How many businesses are actually in a market
Small business counts by state and employer establishment counts by metro, from the SBA and the Census Bureau, with what they do and do not tell you.
The United States has 36.2 million small businesses, and California (4.34 million), Texas (3.52 million) and Florida (3.49 million) hold the most (SBA Office of Advocacy). At metro level the Census Bureau counts something narrower, locations with paid employees. On that measure the New York metro area had 562,802 employer establishments in 2023 (Census Bureau, County Business Patterns).
Those two figures answer how big a market is. Neither answers the question they are usually quoted for, which is whether a page about a city deserves to exist or could rank if you built it.
The short version
- The SBA Office of Advocacy’s 2025 profiles count 36.2 million US small businesses, defined as firms with fewer than 500 employees, including firms with no employees at all.
- California has 4.34 million small businesses, Texas 3.52 million and Florida 3.49 million, which together is about 31 percent of the national total.
- The Census Bureau’s County Business Patterns for 2023 counts 562,802 employer establishments in the New York metro area, 248,519 in Chicago, 222,388 in Miami and 163,103 in Atlanta.
- A business and an establishment are different units. A firm with six branches is one business and six establishments, so the SBA and Census figures cannot be added, divided or compared.
- Both counts are older than their publication dates. The 2025 SBA profiles are built on 2022 Census data, and the 2023 County Business Patterns figures were released in June 2025.
- A business count measures how many firms exist. It does not measure search demand or how hard a search is to win, so check the results page for the city before building anything.
How many small businesses are there in the largest states?
California has 4.34 million, Texas 3.52 million and Florida 3.49 million, according to the 2025 Small Business Profiles from the SBA Office of Advocacy (release, 30 June 2025). The same release puts the national total at 36.2 million small businesses, which it says account for almost 46 percent of private sector employment.
| State | Small businesses | Share of the 36.2 million national total |
|---|---|---|
| California | 4.34 million | About 12.0% |
| Texas | 3.52 million | About 9.7% |
| Florida | 3.49 million | About 9.6% |
The counts are the SBA’s. The share column is our arithmetic on the SBA’s rounded figures. The profiles for every state and territory are on the 2025 Small Business Profiles page.
How many employer establishments are in the largest metro areas?
The New York metro area had 562,802 in 2023, more than twice the count for Chicago. These are the four metro rows we have checked against the Census Bureau’s own file.
| Metro area | Employer establishments | Paid employees, mid-March 2023 | Establishments with fewer than 5 employees |
|---|---|---|---|
| New York, Newark and Jersey City | 562,802 | 8,559,248 | 348,908 |
| Chicago, Naperville and Elgin | 248,519 | 4,285,971 | 146,092 |
| Miami, Fort Lauderdale and West Palm Beach | 222,388 | 2,428,299 | 152,955 |
| Atlanta, Sandy Springs and Roswell | 163,103 | 2,692,513 | 97,111 |
Every figure is from the metropolitan area file of County Business Patterns: 2023, all industries combined. In each of the four metros, more than half of the establishments have fewer than five employees. A large metro count is mostly very small workplaces.
Photo: Erik Mclean on Pexels.
What is the difference between the SBA count and the Census count?
The SBA counts firms, including one-person firms with no staff. The Census Bureau’s County Business Patterns counts physical locations that have paid employees. The two are different units, and that is why the state figures are in millions and the metro figures are in hundreds of thousands.
| Point of difference | SBA Small Business Profiles, 2025 | Census County Business Patterns, 2023 |
|---|---|---|
| What one unit is | A firm with fewer than 500 employees | A single physical location with paid employees |
| Firms with no employees | Included | Not included |
| A company with several branches | Counted once | Counted once per location |
| Year the data describes | 2022 | 2023 |
| When it was published | 30 June 2025 | 26 June 2025 |
What the SBA figure counts
The national profile says small businesses are generally defined as firms with fewer than 500 employees. It gives its sources for the count as two Census Bureau datasets for 2022: Nonemployer Statistics and Statistics of US Businesses (SBA, United States profile). Nonemployer firms are in the number. A self-employed electrician with no staff is one of the 36.2 million.
