No. 01 — Earn
Start an online business
A viable online business can be started for under $200 and take its first payment inside a month. The expensive input isn't money — it's the six months most people spend building something nobody ever agreed to buy.
Five models, honestly compared
Near-zero capital rules out inventory, payroll, and paid customer acquisition. That leaves five realistic models. The column that matters most is the last one, because it decides whether you built a business or a job with worse benefits.
| Model | Startup cost | Time to first dollar | Your time at scale |
|---|---|---|---|
| Productized service | $0–$500 | 1–4 weeks | High, until you document and hire |
| Software / tools | $0–$2,000 | 3–12 months | Low per customer, but support and maintenance never stop |
| Content and audience | $50–$300 | 6–18 months | Medium — revenue decays if publishing stops |
| Digital products / courses | $100–$1,000 | 1–6 months | Low to produce, high to keep selling |
| E-commerce (physical goods) | $2,000–$20,000+ | 1–3 months | High — inventory, shipping, returns, ad spend |
For someone starting from nothing, the productized service wins on every axis that matters early: fixed scope, fixed price, sold before it exists, cash up front. It differs from freelancing in one structural way — you sell a defined deliverable instead of your hours, which is what eventually lets a contractor or a script do the delivering. Software and courses look better on paper and are where most first-timers burn a year. E-commerce is a working-capital business wearing a website; if the capital isn't there, neither is the business.
The first 90 days
Weeks 1–2: pick a problem you already understand
Not a passion, not a trend — a problem you have watched cost someone money. Your last job is the richest source: you know who has the budget, what they already pay for, and which vendors annoy them. Pick a buyer you can name and reach. If you can't list twenty specific companies or people who have this problem, pick a different problem. Whatever skill you already have that someone pays for is the raw material.
Weeks 3–6: sell before you build
Write a one-page offer: what they get, when, for how much. Take it to those twenty buyers directly — email, phone, in person. The goal is not feedback. The goal is money, or a signed "yes, when it's ready." Ten conversations that end in polite interest and zero dollars is a verdict, and it is far cheaper to hear it in week four than in month nine. Charge from the first customer. Free pilots teach you nothing about whether anyone will pay.
Weeks 7–12: deliver manually, then automate
Do the work by hand, badly and slowly, for the first several customers. Manual delivery is how you find out that the part you were going to spend three months automating is the part nobody cares about. Once you have run the same job five or six times, write down every step; automate only the steps that repeat identically. Software built after that point tends to get used. Software built before it tends to get rewritten.
What it actually costs to start
Real ranges, not a fantasy budget. The recurring costs matter more than the one-time ones, because they run whether or not you have customers.
| Item | Typical range | Notes |
|---|---|---|
| Domain name | $10–$20 per year | Renews annually; premium names cost far more |
| Hosting / site builder | $0–$30 per month | Static sites are often free; stores cost more |
| Business registration (US LLC) | $50–$500 one-time | Varies widely by state, plus an annual report fee |
| Payment processing | ~2.9% + ~$0.30 per charge | Standard card-processing pricing; the percentage is the real cost |
| Bookkeeping / accounting software | $0–$50 per month | A spreadsheet is fine below roughly 50 transactions a month |
| Liability insurance (client-facing work) | $30–$80 per month | Often required by contract with larger clients |
Ranges are illustrative and current as of publication; registration fees and processing rates vary by state and provider. Check current pricing before you budget.
Add it up and a service business starts for something like $100–$600 in the first year. Notice what is absent: an office, a logo, a mobile app, and a course. Those are purchases people make to feel like a founder before they have a customer.
The failure patterns, plainly
Building for months before talking to a buyer. The single most common way this ends. Time is the capital you actually have, and a launch is not evidence of demand.
Choosing a market with no budget. Hobbyists, students, and other early-stage founders are the three hardest customers in the world. Businesses that lose money without your product buy fastest.
Competing on price. Being the cheapest attracts the worst clients and leaves nothing to reinvest. Compete on speed, on a narrower promise, or on being the person who actually answers the phone.
Treating "passive income" as the goal. The correct goal for the first year is a first paying customer, then ten. Income that runs without you is an outcome of a working business, not a starting strategy — that distinction is the whole subject of passive income.
Two specific things to skip: dropshipping, which is a low-margin advertising business where you control neither product nor fulfilment, and the great majority of "start an agency" courses. Both sell the course, not the business. The tell is that the revenue being demonstrated comes from teaching the method rather than running it.
Where the profit goes
Business income only becomes net worth if it leaves the business. Pay yourself on a schedule, hold back a cash reserve for taxes, and route the rest into investments the same week it arrives — the discipline described in savings rate applies harder to lumpy income than to a paycheck. An extra $1,000 a month invested at 7% is $520,927 after 20 years. Run your own figures in the millionaire calculator, see the other levers in increase your income and side hustles, and read how to make a million dollars for where earned income sits in the whole machine.