Startups · Entrepreneurship
Start a startup the right way
Find a real problem, test it cheaply, win your first customers by hand and grow before you raise.
Learning how to start a startup begins with one idea from Y Combinator co-founder Paul Graham: “A startup is a company designed to grow fast.” Everything else, from the problem you choose to the first customers you win and the money you raise, should serve that growth.
Americans are starting companies at a steady pace. The Census Bureau counted 534,947 applications for an Employer Identification Number in September 2026, seasonally adjusted, and 147,331 of them were “high-propensity” applications, the kind most likely to become businesses with payroll. The West, which includes California, accounted for 130,070 applications.
Only a small share of those applications are startups in Graham’s sense: companies built to reach many customers quickly, usually with technology, and often with outside investors. This guide walks through the path from idea to first paying customers. If you are still deciding whether a high-growth company is right for you, read our entrepreneur vs. small business owner guide first.
Startup or small business? Know which you are building
Startup vs. small business
| Question | Startup | Small business |
|---|---|---|
| Goal | Grow fast to a very large market | Steady profit in a known market |
| Product | Often new, usually technology-driven | Usually a proven product or service |
| Funding | Often angels and venture capital | Savings, loans, revenue |
| Risk | High; many fail, a few grow very large | Lower; growth is slower |
| Typical example | Software sold nationwide | Restaurant, salon, contractor |
Neither path is better. A family restaurant that pays its owners well for 30 years is a success. But the steps below assume you want the startup path, where growth is the scorecard.
Step 1: Start with a problem, not a product
The costliest early mistake is building something nobody needs badly enough to pay for. Start instead with a problem you understand, ideally one you have lived through at work or at home. Write it in one sentence: who has the problem, how often and what it costs them today.
- Look in your own industry. Years in a field show you problems outsiders never see.
- Watch for workarounds. Spreadsheets, sticky notes and manual steps point to a missing product.
- Ask “who pays?” A problem matters to a startup when someone will pay to solve it.
Step 2: Test the idea before you build
Before writing code or ordering inventory, talk to 20 to 30 people who have the problem. Ask how they handle it now, what it costs them and what they have tried. Do not pitch; listen. Then test demand with something small: a one-page website with a sign-up form, a mock-up or a pre-order.
Ask about the past, not the future. “Would you buy this?” invites polite yeses. “When did this last happen, and what did you do?” gets the truth. If no one has tried to solve the problem, it may not hurt enough to build a company on.
Step 3: Build the simplest first version
Your first product, often called a minimum viable product (MVP), should do one thing well for one kind of customer. It can be rough. Many founders start with no-code tools, a spreadsheet behind a form or a service they deliver by hand while they learn what to automate.
From idea to first paying customers
- Write the problem statementWho has the problem, how often and what it costs them.
- Interview 20 to 30 potential customersLearn how they solve it today and what they pay.
- Test demand cheaplyA landing page, a mock-up or pre-orders before you build.
- Build a minimum viable productOne core feature for one customer type, shipped in weeks, not months.
- Recruit your first users by handEmail, call and visit them; set them up yourself.
- Measure weekly growthPick one number, such as revenue or active users, and track it every week.
- Decide on fundingRaise money only when it speeds up growth you can already see.
Step 4: Win your first customers by hand
In his July 2013 essay “Do Things That Don’t Scale,” Graham wrote that the most common unscalable thing founders must do at the start is to recruit users manually. That means personal emails, calls, visits, demos at a customer’s desk and setting up accounts yourself. Early customers forgive a rough product when the founder answers the phone.
- Start close to home. Former co-workers, local businesses and industry groups are your first market. Sacramento founders can meet them at events listed on Inksight.
- Charge early. Payment is the clearest proof that the problem is real.
- Write down every complaint. Your first 10 customers design your next version.
Step 5: Measure growth every week
Graham’s September 2012 essay set a benchmark: a good growth rate during Y Combinator is 5% to 7% a week. Small weekly gains compound into very large yearly ones, which is why startups track one number every week.
What weekly growth compounds into over a year
Illustrative calculation: 1.05 to the 52nd power is about 12.6. At 7% a week, the same math gives about 33.7 times in a year. Few companies hold these rates for long.
Five-year survival of U.S. start-ups by birth year
Source: U.S. Bureau of Labor Statistics, Business Employment Dynamics, Spotlight on Statistics (January 2024). Share of new establishments still operating five years after birth.
The survival numbers are a reminder to manage cash carefully. About half of new establishments close within five years, and venture-style startups take even bigger risks. Track how many months of cash you have left, and keep personal and business money apart from day one.
