How to Size Your TAM With Live B2B Data
A reproducible method for turning an ICP into account, contact, and reachable-market counts, with dated filters, zero-cost iteration, exclusions, and scenario ranges.
Reviewed and updated August 20, 2026 by Argorant
A defensible TAM is not a single database count. Define the market as filters and exclusions, count companies and buyer roles separately, compare conservative, base, and broad scenarios, and save the exact query with a date. Argorant count and masked-preview tools can test the population before any contact output is purchased.
What is actually supported
These labels separate a code-backed direct path from documentation, generic HTTP, CSV, and workflows that still require an end-to-end test. A marketing logo alone does not count as support.
Live count
DirectUse Argorant search and count tools to measure a current market without revealing contacts.
Masked validation
DirectPreview representative rows to check filter interpretation before using the count.
Public registry cross-check
RESTUse Companies House or SEC EDGAR APIs for bounded official-universe checks where applicable.
Reproducible workflow
- 1
Write an ICP as explicit filters
Define industry, geography, company size, keywords, exclusions, buyer titles, and seniority.
Expected result: A query another analyst can reproduce. - 2
Count companies and people separately
Measure target accounts first, then relevant roles per account.
Expected result: Account TAM and contact TAM are not conflated. - 3
Inspect the edges
Preview rows from each region and size band, then adjust ambiguous filters.
Expected result: The count represents the intended market rather than label noise. - 4
Publish scenarios and date
Save conservative, base, and broad counts with query definitions and access date.
Expected result: A transparent range that can be refreshed later.
Inspect the market before a paid action
The result count is the first control point. It lets an operator or agent confirm that the interpreted filters produce a plausible market before revealing or exporting contact data.
Plans, credits, and access gates
| Provider or layer | Restriction to model | Published price where verified |
|---|---|---|
| Argorant | Count and masked preview should occur before paid reveal or export; confirm the account's current non-spend tool limits. | Confirm in the current account or quote. |
| Public registries | Registry records can validate bounded universes but do not automatically supply current buyer roles or verified work emails. | Confirm in the current account or quote. |
Best fit by job
| Option | Best for | Access | Cost model | Verdict |
|---|---|---|---|---|
| Argorant count | Current B2B company and role population | UI, MCP, API | Discovery before paid output | Primary operational TAM layer. |
| Companies House | UK registered-company cross-check | Official API | Public API terms | Useful boundary check, not a replacement for buyer contact data. |
| SEC EDGAR | US filer cross-check | Official submissions and company facts APIs | Public API terms | Useful for public companies, not the full private-company universe. |
Verification gates
- Re-run every published scenario and capture query JSON, count, timestamp, and preview.
- Manually review a stratified set of rows from every geography and size band.
- Confirm count endpoints do not consume contact-output credits in the tested account.
Claims this guide does not make
- Do not present one live count as a permanent audited market size.
- Do not mix account count and person count.
- Do not use a registry subset as the full private B2B universe.
Frequently asked questions
Is contact TAM the same as account TAM?
No. Account TAM counts companies; contact TAM counts relevant people within them. Report both and avoid multiplying by an assumed contacts-per-account ratio when live role counts are available.
Why save the access date?
Companies, roles, and source coverage change. A dated query makes the number reproducible and enables a scheduled refresh without pretending the count is permanent.
Sources reviewed
Official product, developer, pricing, and policy pages reviewed on August 20, 2026. Test-gated behavior remains labelled until a live end-to-end run proves the current account and plan path.
4 sources checked and datedShow sources
Continue the workflow
Prove the workflow
before you scale it.
Count and preview first, then authorize verified output behind a visible spend boundary.
