Two companies land in your database on the same day. One is a 40-person software startup in Austin. The other is a 12,000-employee manufacturer in Ohio. Same lead source, same form fill, wildly different value to your business. The thing that tells them apart is firmographic data.
Firmographics describe what a company is, the way demographics describe a person. Most B2B marketers track the same handful of attributes to sort, score, and segment their accounts. Five of them do the heavy lifting.
Track these well and your targeting gets sharper. Track them without a plan and you are just hoarding fields nobody uses.
What counts as a firmographic attribute
A firmographic attribute is a fixed trait of a company itself, not its behavior and not its tech stack. Behavior is intent data. Tools are technographic data. Firmographics sit underneath both, describing the organization you sell to. Get them right and everything downstream, from how you segment a database to how you route a lead, has a solid base to stand on. Get them wrong and every play built on top inherits the error. Here are the five that matter most.
1. Industry and vertical
Industry is usually the first filter marketers reach for, and for good reason. A company’s sector shapes its budget, its buying process, its regulations, and the language that lands in a campaign.
Most teams tag industry with a standard code, NAICS or SIC, then group those codes into the verticals they actually sell to. The trap is stopping at the top-level label. “Healthcare” covers a rural clinic and a national insurer, and those two need completely different messages. Track the granular code, then roll it up when you need to. A campaign written for “financial services” reads as generic to both a credit union and a hedge fund. The specific code is where the useful targeting actually lives.
2. Company size
Company size, measured by employee headcount, is the most reliable single predictor of fit for most B2B products. Headcount tracks with complexity, budget, and how many stakeholders sit in a deal.
It also tends to be more accurate than revenue, because employee counts are easier to verify from public profiles and harder to fudge. A 50-person company and a 5,000-person company buy differently, staff differently, and need different things. Banding companies into size tiers, then tailoring the offer to each tier, is one of the cleanest segmentation moves available.
3. Annual revenue
Revenue is the attribute everyone wants and the one most often wrong. Private companies rarely publish exact figures, so a lot of revenue data is estimated. Treat it as a range, not a fact.
Even so, revenue earns its place. It maps to spending power and deal size in a way headcount alone can’t. A lean 30-person firm with high revenue per employee is a very different buyer than a 30-person firm scraping by. Use revenue to sanity-check size, not to replace it.
4. Location and geography
Location covers headquarters, regional offices, and the markets a company operates in. It drives more than you might expect. Time zones affect outreach timing. Regional rules affect what you can promise. Language and currency affect the campaign itself.
For teams with field sales or territory structures, location is the attribute that assigns accounts to reps. For everyone else, it shapes localization and compliance. And if you sell into regulated markets, a company’s operating regions can decide whether a deal is even possible.
5. Ownership and company structure
The fifth attribute gets overlooked, which is a mistake. Ownership and structure cover whether a company is public or private, independent or a subsidiary, venture-backed or bootstrapped, and where it sits in a corporate family tree.
This matters because it changes who holds the budget. A subsidiary might route buying decisions through a parent. A newly funded startup often has cash to spend and pressure to grow fast. A recent acquisition may be mid-consolidation and frozen on new spend. Growth stage and ownership tell you not just whether a company can buy, but whether it is in a position to.
How to combine the five into a usable segment
Any one of these attributes on its own is weak. The power is in the combination, layering them into a filter that describes your actual best customer instead of a vague wish list.
A simple ICP filter
Say your closed-won data shows your best accounts share a pattern. You would build a segment like this.
- Industry, software and technology services
- Company size, 200 to 2,000 employees
- Annual revenue, 25 million and up
- Location, North America
- Structure, independent or recently funded, not a subsidiary
That filter turns a bloated database into a target list you can actually work.
Weighting the attributes
The five are not equally predictive for every business. Figure out which ones separate your winners from your losers, then weight your scoring accordingly. If size drives your deals, size gets the most points. If geography constrains you, location becomes a hard filter rather than a soft signal. Let your own win data set the weights, not a template. Rerun that analysis once a year. The attribute that predicted deals last year can lose its edge as your product and your market shift.
Common mistakes when tracking firmographics
A few habits quietly wreck an otherwise good dataset.
Letting the data go stale. Companies grow, get acquired, relocate, and change stage. A record from three years ago describes a company that may not exist anymore.
Treating estimates as facts. Revenue and sometimes headcount are best guesses on private companies. Build your rules to tolerate ranges.
Collecting fields you never use. If an attribute never changes a decision, it is clutter. Track what you act on.
Get these five right, keep them current, and combine them with intent behavior, and your segmentation stops being guesswork. You spend your budget on the companies that look like the ones already buying from you.
Build target lists on firmographic data you can trust
HG Insights maps firmographic, technographic, and install data across millions of companies so your segments reflect reality, not stale records. Explore HG Insights