Key finding
Across 130,030 size-classified companies in a 933,544-company B2B database, 69.1% have 200 or fewer employees, 49.7% have 50 or fewer, and only 14.2% have more than 1,000; among located companies, 55.2% are in the United States.
What we measured
TAM slides are built on company databases, and company databases are rarely audited in public. This study audits one: the 933,544 company records behind Reachium's B2B data, with exact counts of what the records actually contain, coverage gaps first.
It is the company-side companion to our contact-side audit, and the ground truth behind two findings elsewhere in this research: why "C-Suite" tags are mostly small-company owners, and what universe outreach targeting skews against.
Methodology
Exact counts, single snapshot, taken 27 August 2026 on the post-deduplication database.
Coverage before composition. Not every company record carries every field, and composition claims are honest only against their covered denominator. The coverage table below is therefore the first finding, not a footnote: in particular, only 13.93% of companies carry an employee-count band, and every size share in this study divides by that 130,030-company subset, never by 933,544. Enrichment is not random, so the sized subset is not a random sample of the whole; the direction of its bias is discussed under Limitations.
The findings
Field coverage across 933,544 companies:
| Field | Companies covered | Coverage |
|---|---|---|
| Website domain | 517,863 | 55.47% |
| Country | 500,686 | 53.63% |
| Industry | 305,814 | 32.76% |
| Employee-count band | 130,030 | 13.93% |
Size distribution, of the 130,030 with a size band:
| Employees | Companies | Share |
|---|---|---|
| 1-10 | 31,695 | 24.38% |
| 11-50 | 32,928 | 25.32% |
| 51-200 | 25,191 | 19.37% |
| 201-500 | 13,621 | 10.48% |
| 501-1000 | 8,152 | 6.27% |
| 1001-5000 | 11,580 | 8.91% |
| 5001-10000 | 2,718 | 2.09% |
| 10001+ | 4,145 | 3.19% |
Geography, of the 500,686 with a country: United States 276,437 (55.21%), United Kingdom 31,335, Canada 21,809, Germany 17,566, France 13,700, with a long tail behind them.
Industry, of the 305,814 classified: Financial Services 91,121 (29.80%), Technology 43,138, Consumer 24,956, Business Services 24,095, Healthcare 22,466.
What this means
The B2B world in this database is small, American, and financial. Half of size-classified companies would fit in one meeting room, and the 1,000-plus enterprises that anchor most sales strategies are a seventh of the classified universe. That matches the real economy more than it matches TAM slides, which is the point: company counts in databases reflect how many companies exist, and most companies that exist are small.
The industry skew is a property of sourcing, and worth reading as one. Financial Services at 29.80% of classified companies does not mean a third of the economy is finance; it means this database's sources and its users' demand pull finance-heavy. Every vendor's database has an equivalent skew and almost none publish it; and among the outreach requests that match to a classified industry, 53.6% flow into Financial Services, so database composition and user demand are feeding each other.
The coverage table may be the most broadly useful part of this study. A company record with no size, no industry, and no country still counts toward a vendor's headline number. Here, the honest multipliers are on display: a 933,544-company database is a 130,030-company database if your segmentation needs a size band, and a 305,814-company database if it needs an industry.
In practice
A founder sizing a market from any database filter should check what coverage the filter silently imposes. In this database, filtering by employee count discards 86.07% of company records before the filter's own logic runs (100 minus 13.93, derived). A "5,000 companies match your ICP" result sits on top of that discard, and the true matching population is larger by some unknowable multiple.
A RevOps lead building territories on company size should expect the same coverage cliff in any vendor's data, and treat "size unknown" as its own large territory rather than noise: here it would be the biggest segment by far.
Limitations
Coverage bias runs toward notable companies. Enrichment prioritizes companies people look up and target, which skews the sized subset toward established firms. The true universe of companies is almost certainly smaller-skewed than even these numbers, so the 69.1% figure is more likely an understatement of small-company dominance than an overstatement.
Size bands are LinkedIn-style self-reports collected at enrichment time and not continuously refreshed.
One database, sourced for sales use. Composition reflects what a sales-data pipeline collects, not a census. National business registries show far more micro-companies than any sales database; this study's claim is about what B2B tooling contains, not about the economy.
Industry taxonomy is coarse, as in the industry acceptance study: broad classes, one per company.
Counts drift. Records are added continuously; these are exact counts for 27 August 2026 on the post-cleanup database, and are not comparable to any pre-cleanup figure.