How to start any business for under $100

Executive overview

Most founders waste months building a product nobody wants. The fix is to separate validation from execution — prove people will pay before you build anything.

Pick a massive, universal market: everyone eats, sleeps, wears clothes, and wants entertainment. Then find the cheapest possible way to deliver the outcome manually and collect real money.

The core insight: get three paying customers first — making the product is the trivial part.

The manual validation method

• Strip any business down to the core problem it solves, ignoring the tech layer
• Solve it manually for a small group to confirm demand before building anything
• Collect actual payment (Venmo, cash) — interest is not validation, money is
• Only invest in infrastructure once manual delivery becomes the bottleneck

Uber: ride-sharing without an app

• Text friends: "I'll get you a driver within 10 minutes — text me when you need a ride"
• Recruit drivers via Craigslist and Facebook Marketplace
• Build the customer list and driver pool manually; add tech only when texting can't scale

Chipotle: fast-casual food without a restaurant

• Hire a local chef to cater a test dinner at your home for $10 a head
• Run it repeatedly until you outgrow the space — only then find a location
• Model unit economics first: ingredient cost × volume × price = viable or not

Patron: a drinks brand without a distillery

• Identify a gap in the market (e.g. flavoured or spicy tequila) before making anything
• Text potential buyers and ask if they'd pay ~$30 a bottle — collect payment upfront
• Find production via YouTube tutorials and paid consultants (sommeliers, co-packers)
• Document the journey publicly; the audience becomes the launch customer base

TikTok: an entertainment platform without an algorithm

• Send one curated gif or short video per day to 10 friends; track what they like
• Source content from highly rated Reddit threads — the crowd has already validated it
• Use an email list (e.g. SendFox) to measure open and click rates before building a platform
• The algorithm and the app are the easy part; proving people want the content is the hard part

Patagonia and Gymshark: clothing without a factory

• Identify an underserved niche (ugly cycling jerseys, gym clothes for young influencers)
• Borrow materials or design language from adjacent industries
• Contact manufacturers directly — many (e.g. Giordana) will help design and produce small runs
• Distribution channel choice matters as much as product: Gymshark used YouTube, not retail

Quest Bar: food products without a lab

• Test recipes at home first; professional production comes after demand is proven
• Use a co-packer — a contract manufacturer that produces to your specification at scale
• The co-packer model is how most consumer packaged goods actually get made