What Are Firmographics? Definition, Examples and Sources

Firmographics are the descriptive facts about a company that you use to sort organizations into groups. The main ones are industry, size by employees or revenue, location, legal status, ownership and age. They do for companies what demographics do for people. "Firmographic" is the adjective, as in a firmographic field. "Firmographics" is the noun for the whole set.

Firmographic segmentation means grouping target accounts that share those attributes. A segment might read: civil engineering contractors in England with 50 to 249 employees, privately owned and outside any larger group.

The most dependable free sources are official ones:

• In the US, the NAICS industry codes and the Census Bureau's business statistics.
• In the EU, the NACE classification and the legal SME definition.
• In the UK, UK SIC 2007 and the Office for National Statistics (ONS).

These sources give you definitions and counts. None of them gives you a single company name, so the list itself comes from a data tool or a company register that you query separately.

Why do firmographics matter so much for outbound?

When a founder tells me their cold email gets no replies, I ask for the list first. The email can wait. Why? The list decides more than the copy. Whether you sell is mostly settled before anyone opens the message.

Copy gets the blame because you can see it and edit it. Nobody has opened the spreadsheet underneath since the import. I have seen the exact same message sent for two different companies. One got 2% replies. The other got 7.5%. So the words were never the whole story. No copy can save a bad list.

Firmographics are the first filter on that spreadsheet. If the filter lets in companies that are too small, in the wrong industry or in the wrong country, no subject line fixes that. A three-person firm will not buy a project scoped for a company of 500, however well you write to it.

Which firmographic fields exist, with an example of each?

Seven fields carry most B2B targeting. The table gives each one an example value, a free source and the trap that catches most lists.

Industry codes: NAICS, NACE and UK SIC

NAICS is the standard US federal statistical agencies use to classify business establishments. It was adopted in 1997 to replace the older SIC system. It was built jointly with Canada and Mexico so the three countries' statistics can be compared (US Census Bureau, NAICS). Plenty of data tools still show SIC codes, so check which system your filter is using.

The EU equivalent is NACE. Its newest version, NACE Rev. 2.1, applies to European statistics from 2025 onwards. Statistics for 2008 to 2024 use NACE Rev. 2 (Eurostat). An EU dataset that runs across that line can mix two versions of the same code list.

The UK classifies businesses by UK SIC 2007 (ONS, UK business activity, size and location).

Every code system shares one weakness: a code describes the main activity only. Take a contractor that also hires out equipment: it will usually sit under construction, so add a second code to your filter when your buyers do two things.

Size: whose definition of "small"?

Two legal definitions get quoted as if they were sales segments. They are program rules, written to decide who qualifies for support or contracts, and they count people differently.

The EU's SME Recommendation of 2003 sets these ceilings (EUR-Lex, Recommendation 2003/361/EC):

Two details change who counts as small:

• Headcount is measured in annual work units, so part-time and seasonal staff count as fractions.
• A company changes category only after it crosses a ceiling in two consecutive accounting periods.

The European Commission says SMEs make up 99% of EU businesses (European Commission, SME definition). So the label "SME" on its own filters out almost nothing.

In the US, the Small Business Administration (SBA) sets a size standard for every NAICS code. Most manufacturers with up to 500 employees count as small, and so do most non-manufacturers whose yearly receipts, averaged, stay below $7.5 million.

Employees are averaged over each pay period of the latest 24 calendar months. For federal contracting, receipts are averaged over the latest five complete fiscal years (SBA size standards).

The two systems can disagree about the same company. A manufacturer with 260 people falls outside the EU SME definition. It can still be a small business under the SBA rule for its code.

Decision rule: pick one size definition, write it next to your filter, and never compare counts built on two different ones.

How fresh is official firmographic data?

Expect SUSB figures to trail the year they describe by roughly two to two and a half years. Receipts are published only for reference years that end with 2 or 7 (Census SUSB).

Official data is good for sizing a market and designing a filter. It is the wrong place to look for whether one company still has 120 staff. Inside that two-year gap a single company can hire, shrink, move or be bought.

Why do firm and establishment counts never match?

Ask three sources how many companies sit in your segment and you get three numbers. Usually all three are right, because each counts a different unit.

CBP counts establishments: single physical locations, grouped by the employment size of each location, down to ZIP code.

The Bureau lists uses such as analyzing market potential and setting sales quotas (Census CBP). The page also carries a notice that its information is no longer current while the Bureau reviews its disclosure-protection method. Read the method notes before you rely on a new release.

SUSB counts firms and establishments, grouped by the size of the whole enterprise. Take a company with 40 branches of 50 people each:

• In SUSB it is one firm of 2,000 people.
• In CBP it is 40 establishments of 50.

