AI Side Hustles Half-true — works only if you do the unspoken work

By the Income Reality Check Editorial Team · Updated 2026-05-15 · 7 min read

Selling Claude Skills for $99: the side hustle the pitch leaves out

Verdict: Half-true — works only if you do the unspoken work. The product is real and the $3,000 number is plausible, but the guest is a working marketer with an audience, an email list and a ranked website doing most of the heavy lifting.

On the May episode of The Koerner Office Podcast, host Chris Koerner interviews Ryan Dozer, an AI marketer who claims he made “over $3,000 just passively as a digital product” in 30 to 45 days by selling a “Claude Code Skill Stack” — a bundle of around 20 to 25 markdown files for $99. The pitch is that anyone with knowledge of anything (“a gardener, a marketer, or you work in nonprofits”) can package it as Claude Skills and sell it. The product behind the pitch genuinely exists. The income-replication story is more complicated.

What the video actually claims

Dozer is introduced as someone who “has made a living from building and selling Claude skills.” He shows his Stripe dashboard on screen and tells Koerner that the bundle has cleared roughly $3,000 in a month and a half — “all profit, right? Like this is just your time.” Then he notes a detail viewers might miss: “This one actually came as like a consulting upsell. Some guy wanted my skills and was so impressed by one of them and I was like, ‘All right, man. $1,000 for an hour.’” So at least $1,000 of the headline figure is one-off consulting, not the $99 digital product running on autopilot.

The method itself is straightforward. A Claude Skill is a markdown file with YAML frontmatter that tells Claude how to handle a specific recurring task — Dozer compares it to a standard operating procedure or a recipe. He recommends building one skill per task (SEO blog post writer, email newsletter writer, YouTube thumbnail designer), bundling them and selling the bundle through a Stripe checkout on your own site. He shows the landing page he vibe-coded for the bundle, complete with FAQs, social proof and floating popups, and says the site is “gaining significant traffic across Google search and LLMs.”

The framing throughout is that this is beginner-friendly. “If you can use Chat GPT, you are technical enough to build a Claude skill.” Koerner adds, “my 10-year-old could create a skill.” The implication is the workflow itself is the bottleneck, and once you cross that bar, sales follow.

What the method actually requires

The technical part is genuinely real and genuinely simple. Anthropic published Agent Skills as a public standard in late 2025, and the official Claude API docs describe a Skill as just a SKILL.md file with a YAML frontmatter block — name, description, then markdown body. Anthropic also maintains a public GitHub repository of free Skills and ships pre-built ones for PowerPoint, Excel, Word and PDF on claude.ai and the API. The format is open, portable across Claude Code, claude.ai and other coding agents, and according to Anthropic’s engineering write-up has been adopted by 30+ tools including Cursor, VS Code Copilot, Gemini CLI and OpenAI Codex.

So the production step is roughly what Dozer describes: voice-dump expertise into Claude, ask it to use the skill-creator skill, get a .md file back. The cost is low — a few dollars in API or subscription usage to author and test a skill.

The skipped step is everything between “you have a .md file” and “money lands in Stripe.” Dozer’s own landing page is doing five things at once: organic Google ranking, LLM citation traffic, a floating popup, blog-post CTAs, and “CTAs in several of my emails and email newsletters and drip sequence emails.” Each of those is a marketing channel that took years to build. He mentions in passing that he started his SEO and content marketing agency in 2019, that he charges clients “about 10 grand a month, a little more,” and that he runs a community called AI Marketing Insiders. That is the actual machine behind the $3,000.

On unit economics, Stripe takes about 2.9% + $0.30 per online card transaction in the U.S. — closer to $3.17 on a $99 sale, not the flat 3% Dozer cites — and international cards add another 1.5% cross-border fee plus 1% conversion. Net per sale is around $95. To clear $2,000 from the $99 product (subtracting the $1,000 consulting upsell), you need roughly 21 sales in 45 days, or about one every two days. That is achievable when you have an existing email list and a ranking blog feeding it. It is dramatically harder cold.

