Launch a real online business
this weekend.
You talk. Claude builds. Start & Ship guides you from niche research to a live affiliate content site on your own domain, and it assumes you've never built a website. Every decision it makes for you was extracted from four live sites you can verify. No income promises, no guru math. The numbers behind this product are public, losses included.
one-time · lifetime updates · all sales final
you'll need: a computer · Claude (~$20/mo) · a domain (~$12/yr) · a free weekend
Extracted from sites you can visit right now:
“AI search broke the traffic math.”
I'm not going to pretend it didn't. ChatGPT answers the question and most of the clicks never happen. That's the market this kit was built in, not the one from 2019. Every template, schema, and guardrail in the repo is engineered for the post-click, post-citation era: pages built to be cited, quoted, and trusted when a visit does land.
So, plainly: I make zero promises of rankings, citations, or income. Nobody honest can. What I promise is working code, extracted from four sites you can visit, with the books kept in public: losses and the clawback included.
— Charles McQuain
[signed 2026-07-14 · verify at realnumbers.online]
How the money works.
An affiliate site earns a commission when a reader clicks a recommendation and buys. The merchant pays, never the reader. Every one of the kit's six page templates exists to serve that moment honestly: earn the trust, make the right recommendation, take the better commission.
You own the site, the content, and the list. Nothing rides on a platform's algorithm change surviving the month.
What merchants pay, per sale
Rates are the merchants' published terms, not ours. The kit's dual-CTA component simply puts the better link first.
More than one way to earn from the same pages
What it costs you
Hosting honestly: Vercel's free tier runs this architecture easily, but its terms ask commercial sites for the $20/mo plan — most start free and upgrade when the site earns it. What you earn depends on your niche and your work. No projections here; see the honest part above.
What's inside
Seven entries. Everything below ships in the private repo on day one.
Click through all six templates, live ↗Decision-hours, estimated. The code is the easy part.
Six money-page templates
The “Best X for Y” roundup, single product review, head-to-head comparison, gift guide, price-tier ladder, and supporting editorial: the six page shapes that actually earn, with the layout decisions already made.
AI-surface layer
JSON-LD factory, llms.txt + llms-full.txt, answer-first blocks, schema CI. The part that makes the site legible to machines.
Agentic coding layer
CLAUDE.md, skills, and typed guardrails so agent errors become build errors.
export const Review = z.object({
title: z.string().max(60),
rating: z.number().min(1).max(5),
testedOn: z.string().date(),
sources: z.array(Source).min(3),
affiliate: z.object({
href: goLink,
disclosed: z.literal(true),
}),
});schemas/review.ts — actual file from the kit
Affiliate plumbing
/go/ redirects, dual-CTA components, FTC disclosure components wired in by default.
Source ledger + staleness CI
Every claim carries a source; expired citations fail the build.
Editorial & compliance pages
Correction log, methodology, author schema, privacy/terms/disclosure templates.
Operator's Notes
Real timelines, real revenue, and the $0-site teardown. The docs nobody else ships.
Why not just ask Claude to build this myself?
You could. The question is whether you'd know to ask for a tracked-redirect affiliate registry, FTC disclosure on every recommendation rather than a footer, the correct JSON-LD type per page (a comparison isn't a Review), and a citation-staleness gate that fails the build instead of letting sources quietly rot. An agent starting from a blank repo only builds what it's told to build, and it will happily ship something that looks finished without any of that.
A fresh agent, given only this repo's docs and no peek at the internals, shipped two working pages on the first try without opening a single schema file. That's not a claim you couldn't build a Next.js site with an agent — of course you could. It's that getting the specific decisions right, in the right shape, on the first pass, is what you're paying for.
Point an agent at it.
One copy-paste setup prompt takes Claude Code from clone → configure → deploy. Typed frontmatter, zod schemas, and CI guardrails mean the agent physically cannot publish a page that doesn't validate.
Agent errors become build errors.
And that prompt ships in the box. If you can run Claude Code, you can run this kit. You don't have to start from a blank repo, or a blank skillset.
$ claude
> /add-review-page --product "Old Forester 1920"
● Reading skills/add-review-page/SKILL.md
● Validating frontmatter against schemas/review.ts
● Writing content/reviews/old-forester-1920.mdx
● pnpm check:sources ......... 12/12 citations fresh
● pnpm build ................. passed
Page ready for editorial review.
Pricing
one-time purchase · lifetime updates via the repo · no subscription
[tier_01]
Founder
$149 /one-time
- +the /start-and-ship guided launch — you talk, Claude builds
- +the full site codebase, delivered to your GitHub automatically at checkout
- +every skill in the box: niche interview, affiliate-program finder, deploy walkthrough, launch-week plan
- +lifetime updates · unlimited sites you own
- −no pre-researched niche dossiers — the interview gets you there yourself
[tier_02]
Founder Pro
$249 /one-time
- +everything in Founder
- +3 niche dossiers: real research on three niches I vetted myself — demand signals, every affiliate program with real rates, the first 20 pages mapped, and why each one passed or didn't
- +the decision stays yours — use one, or reject all three and run the interview instead
- +capped at 10 buyers, so no dossier gets handed to a crowd of identical sites
[tier_03]
Agency
$399 /one-time
white-label · client work
For people who build sites for clients — if that's not you, Founder is your tier.
- +everything in Founder
- +unlimited client sites
- +white-label rights
- −no vetted niche dossiers
what it actually costs, year one: the kit once, plus ~$20/mo Claude while you're building and ~$12/yr for a domain — no subscription to us, ever
Built and operated by Charles McQuain — every site's numbers, wins and losses, published at realnumbers.online.
