How US Census and BLS data size a market before you buy a report

What to take away

Why free federal data beats a bought report for a first market size

A bought report answers someone else's question. Federal data answers yours, at the geography where you plan to sell.

Most off-the-shelf reports cover a national market or a handful of large metros. If you are opening in Boise, Spokane or Fort Worth, the report's number does not describe your buyers.

Federal series are reproducible. You can hand a lender, an SBA district office counselor or an angel investor the exact table you pulled and the filters you used.

A paid report hides its method. You cannot tell whether its market size came from a survey of forty people or a top-down guess built on a press release.

Cost is the smaller advantage. Public data lets you test a claim, then test it again next quarter when the new vintage lands.

Start with the free sources. The market validation checklist walks through that sequence from first question to sign-off.

One caveat: federal data is backward-looking. It describes what existed, not what will exist. That is enough for a first size and often enough for a decision.

data.census.gov: pulling establishment and population counts by geography

data.census.gov is the Census Bureau's main access point. It replaced American FactFinder and holds the Economic Census, County Business Patterns and the American Community Survey in one search box.

For market sizing you want two different things from it: businesses and people.

County Business Patterns reports establishments by industry and geography. An establishment is a single physical location, so a chain with twelve stores counts as twelve.

The American Community Survey gives you households, age bands, income brackets and commuting patterns at the same geography.

Pulling a first count

The path below needs no account and no fee.

For the household side, search the American Community Survey with the same geography. Open table B11001 for household counts, B19013 for median household income, and S1901 for the income distribution. The ACS program page is https://www.census.gov/programs-surveys/acs.html.

County Business Patterns is described at https://www.census.gov/programs-surveys/cbp.html. Both programs publish county and metro tables.

What the two numbers give you

Divide establishments by population to get business density. A county with far fewer establishments per resident than its neighbors may be underserved, or may have different buying habits.

Multiply households by the share you think would buy, and you have a rough revenue ceiling. That share is an assumption, so write it down and label it as one.

Do not confuse establishment counts with firm counts. County Business Patterns counts locations. A ten-location firm appears ten times.

For a first pass that is fine. You are checking whether the buyer pool is in the thousands or the dozens, and whether it is growing or shrinking.

If you want to compare that quick estimate against a full build, the top-down market sizing comparison shows where each method breaks.

The Annual Business Survey and what it says about firms like yours

The Annual Business Survey is a joint Census Bureau and National Science Foundation product. It covers employer firms and reports on owners, research and development, technology use and business practices.

Its value is not size. It is character. The survey tells you what the firms already in your sector look like.

You can see the age distribution of firms, the share that are family owned, the share with female or minority ownership, and the revenue classes they fall into.

That shapes strategy. If most firms in your sector are ten years old with fewer than five employees, you are entering a fragmented market with low barriers and thin margins.

If the sector is dominated by firms with more than a hundred employees, your plan needs a niche, a service angle or a partner channel.

The survey also reports why firms did not seek financing and what obstacles they named. Read those tables before you assume capital is your main constraint.

Access runs through the Census Bureau's data site, and the tables are downloadable. The ABS program page is https://www.census.gov/programs-surveys/abs.html. Coverage is employer firms only, so sole proprietors without payroll do not appear.

Pair the survey with your own numbers. For a second market in Canada, where none of these US series apply, the canadian small business grants idea validation approach explains how to derive a benchmark from your own pipeline instead of borrowing one from a report.

One practical note: the Annual Business Survey releases lag. Check the vintage on every table before you put a figure in a pitch deck.

BLS employment and wage series as a demand proxy

The Bureau of Labor Statistics collects employment and wage data that works as a demand proxy. If an industry employs people in a metro, those people earn paychecks, and paychecks become spending.

The Bureau publishes an overview of its statistics that explains which series cover what, which is the fastest way to avoid pulling the wrong table. The Overview of BLS Statistics is worth reading once before you build a spreadsheet.

The two series you will use most

First, the employment series. It tells you how many people work in an industry, in a state or metro, and whether that count is rising or falling over the last several years.

Second, the wage series. It tells you what those jobs pay, by occupation and by area. Average pay in a metro tells you whether your price point is plausible.

The Occupational Employment and Wage Statistics program reports wages by occupation for metros and states. It is the series most founders should start with. OEWS tables are at https://www.bls.gov/oes/.

The Current Employment Statistics program reports monthly employment by industry for states and metros, which is better for spotting a trend. CES data is at https://www.bls.gov/ces/.

Turning employment into demand

Take the employment count for your target industry in your metro. Multiply by average annual wages for the occupations you serve. That product is the wage base.

Now apply a share: what fraction of that wage base could plausibly flow to your product in a year? Even a small share of a large base can be a real business.

Here is the arithmetic with placeholder numbers. Replace them with the counts you pull. The method is what matters.