If your filter says "50 to 99 employees" and your source counts sites, that one company can land in your mid-size segment 40 times.

NAICS classifies establishments, so a company's factory, warehouse and head office can carry three different codes.

The UK has the same split under different names. The ONS defines an enterprise as essentially a business. A local unit is an individual site, such as a shop or a factory.

Ownership shifts the count too:

• US. Under the SBA rules, affiliates count toward a firm's size, and the power to control exists from 50% ownership.
• EU. An investor with a stake of at least 25% in the capital, or in the voting rights, makes the two firms partner enterprises. A firm inside a group may need to add the group's headcount and turnover.

Decide first whether your filter describes the buying entity or the site. Then count everything in that unit and nothing else.

For most B2B offers the buying entity is the enterprise, because one contract covers every site. When each site orders and pays on its own, as with local maintenance contracts, the site view fits better.

How do firmographics compare with demographics, technographics, psychographics and intent?

The difference between demographics and firmographics is the unit. Demographics describe the human reading your email. Firmographics describe the company that would pay the invoice.

In the nested model Bonoma and Shapiro published in 1984, industry, company size and customer location form the outer, macro layer. Company size works best when combined with other factors (Wikipedia, Industrial market segmentation).

Which data layer answers which question?

The last row needs the most care. Intent data earns a place in my process only if the total addressable market is big. Picture a market of about 2,000 target companies where intent signals show up on about 10 of them. In that case I skip intent data and put the work into an offer the whole market cares about.

Three more decision rules:

• If your whole market fits in one spreadsheet, firmographics plus real titles are the list. The offer does the rest.
• If you sell an integration or an add-on, put technographics straight after industry.
• If the company data and the contact data point different ways, trust firmographics on fit and demographics on who gets the message.

A tight filter can still leave you with too few accounts. So once the filter is set, count what it returns before anyone writes a sequence.

My rule: you need the right people and enough of them. Sending to 50-100 perfect-fit leads a month is a coin flip, because even an ideal fit can be busy, not ready or not interested. My planning number is at least 10,000 people contacted each month for every 30-35 booked sales calls, and the real figure varies by industry. If your count falls far short of that, widen the filter.

What can firmographics not tell you?

Firmographics tell you what a company is. They say nothing about need, timing, budget or who decides. That gap is where a list with a perfect filter can still fail.

Take two contractors with the same code, headcount and city. They can still be in opposite positions. One signed a five-year supply deal last month. The other is unhappy with its supplier and reviewing it this quarter. Neither fact lives in a firmographic field.

The Ehrenberg-Bass Institute's summary of the 95:5 Rule says only about 5% of target B2B buyers are actively looking at any given time. Forrester argues the share can run higher: in one category it measured, roughly 20% of B2B organizations plan to change their primary marketing technology provider. Either way, most of a firmographic match is not buying right now. Adding technology data, covered in technographic data providers, narrows the match further, and lookalike companies built from your best customers narrow it from the other direction.

A clean segment still needs one message. The least popular advice I give is to commit a single offer to a single audience and keep it there for 12 months. A message built to fit every company your filter lets in reaches nobody in particular. That is why I argue for committing to a single offer and audience before anyone tunes a filter.

The filter is only half of an ideal customer profile, which also needs the buyer and the problem. And before you commit, count your target market to check the segment is big enough to hold 12 months of one offer.

How do you turn a firmographic filter into a list?

Run these five steps in order:

1. Start from your clients. I start every list with the client's past and current customers, then look for companies that match them. Fill in the seven fields from the first table for each customer and look for the columns that repeat.
2. Write the filter in one system and one unit. Use one industry code system, one size band counted one way, specific geographies and an explicit entity or site choice.
3. Add real titles per vertical. Go back through past deals, note who actually made each buying decision, and build the search string from those titles, because a plain "manager" says nothing about who signs. I remove generic titles such as Assistant, Manager and Associate on purpose.
4. Verify before sending. Run every address through a verification check and drop the ones that fail before the first email goes out.
5. Split into tracks. Independent firms and group subsidiaries, or two size bands, rarely want the same message. Give each track its own message.

Here is what steps 2 and 3 look like for a supplier of temporary site equipment selling to UK contractors:

A disclosure, since this is the work my agency does. My team builds these lists from each client's ideal customer profile (ICP) for cold email and LinkedIn campaigns. We do not sell a firmographic database, and the client pays for data tools directly.

Open your CRM this week. Export the last 20 clients you signed and fill in the seven fields for each one. Wherever those clients and your current filter disagree, trust the clients. Your filter is a guess. They signed.

Unsure which firmographic filters belong in your list? Book a call and bring your best customers. We will look for what they share.