There is also a competitive wrinkle worth naming. Buyers of a Claude Skills bundle are, almost by definition, Claude power users — exactly the people who already know Anthropic ships free first-party Skills and that there is a public repo of open-source ones on GitHub. The pitch competes against $0 alternatives that any savvy buyer can find in two minutes. Dozer’s actual moat is his branded prompts, his SEO-optimized variants, and the trust his audience has in him — not the format itself.

Who actually wins this game

The pattern is familiar from every prior “creator economy” wave, and the data is consistent across them. A 2025 Influencer Marketing Hub creator earnings report found that more than 68% of creators earned less than $50,000 in 2025 and only about 4% cleared $100,000 or more. Course and digital-product creators echo the split: a small minority with diversified income streams and an engaged audience pull most of the revenue, and the rest sell to crickets. NerdWallet’s round-up of realistic side hustles notably leaves “sell digital products” off its list of eight, and warns that even Etsy increasingly requires paid ads to get visibility.

The winners in selling Claude Skills, as far as anyone can tell six months into the format existing, fall into three groups. First, marketers with pre-built audiences in a software-buying niche — Dozer fits this exactly. Second, technical creators with developer followings on YouTube, X or LinkedIn who can route attention to a Stripe page. Third, agency operators who use the bundle as a lead magnet for a higher-ticket service. Notice that “complete beginner with no audience” isn’t on that list. Dozer’s own admission in the interview — that one of his $99 buyers turned into a $1,000 consulting client — illustrates which dollar is doing the work.

What you’d realistically earn

A realistic first-year range for a beginner with no email list, no ranked content and no existing community is roughly $0–$300 in the first 90 days, climbing to maybe $200–$1,500 a month after a year of consistent posting, SEO and email building — and only if the niche is one buyers will actually pay for. That’s directionally similar to digital-product income figures aggregated across platforms: median monthly profit for digital-product Etsy shops sits in the low hundreds of dollars, not thousands. If you treat skills as a lead magnet that funnels into consulting at $100–$1,000 per hour, the math improves quickly, but at that point it isn’t a “passive digital product” — it’s a service business with a free sample.

Compare that to the impression the video leaves: $3,000 in six weeks as a non-technical person with one prompt and a markdown file. The number is real for Dozer. The implied path to it is not the path most viewers will actually walk.

It’s also worth noting the regulatory backdrop. The U.S. Federal Trade Commission’s Consumer Reviews and Testimonials Rule, in force since October 21, 2024, can fine businesses up to $51,744 per violation for deceptive testimonials, fake social proof, or insider reviews without disclosure. The FTC has separately announced enforcement action against deceptive AI income claims, including one operator forced to pay $193,000 in a refund order for pitching an AI side hustle. These rules apply to U.S. operators; U.K. readers should check the ASA, Indian readers SEBI/CCPA, EU readers ESMA. If you do sell a Skills bundle and your landing page uses customer testimonials or income screenshots, the disclosure bar is higher than it was two years ago.

Who this is (and isn’t) for

This makes sense for someone who already has an audience adjacent to AI, software, marketing or productivity — a newsletter with 1,000+ engaged readers, a YouTube channel in the niche, or a consulting practice where Skills naturally become a deliverable. It also makes sense as a lead magnet rather than a primary product: free or $9 bundles that funnel into a $500–$2,000 service. Budget 8–15 hours a week for content, distribution and customer questions, plus a small Stripe and hosting bill.

It does not make sense if you have no platform, no domain expertise, and are expecting the product itself to drive traffic. The Skills format is open, the free alternatives are good, and the buyers who exist mostly already follow someone like Dozer. Without a wedge — an audience, an unusual niche, or a paid traffic budget you’re willing to test — the most likely outcome is a Stripe dashboard at $0.

For deeper background on the broader category, our look at the broader Claude AI side hustle picture walks through similar pitches, and our review of the Claude design side hustle covers an adjacent niche.

What to remember

Dozer’s $3,000 is plausible, the Claude Skill format is real, and packaging expertise as markdown is a legitimate product idea. The video undersells the role of his decade of SEO work, his existing client roster, his email list and his community in making any of those sales possible. Treat the workflow as a useful new packaging format for people who already have distribution, not as a beginner-friendly path from zero to passive income.