Say the market is the Boise City, Idaho metro and the sector is NAICS 722511, full-service restaurants. Suppose the pull returns 620 establishments and 12,400 jobs at an average annual wage near $32,000. The wage base is 12,400 times $32,000, or about $397 million a year.

Apply a share. At 0.5 percent, the revenue ceiling is about $2 million a year. Split across 620 establishments, that is roughly $3,200 per location per year. That figure tells you whether a product priced near $3,200 per location is worth building.

Now replace the assumed share with evidence. If your own pipeline shows ten buyers per thousand contacts, use that rate instead of a round number.

The method has limits. Employment counts include people who will never buy from you, and wage data is an average, not a distribution.

BLS also publishes industry-level summaries that group related series, so you can size a sector without hunting table by table. The Overview of BLS Statistics by Industry lists what exists for each sector.

If you know the series name but not its code, the A-Z Index is the fastest route to the right table.

Before you commit to a geography, confirm what the agency actually measures. The About the U.S. Bureau of Labor Statistics page states the mission and the scope of the data, which is useful when a lender asks where your numbers came from.

BLS establishment birth and survival rates as market entry risk

Demand and risk are different questions. A market can be large and still be a bad place to open, because new firms there die quickly.

The Business Employment Dynamics program tracks establishment births, deaths, expansions and contractions. Its survival tables show what share of new establishments are still alive after one, two, five and ten years.

The BDM Home page is the entry point for those tables, and the survival data is broken out by sector and by state in places. The headline figures are published as Table 7, Survival of private sector establishments by opening year.

Reading survival rates correctly

A low one-year survival rate in your sector is a warning about entry conditions, not a verdict on your idea. It often signals low capital requirements and easy entry, which cuts both ways.

A high survival rate usually means high barriers: licensing, equipment, leases or incumbency. That protects you once you are in and makes getting in harder.

Compare your sector's survival rate to the all-sector average. A sector well below average deserves a written explanation in your plan.

Look at births as well as deaths. A sector with many births and many deaths is churning. A sector with few of both is stable and probably consolidated.

Use the numbers to set expectations. If half of new establishments in your sector are gone within five years, your plan should say what makes yours different, in specific operational terms.

This is the part of market validation data that founders skip, and it is the part that most changes a plan. The market validation tools comparison covers which sources carry this kind of risk data and which do not.

Reconciling Census and BLS numbers when they disagree

The two agencies will disagree, and that is normal. They measure different things with different methods.

Census establishment counts come from business registers and surveys of employer firms. BLS employment comes from a monthly survey of establishments plus administrative records.

Census counts locations. BLS counts jobs. One firm with three locations and forty workers is one firm, three establishments and forty jobs.

Census coverage of small employer firms depends on the program. BLS coverage depends on unemployment insurance records, which captures most employers but handles some sectors differently.

A short reconciliation routine

When the gap persists, prefer the source whose unit matches your claim. Sizing a customer count points to Census. Sizing a payroll or wage base points to BLS.

Document the reconciliation in one paragraph. Anyone reviewing your plan will ask, and a written explanation is worth more than a precise number you cannot defend.

If you have not yet confirmed that money is moving in the category at all, settle that first. The question of whether buyers already validate the market with spending matters more than the exact size.

When a paid report is still worth buying

Free data has three real gaps: it lags, it stops at the sector level, and it says nothing about your specific competitors' pricing or unit economics.

Buy a report when you need one of those three things and cannot get it another way.

Buy when you need granular competitor pricing, channel margins or a forecast that a lender or investor has asked for by name. Forecasts are the one thing federal data does not produce.

Buy when the report's methodology is disclosed and you can see the sample, the dates and the definitions. A report that hides its method is not worth the price.

Do not buy before you have pulled the free numbers. You will not know which questions to ask, and you will pay for pages that restate what data.census.gov already gave you.

Do not buy a national report to size a single metro. The geography mismatch is the most common wasted purchase.

Set a budget rule. If a report costs more than the first month of revenue you project from the market it describes, wait until the free data narrows the question.

For most first-time US founders, the sequence runs from Census counts to the Annual Business Survey, then BLS employment and wages, then survival rates, then reconciliation.

Common questions

Is data.census.gov free to use?

Yes. It is a public Census Bureau site with no account required for most tables. You can download data as CSV or Excel.

What is the difference between the Annual Business Survey and County Business Patterns?

County Business Patterns counts establishments and employment by industry and place. The Annual Business Survey describes characteristics of employer firms, such as ownership, age and business practices.

Can I size a market with BLS data alone?

You can size a wage base and an employment trend, which is a demand proxy. For a customer count you generally need Census establishment or population data as well.

How old is the newest federal data?

It varies by series. Economic Census data appears every five years, the Annual Business Survey lags by roughly a year or more, and BLS employment series update monthly or quarterly.