That makes the figure large, and a fair measure of how much business activity a state has. It also means part of the count is one-person firms, which matters if you are sizing a market for something a one-person firm rarely buys. The profile does not say what that share is on the page we checked, so we do not quote one.
What the Census figure counts
County Business Patterns gives yearly data for businesses with paid employees (Census Bureau). The Census Bureau defines an establishment as a single physical location at which business is conducted or services or industrial operations are performed (CBP methodology). An establishment is a place, not a company. A regional chain with eleven branches contributes eleven.
The count leaves some things out. The programme excludes crop and animal production, rail transportation, the Postal Service, private households and most government employment (About CBP).
If your customers are companies with staff, this is the closer count, because every location in it has a payroll.
Photo: Audy of Course on Pexels.
How to use these numbers
- Use them to size a market, not to choose one. A large count means demand exists somewhere in it. It does not show whether you can win any particular search inside it.
- Keep the two sources apart in any calculation. Dividing an SBA business count by a Census establishment count gives a number with no meaning.
- Check the year the data describes before quoting a figure, and quote that year, not the year of publication.
- If you are using market size to justify location pages, look at the search results for the target city first. In our experience that takes about ten minutes, and it answers the real question, which is whether a page can rank.
Does a bigger market mean you should build a page for it?
No. A business count tells you firms exist in a place. It does not tell you how many of them compete for the search you care about, and it does not tell you whether you have anything to say about that place.
This is why the data is on this site. We were asked to build three hundred city pages for a client, declined, and were asked to justify the decision with something better than an opinion. The counts show the markets are real and large. They cannot show that a page per city is worth having.
Google’s spam policies name the risk. Among its examples of doorway abuse is having multiple pages targeted at specific regions or cities that funnel users to one page (Google Search Central). A set of pages that differ only by the city in the title fits that description closely.
The check we use is the search itself. Type the term with the city name in it and read what is ranking. If the results are national directories and companies with no local presence, there may be an opening. If they are established local firms with real pages, a new page whose main feature is the city name in its title has little to compete with. A metro with two hundred thousand establishments could have a handful of firms seriously competing for your term, and a market a tenth of that size could have dozens, because the category is different.
Why is the newest Census figure two years old?
Because it is a full count built from records, and that takes time. The 2023 figures were released on 26 June 2025 (Census Bureau), and the SBA’s 2025 profiles rest on 2022 data.
County Business Patterns is drawn from the Census Bureau’s Business Register, which holds records for establishments with paid employees and is updated from administrative records and Census surveys. The Census Bureau says the data are not subject to sampling error for that reason, though they are subject to other errors, such as missed or misclassified establishments (CBP methodology). It also adds statistical noise to employment and payroll figures to protect confidentiality, so treat the employee column above as very close, not exact.
For deciding which markets deserve serious work, we would rather have a full count from two years ago than a modelled estimate from last month. That is our preference, and for a fast-changing market you may weigh it differently.
Where the figures come from
Two government sources, kept separate because they count different things. The state figures come from the SBA Office of Advocacy’s 2025 Small Business Profiles. The metro figures come from the Census Bureau’s County Business Patterns for reference year 2023. Neither figure was collected, adjusted or modelled by us. The only numbers of our own are the percentage shares in the state table, which are simple division.
Only the rows we have checked against the published source are listed, which is why the tables are shorter than they could be. The state figures were checked against the SBA release and the metro figures against the Census metropolitan area file.
What these figures do not tell you
- The two sources count different things and cannot be compared with each other. A single company can be one business and several establishments.
- Both describe a year that has already passed: 2022 for the SBA profiles and 2023 for County Business Patterns.
- Market size says nothing about how hard a search is to win.
- Neither source says anything about search demand. Business counts and search volume measure different things, and treating one as a stand-in for the other is how location page programmes get justified.
Last checked against the sources on 2026-10-02.
The bottom line
A business count is a fact about an economy. Whether a page should exist is a question about the reader and the search results. Use the first to decide where to look, and the second to decide what to build.