Auto-delivery in minutes · failed delivery is a same-day refund, no argument · beyond that, all sales are final: repo access can't be un-downloaded, so a blanket refund window on a digital product would just be an invitation to clone the code and ask for the money back. Prices move in one direction, up, and never quietly; the terms below are signed.
The terms, pre-registered
[2026-08-24]- i.
The price is $149 / $249 / $399, paid once. No subscription, no seat count, no upsell after checkout. Updates to the kit are free for as long as it exists.
- ii.
Prices move in one direction, up, and any change is posted here and dated before it takes effect. There is no discount code, no seasonal sale, and no coupon that mysteriously always works. What you see is what everyone pays today.
- iii.
Every sale is final once repo access is granted, the honest tradeoff of a digital download that can't be un-downloaded. If delivery itself ever fails, you're made whole immediately. Every sale is logged on the public ledger. No hidden numbers, no quiet pivots.
Charles McQuain
signed for Start & Ship · 2026-08-24
one-time purchase ........ no subscription
Is this for you?
For you if —
- +you've never built a website. The guided path assumes exactly that: you answer questions and make the calls, the agent does the technical work.
- +you're a developer who'd rather not start from a blank repo, or you already run a coding agent.
- +you want the affiliate and AI-search machinery pre-decided, not researched for six weeks.
- +you'd rather verify receipts than trust testimonials.
Not for you if —
- −you want passive income. The setup is fast. The earning part is a real business, and it's earned.
- −you need traffic, ranking, or citation guarantees. There are none here.
- −you want someone else to pick your niche and write your reviews. The kit guides both, hard; it does neither for you.
- −you're looking for a drag-and-drop page builder. This is a real codebase with an AI operator.
FAQ
What exactly do I get?
Access to a private GitHub repo containing the full kit described above, plus lifetime updates via git tags. One-time purchase, no subscription.
What literally happens after I buy?
Checkout runs through Polar (the merchant of record). The moment payment clears, you land on our start-here page, and Polar grants your GitHub account access to the private repo — if you don't have a GitHub account yet, that page walks you through creating one (free, three minutes) and connecting it. Lifetime updates arrive as new versions on that repo. For the day-by-day path from there to a live site — guided or manual — see /how-it-works.
What's the refund policy?
All sales are final. Once Polar grants your GitHub account repo access, the code can't be un-downloaded, so a blanket 'no questions asked' window on a digital product isn't honest trust-building, it's just an invitation to clone the code and ask for the money back. The one exception: if delivery itself fails, we fix it or refund you immediately, no argument. That's on us, not a favor.
What if I get stuck?
Email hello@startandship.com or open an issue on the kit's repo if you have access — I answer within one business day. I'm one person, so that's a real day, not a chatbot minute. If /start-and-ship itself gets stuck on your machine, that's a bug in my product, not a failure in you: I fix it, and the fix ships to every buyer as an update.
Will there be a sale if I wait?
No. There's no discount code, no seasonal sale, and no coupon that mysteriously always works. When the price changes it goes up, and it's announced and dated in the signed terms above before it does. Waiting is a fine decision; it just won't get you a cheaper price.
I've never built a website. Can I actually do this?
Yes, and you're who the guided path is built for. The kit ships with /start-and-ship: open Claude Code, type it, and the agent walks you from never-opened-a-terminal to a live site. It checks your machine and installs what's missing, interviews you through niche selection, sits with you while you apply to affiliate programs, writes the configuration, deploys, and connects your domain. You make every decision and click anything involving money or accounts; the agent does the technical work. Progress saves as you go, so a weekend split across sessions works fine.
Do I need Claude Code or Cursor?
For the guided experience, you need Claude Code — it runs /start-and-ship and the kit's skills, and it's included in a Claude subscription (about $20/mo). Developers can skip the guided path entirely: it's a clean Next.js + TypeScript codebase that works by hand or with Cursor/Codex, and the docs ship equivalent copy-paste prompts for agents without skills support.
What does it cost to run, all-in?
The kit is a one-time purchase. Running the site: a domain is about $12/yr, and Claude is about $20/mo while you're actively building (cancel anytime; the site runs without it). Hosting, honestly: Vercel's free tier handles this architecture easily, but Vercel's terms ask commercial sites for the $20/mo plan — most people start free and upgrade when the site earns it. There is no subscription to us, ever.
How is this different from ShipFast?
Same shape: one-time repo purchase, opinionated architecture. Different substance: ShipFast encodes SaaS plumbing; this encodes 4 months of operating content and affiliate sites, from schema factories to FTC components to the source ledger.
Why not just ask Claude to build this myself this weekend?
You could ask it to build *a* site. The question is whether you'd know to ask for a tracked-redirect affiliate registry, FTC disclosure on every recommendation (not just a footer), the correct JSON-LD type per page (a comparison isn't a Review), and a citation-staleness gate that fails the build instead of letting sources rot — an agent starting from a blank repo only builds what it's told to, and it'll happily ship something that looks finished without any of that. We ran exactly this test: a fresh agent, given only this repo's docs and no peek at the internals, shipped two working pages on the first try without opening a single schema file. That's not a claim you could build a Next.js site with an agent — of course you could. It's that getting the specific decisions right, in the right shape, on the first pass, is what you're paying for.
Why sell this instead of just running your own sites?
They're not in competition. This kit is a fixed, one-time product; my own sites earn or don't independently of how many people buy it. Selling working architecture isn't inconsistent with it being early — everything in the kit is extracted from real, running sites, not planned for one. And if it ever stops being honest to sell — if the receipts stop backing it up — the fix is public: I'd say so on the ledger, not quietly pull the page.
What are the license terms?
Founder and Pro: unlimited sites you own. Agency: white-label rights for unlimited client sites. No resale or redistribution of the kit itself.
all sales final · one-time · no subscription
