase, you collect the commission.
He claims this requires no inventory, no customer service, no ad spend, just posting AI-generated content and letting Instagram's algorithm distribute it for free.
WHAT'S CREDIBLE
The underlying affiliate marketing mechanics are well documented. Affiliate networks do exist, commissions in the 40–75% range are standard for digital products, and some supplement companies do offer aggressive front-end payouts (100–300%) banking on subscription retention. The AI tools he names,Wix, HeyGen, and similar platforms, can perform the tasks described. The example Instagram accounts he shows (Yangmong, Susan Miller) are real and do use AI avatars. Wix is indeed a publicly traded, multi-billion-dollar company. The basic framework: find a product, build landing page, drive traffic via content, is how digital marketing has worked for two decades. What he's describing is not novel; he's packaging traditional affiliate marketing with newer generative AI tools.
WHAT'S OMITTED OR OVERSTATED
The income claims lack context entirely. We're told he personally made $32.7 million, but over what time period, with what team size, and from which specific activities? His students' results, Priscilla's $200K, Sean's $250K, are presented without failure rates, sample sizes, or disclosed time investment beyond "2–3 hours a day." The math model he walks through (1,000 views per video, 1% click rate, 5% conversion, $200 average commission) is purely hypothetical and ignores platform saturation, audience quality variance, and the likelihood that conversion rates on cold traffic to affiliate offers are typically well below 5%.
He glosses over the most material friction point: Instagram does not reward affiliate content the way he implies. Accounts posting repetitive AI-generated videos with external links in bio face algorithmic suppression, and Instagram has actively discouraged link-farming behavior. The example accounts he shows may have succeeded before the platform's recent crackdowns on AI spam. He never addresses deliverability, account suspension risk, or what happens when thousands of people post similar AI content simultaneously, which is the logical outcome if his system scales.
The Wix "scholarship" framing is misleading. This is not a corporate philanthropy program; it's a standard affiliate upsell where he earns commission when viewers sign up for Wix's paid tiers. The claim that Wix "sponsors your tuition" obscures the fact that you're paying Wix for a website builder. His course is the incentive for you to subscribe through his link, not a gift from a billion-dollar company moved by his impact. The pricing pressure ("limited spots," "can't guarantee tomorrow") is classic direct-response urgency tactics, not a reflection of genuine scarcity.
He repeatedly emphasizes "no prior experience needed" but never addresses the skill gap in content strategy, copywriting for conversions, or navigating affiliate program compliance rules. The supplement example with 300% commissions is particularly opaque. Such payouts exist, but they're contingent on recurring subscription conversions, and the viewer would need to understand whether they're liable for refund chargebacks or customer disputes. None of this is mentioned.
BOTTOM LINE
This could work for someone with existing content creation discipline, comfort testing AI tools, and patience to iterate through platform restrictions and low initial conversion rates. But the income projections are divorced from typical outcomes, and the "passive" framing ignores the fact that algorithmic content distribution is anything but guaranteed. If you already understand affiliate marketing and want to experiment with AI content efficiency, the toolchain he describes is real. If you're expecting to deploy this and see four-figure monthly returns within 30–60 days as implied, you're walking into a structure where most participants will earn nothing and a small fraction will succeed, and he's provided no data on which outcome is likelier.
OPPORTUNITY DESK RATING
Concept Viability: 3/5
Affiliate marketing works, AI tools reduce production challenges, and short-form content can drive traffic, but platform risk, audience saturation, and conversion rate realities make this far harder than what is presented.
Presentation Honesty: 2/5
Income claims lack context, the Wix "scholarship" is standard affiliate upselling misframed as corporate generosity, the math model is hypothetical, and critical execution barriers (platform policies, conversion variance, skill requirements) are omitted.