B2B Market Sizing Using Lead Data: From TAM Estimation to Account Selection

Lead data helps estimate how much of a defined audience a source can represent. It cannot, by itself, establish total market size, purchase demand or the revenue your team will win.

Keep three outputs separate: the external market estimate, observed database coverage and the account workload your team can execute. Connect them through documented assumptions rather than treating a contact count as a company count.

Specify the product use case, eligible customer, geography, period and economic unit. Decide whether the model counts legal entities, business groups, locations, accounts or contact people.

The Census Bureau's SUSB glossary distinguishes an establishment, a physical business location, from broader business units. Its employer-business coverage also differs from nonemployer statistics. Choose external evidence that matches your intended audience rather than mixing incompatible totals.

A group with several offices can be one purchasing account or several relevant operating units. State your decision and use it consistently in the market estimate and account list.

Total addressable market, or TAM

The full potential market under the product and customer definition. When expressed as revenue, it needs a price, usage or spending model and a stated period.

Serviceable available market, or SAM

The part of that market the product and distribution can serve under actual geographic, product and operating constraints.

Serviceable obtainable market, or SOM

A reasoned estimate of what the business could obtain over a specified horizon, considering competition, adoption, capacity and execution assumptions.

A count of accounts you can research this quarter is an operating capacity plan. It can inform an obtainable-market scenario, but it is not automatically SOM or a revenue forecast.

The TAM, ICP and segment guide explains the different planning decisions these terms support.

Use appropriate primary statistics, registries, sector data and documented customer research. Record the reference period, geography, industry classification, unit and exclusions for each source.

• Does the source include the sizes and types of businesses you serve?
• Are locations being counted as companies?
• Does the category include businesses outside your use case?
• Are parent groups and subsidiaries handled consistently?
• Does the evidence describe potential buyers, actual spending or expressed demand?

Industry classifications and broader reports can be useful even when they do not directly produce a campaign list. Investigate differences rather than assuming the contact database is a complete or superior census.

Translate the audience into supported search criteria and record the exact query and date. In Dievio, a free count preview measures matching contact records under those criteria.

Use title, industry, size and location criteria as appropriate. Stored company revenue and funding dates are supported where covered. Native technology-stack, funding-stage, hiring-growth and intent filters should not be assumed.

A controlled query comparison can show how a supported criterion affects the observed record count. It does not establish whether excluded or missing records represent unsuitable companies or incomplete data.

1. Retrieve records for a defined test. Document the selection and result limits; do not call a convenience sample representative of the entire market.
2. Resolve company identities. Normalize domains and distinguish parent groups, subsidiaries and locations using the chosen account rule.
3. Review fit. Record suitable, unsuitable and uncertain cases separately.
4. Assess role and field coverage. One contact at a company does not establish that the needed buyer role or channel is available.
5. Compare known accounts. Investigate why suitable independently identified accounts are present or absent.

Search retrieval uses one credit per returned lead; repeated retrieval can charge again. Downloading an existing saved list is free and does not refresh it. Plan a bounded test around the evidence needed rather than collecting the entire source to get a count.

The segment validation workflow and niche coverage guide develop these checks.

A historical opportunity win rate describes qualified opportunities, not all cold accounts in a database. Applying it directly to the contact preview overstates what the evidence supports.

A planning scenario can instead model the stages:

Expected wins = accounts actually worked × account-to-qualified-opportunity rate × opportunity-to-win rate.

Use rates measured on comparable cohorts and a suitable horizon. Multiplying expected wins by an appropriate contract-value assumption gives a conditional bookings scenario. It is not realized revenue, profit or proof of obtainable market.

Do not multiply several arbitrary “accuracy discounts” as though they are independent. Overlapping exclusions, conditional rates and repeated contacts can make that arithmetic misleading. Model observed transitions and show uncertainty.

Show lower, expected and higher cases for material uncertainties: eligible accounts, represented accounts, role availability, team capacity and conversion. Do not label invented values as industry benchmarks.

If an external source and your database test differ, compare definitions first. Geography, reference dates, business units, category boundaries, employer coverage and source omissions can all matter. A larger contact count does not resolve those differences.

Choose accounts supported by the evidence and the team's review capacity. Assign owners, preserve exclusions and record the research gaps. Keep contact-level and account-level counts visible throughout the handoff.

Revisit the model when product eligibility, geography, source coverage or observed conversion changes. Use a count preview to update the coverage layer, while retaining the independent evidence and assumptions behind the market estimate.