Questions people ask about this
Why are there no pages on this site for individual cities?
Because we have not yet got something specific to say about most individual cities that is not already on the parent page. A page whose only difference from another is a city name is the kind of page Google’s spam policies describe under doorway abuse. Building hundreds of them is a risk we would not take with a client’s site. Where there is genuine local substance, meaning market conditions, named local competition and work actually done there, a page gets built.
What is the difference between a business and an establishment?
A business is a company. An establishment is a physical location with paid employees. A firm operating from six branches is one business and six establishments, and a self-employed consultant with no employees is one business and no employer establishments. This is why the SBA and Census figures on this page are so different in scale, and why they must not be combined.
How current is this data?
The SBA profiles were published in June 2025 and are built on Census data for 2022. County Business Patterns is published for reference year 2023, released in June 2025, and that was the newest year available when we last checked on 2026-10-02. The lag comes from how the data is compiled from records and surveys of every employer, not from a sample.
How do I check whether a city page can rank before building it?
Search the exact term a customer would type, with the city name in it, and read the first page of results properly. Note what type of business or site holds each position and whether it has a genuine local presence. If the results are national directories, aggregators or firms based elsewhere, there is room. If they are established local companies with real pages, reviews and addresses, a new page with the city in the title has nothing to beat them with.
Should I use the SBA or the Census figure to size my market?
It depends on who you sell to. If your customers are companies with staff, the Census establishment count is the closer fit, because every location in it has at least one employee. If you sell to sole traders and the self-employed, the SBA count is the one that includes them. Pick one, state which you used, and never divide one by the other.
Can I estimate search demand from business counts?
No. Business counts measure how many firms exist. Search demand measures how many people type a phrase into Google each month, and the two move independently. A city can have thousands of establishments and almost no searches for your category, or a small population with heavy demand because of one local industry. Use a keyword tool or your own Search Console data for demand, and use the counts only for the size of the economy.
What if the metro I want is not in the table?
Go to the source rather than to a blog that has copied it. County Business Patterns lets you pull establishment counts for any metro or county, and the SBA profiles cover every state and territory. Check the reference year on the page you land on before you quote anything. If you only need a rough sense of scale, the table above tells you where your market sits relative to the largest ones.
Does a large establishment count mean more competition?
Not for any specific search. Competition for a term depends on how many businesses in your category are investing in ranking for it, which the count does not show. The check is the search itself. Two markets of the same size can have wildly different results pages, one full of directories and one full of strong local firms. Size tells you the prize might be large. It says nothing about the price of winning it.
Is a smaller market ever a better bet than a large one?
It can be. A large metro with strong local competitors can take years of authority work to crack, while a smaller city may have nobody serious ranking for your term. The right question is where you can win a meaningful search soonest, then whether the demand there is worth the effort. Run the results check on both, compare what is ranking, and start where the gap between you and the current page one is smallest.
Should I quote these figures in a proposal or pitch?
Yes, if you name the source, the reference year and what is being counted. Say “employer establishments, Census, reference year 2023” rather than “businesses in Chicago”. The credibility comes from the precision. A round number with no source is easy to doubt. A reader who knows the data may discount the whole document once they spot one figure being used loosely.
How often should I recheck the source data?
Once a year is enough for a planning figure, because both programmes publish once a year. Recheck before you reuse a figure in anything public, since a new reference year may have been released. The date you last checked belongs next to the figure, as it does on this page, so anyone reading later knows how fresh the evidence was.
Sources for every figure above
- New Advocacy Report Shows the Number of Small Businesses in the U.S. Exceeds 36 million, US Small Business Administration, Office of Advocacy, 30 June 2025.
- 2025 Small Business Profiles for the States, Territories, and Nation, US Small Business Administration, Office of Advocacy, 30 June 2025.
- County Business Patterns: 2023, US Census Bureau, metropolitan area file, released 26 June 2025 for reference year 2023.
- County Business Patterns methodology, US Census Bureau.
Free review
Want this looked at on your own site?
Send the address. A person reads it and writes back with the two or three things that would move your enquiry number first. Free, and yours to keep whether you hire us or not.