Step 6: Set up the business properly
You do not need a corporation to test an idea, but once you take money from customers or partners, set up the basics. Choose a legal structure (our business structure in California guide compares LLCs and corporations), get a free EIN from the IRS, open a business bank account and put co-founder roles, equity and vesting in writing.
Keep receipts for what you spend before launch. The IRS lets a new business elect to deduct up to $5,000 of start-up costs in the year it begins, reduced by the amount total start-up costs exceed $50,000, and amortize the rest over 180 months.
Step 7: Raise money only when it speeds up growth
Many strong companies start with savings and early revenue. Outside money makes sense when you can show growth and need to move faster than revenue allows. Our startup funding stages guide explains pre-seed through Series A, our pitch deck guide shows what investors expect to see and our Y Combinator explained guide covers accelerators.
Common mistakes
- Building before talking to customers. Months of work can answer a question nobody asked.
- Trying to serve everyone. A first product for one customer type learns faster.
- Raising money too early. Selling equity before you have proof costs more of your company.
- No co-founder agreement. Unwritten equity splits cause many early breakups.
- Tracking too many numbers. Pick one growth metric and watch it weekly.
Key terms
- Startup: a company designed to grow fast, usually by reaching a very large market.
- Minimum viable product (MVP): the simplest version that solves the core problem for real users.
- Product-market fit: the point where customers want the product enough that growth comes steadily.
- Vesting: earning founder or employee equity over time, so people who leave early do not keep it all.
- Runway: the number of months a company can operate before it runs out of cash.
The bottom line: start a startup by solving a problem people will pay to fix, test it cheaply, win the first customers yourself and measure growth every week. Money, offices and big hires can wait until the numbers prove the idea.
Free help in the Sacramento region
You do not have to work through these decisions alone. SCORE Sacramento pairs owners with volunteer mentors, many of them retired executives, bankers and accountants, at no cost. The Sacramento Valley Small Business Development Center offers free one-on-one advising on financing, budgeting and financial statements. The SBA’s Sacramento District Office can explain loan programs and connect you with lenders, and community lenders such as California Capital and Opening Doors serve owners who need smaller loans or extra support. Browse our Business Resources and the free business help guide for more.
Key takeaways
- Paul Graham defines a startup as a company designed to grow fast; growth is the scorecard.
- Talk to 20 to 30 people with the problem and test demand before you build.
- Recruit the first users by hand; Graham calls it the most common unscalable task.
- 5% weekly growth compounds to about 12.6 times in a year (illustrative calculation).
- BLS data show 57.3% of 2018 start-ups survived five years; manage cash from day one.
Frequently asked questions
How do I start a startup with no money?
Start with a problem you know, interview potential customers, test demand with a landing page or pre-orders and build a simple first version with low-cost or no-code tools. Charge early so revenue funds the next step.
What is the difference between a startup and a small business?
A startup is built to grow fast, usually with technology and often with investor funding. A small business aims for steady profit in a known market, such as a restaurant or contractor.
How do startups get their first customers?
Usually by hand. Founders email, call and visit potential customers, run demos and set up accounts themselves. Paul Graham calls recruiting users manually the most common unscalable thing founders do.
What is a good growth rate for a startup?
Paul Graham wrote in 2012 that a good growth rate during Y Combinator is 5 to 7 percent a week. Over a year, 5 percent a week compounds to about 12.6 times.
How many startups survive?
Bureau of Labor Statistics data show that 57.3 percent of new establishments born in 2018 were still operating five years later, the highest rate in its table since 1994.
Sources
- Paul Graham: Startup = Growth (September 2012)
- Paul Graham: Do Things That Don’t Scale (July 2013)
- U.S. Census Bureau: Business Formation Statistics, September 2026 (Oct. 14, 2026)
- U.S. Bureau of Labor Statistics: Business Employment Dynamics Twentieth Anniversary! (January 2024)
- U.S. Bureau of Labor Statistics: 1-year survival rates for new business establishments by year and location (March 4, 2024)
- IRS: Instructions for Schedule C (Form 1040) (2025), business start-up costs
Keep reading on Inksight
- Business Credit vs. Personal Credit: What Every Small Business Owner Should Know
- How to Choose a Business Structure in California: Sole Proprietorship, LLC, S Corp or C Corp
- Business Banking 101: How to Choose and Open a Business Bank Account
- Small Business Funding Options: Loans, Grants and Investors
- Business Resources: all guides →
