Startup Statistics: Where to Find Reliable Data and How to Use It

What counts as a startup depends on who is counting

A startup might mean any new business, a new employer firm, or a venture-backed company. Each definition produces very different totals.

Government agencies usually treat a new business as a firm or establishment in its first year or first few years. Venture capital databases define a startup as a young, high-growth company seeking outside investment. A neighborhood bakery and a software company funded by investors are both startups in everyday talk, but they behave very differently.

Before using a statistic, find the definition and the group measured. Many headline figures cover only employer businesses, while most new businesses start with just the owner. Our post on how the SBA defines a small business shows why definitions matter.

Where official startup data comes from

The U.S. Census Bureau, the Bureau of Labor Statistics, the SBA Office of Advocacy and the Kauffman Foundation publish research on business formation. Use their latest releases.

• Census Bureau Business Formation Statistics. Tracks applications for new employer identification numbers, a signal of new business interest, and the Business Dynamics Statistics on firm births and deaths.
• Bureau of Labor Statistics. Publishes business employment dynamics, including establishment openings and closings and survival data.
• SBA Office of Advocacy. Summarizes small business data and research.
• Federal Reserve small business credit surveys. Report on how new and young firms seek financing.
• Kauffman Foundation. Studies entrepreneurship activity and indicators.

We do not print specific figures here because they update frequently. Visit each source for the current release and note the publication date.

Questions to ask before trusting a number

Ask who measured it, how, over what period and for which businesses. If you cannot trace it to an original source, do not rely on it.

A short checklist:

1. Source. Is it an agency, an academic study or a company marketing a product?
2. Method. Was it a count of records or a survey of volunteers?
3. Sample. Does the sample match businesses like yours?
4. Date. Does it cover a period relevant now?
5. Measure. Is it a mean, median or percentage?

Be careful with statements such as most startups fail because of X. Reasons for closure come from surveys of owners and vary. Read small business survival and failure rates for how to read those claims.

Beware of round numbers repeated across many blogs. When a figure appears everywhere with no link to an agency report, it has often been copied from article to article, and the original may have measured something else entirely or be many years old.

Statistics that matter to a founder's plan

Industry margins, local demand, wage data and time-to-profit benchmarks help more than total startup counts. Use the specific over the broad.

What helps in a plan:

• Industry cost structures. Trade associations and agency data show typical margins and expenses.
• Local demographics and spending. The Census American Community Survey gives local data on population and income.
• Wage rates. BLS publishes occupational wage data by region.
• Time to break even. Talk to owners in your field to learn how long it took them.

Then replace averages with your own forecast. The most useful number for a startup is the one in your own cash-flow plan. See financial forecasting for small business and budget forecasting.

One more practical source is people. Call three owners in your field and ask how long it took to reach break-even, which costs surprised them and what they would do differently. Their answers are small samples, but they are specific to your industry and place.

What statistics cannot tell you

A national average cannot predict your result. Your market, price, costs, skill and cash cushion matter far more than any population-level figure.

Statistics describe groups. Your business is one case. Two coffee shops on the same street can have very different outcomes because of location, pricing, management and cash reserves.

Use data to avoid obvious mistakes, such as entering an oversaturated niche without a difference, or underestimating how long it takes to become profitable. Then test your own idea with real customers. Our guide to starting a small business lays out a practical test-and-launch sequence.

Using data when you seek outside money

Investors and lenders want your numbers, with market data as context. Cite the source and year, and tie claims to your own evidence.

A good plan includes a short market section with sourced data, but funders weigh your own track record more. Use statistics to show that demand exists, then use your own sales, pre-orders or contracts to show you can capture it.

Startups under 6 months are usually not a fit for ROI Funder. We consider businesses with 6+ months in business, $20,000+ monthly revenue and 500+ FICO. Once you reach that stage, bank statements carry the file for products like working capital. Until then, review how to get startup funding and small business statistics for related research.