Firmographic data is the demographic data of companies: industry, employee count, revenue, location, ownership structure, growth stage. It is the oldest and most widely used filter in B2B targeting, and the one most often applied badly.
The core fields
| Field | Typical use | Where it goes wrong |
|---|---|---|
| Employee count | Segmenting SMB, mid-market, enterprise | Counts include contractors and lag by months |
| Industry / SIC / NAICS | Vertical targeting | Self-reported codes are frequently wrong or too broad |
| Revenue | Deal-size prediction | Estimated for private companies, often badly |
| Location | Territory assignment | Registered office is not where the buyer sits |
| Growth stage / funding | Timing | The single most actionable field, and the most perishable |
Why headcount beats revenue for most teams
Revenue is estimated for private companies and those estimates carry wide error bars. Employee count is observable, updates more often and correlates well enough with budget for most segmentation. If you have to pick one, pick headcount, and pick a source that reports how recently it was refreshed.
Using firmographics without over-filtering
The common failure is stacking filters until the list is small and comfortable rather than large and correct. Every additional field multiplies the effect of that field being wrong. Three well chosen filters on accurate data beat eight on stale data, every time.
- Start with two or three fields you are confident in.
- Add a timing signal such as funding or hiring before adding another descriptive field.
- Check what share of your list is excluded by each filter. If one drops 80 percent, verify it before trusting it.
Firmographics work best combined with technographic data, which tells you what a company runs rather than what it is. For appending these fields to records you hold see B2B data enrichment, and for provider coverage see the best B2B data providers.
Each core field, and what it is actually good for
Employee count
The most useful single field. It is observable, updates relatively often, and correlates well enough with budget to drive segmentation. The traps: headcounts frequently include contractors and offshore teams, they lag reality by weeks or months, and a global figure hides the fact that the buying unit might be 30 people in one country.
Industry classification
SIC and NAICS codes are the standard, and they are frequently wrong. Codes are self-reported at registration and rarely updated, so a company that pivoted years ago still carries its original classification. Worse, the categories are too coarse for most B2B targeting: “Computer Software” covers an eight-person agency and a public enterprise platform equally. Where accuracy matters, a plain-language description of what the company sells beats the code.
Revenue
Reliable for public companies, estimated for everyone else. Private-company revenue estimates are typically modelled from headcount and sector, which means using them alongside headcount adds far less information than it appears to. Treat revenue as a rough band, never as a filter boundary.
Location
Registered office, headquarters and the location of the actual buyer are three different things. Territory assignment built on registered office produces reps calling the wrong country. Prefer office locations over legal registration where the provider offers both.
Funding and growth stage
The most actionable field, and the most perishable. A Series B raised last quarter implies budget, hiring and new initiatives. The same signal eighteen months later implies very little. Use funding as a timing trigger, not as a persistent attribute.
Building an ICP from firmographics without over-filtering
The common failure is stacking filters until the list feels comfortable. Each additional field multiplies the effect of that field being wrong, and the fields are not independent: headcount, revenue and funding stage all measure roughly the same thing.
- Start with two fields you trust. Usually headcount and a plain-language industry description.
- Add one timing signal, such as funding or hiring activity.
- Stop. A third and fourth descriptive filter rarely improves conversion and reliably shrinks the addressable list.
- Measure exclusion. If a filter removes 80 percent of your list, verify it on 20 accounts by hand before trusting it at scale.
A worked segmentation
| Segment | Firmographic definition | Why it holds together |
|---|---|---|
| Emerging | 20 to 100 employees, raised in the last 12 months | Budget exists, processes are not yet entrenched, one champion can decide |
| Core mid-market | 100 to 500 employees, no recent raise | Established process, committee buying, ROI case required |
| Enterprise | 500+ employees | Procurement, security review, long cycle, different motion entirely |
Three segments built on two fields will outperform twelve segments built on six, because each one is large enough to test and distinct enough to write different copy for.
Where firmographic data comes from
- Company registries and filings. Authoritative for legal entity data, slow and incomplete for anything commercial.
- Web and public profile scraping. Broad and current, inconsistent quality.
- Self-reported directories. Whatever the company chose to publish about itself.
- Modelled estimates. Revenue and sometimes headcount are inferred rather than observed. Providers rarely label which fields are modelled, and it is worth asking.
The question to ask every vendor
Not “how many records do you have” but “when was this field last verified, and was it observed or modelled”. Those two answers explain almost every quality difference between providers, and most vendors will answer them directly if asked.
Frequently asked questions
What is firmographic data?
Company-level descriptive data: industry, employee count, revenue, location, ownership and growth stage. The company equivalent of demographics.
What are the main firmographic fields?
Employee count, industry classification, revenue, location, and funding or growth stage.
Is revenue or employee count more reliable?
Employee count. Revenue is estimated for private companies with wide error bars, while headcount is observable and refreshed more often.
