What Is Market Sizing? #
Knowing how to calculate market size is the first real test of any business case. Market sizing is the process of estimating the total revenue opportunity available for a product or service within a defined geography and time period. Get it wrong and every downstream forecast — headcount, capital requirements, investor pitch — is built on sand.
Market sizing is not guesswork dressed in spreadsheets. It is a structured discipline that draws on establishment counts, payroll data, consumer expenditure surveys, and revenue aggregates — all of which are publicly available from federal statistical agencies. The difference between a credible market size estimate and a fabricated one is sourcing.
There are two broad approaches — top-down and bottom-up — and three nested market tiers you must define before running any numbers: TAM, SAM, and SOM. Each serves a different analytical purpose. A rigorous market sizing exercise uses at least one of these methods, anchored to a named federal data source, with explicit assumptions documented at every step.
This guide walks through each method in plain terms, shows where to find verified federal data, and works through a real example so you can replicate the process for any NAICS-classified US industry.
TAM, SAM, and SOM Explained #
TAM, SAM, and SOM are the three concentric rings of any market size calculation. Conflating them is one of the most common errors in business planning — and one of the most damaging to credibility.
- TAM (Total Addressable Market): The total revenue opportunity if your product captured 100% of the defined market with no competitive or geographic constraints. This is a theoretical ceiling, not a target.
- SAM (Serviceable Addressable Market): The portion of TAM your business model can realistically serve — constrained by geography, distribution channel, customer segment, or regulatory scope.
- SOM (Serviceable Obtainable Market): The share of SAM you can realistically capture given your current resources, go-to-market capacity, and competitive position. This is the number that should drive near-term revenue projections.
A venture-backed startup pitching a $50B TAM that projects $500M in Year 3 revenue is implying a 1% SOM — which may be reasonable or absurd depending on the competitive landscape. The point is that each number must be derived independently, not reverse-engineered from a desired outcome.
Federal data sources map cleanly onto this framework. Census Bureau County Business Patterns (CBP) data gives you establishment counts and payroll by NAICS code — a direct input for TAM. Filtering by state, metro area, or firm size narrows to SAM. Your own sales capacity defines SOM.
Want the full how to calculate market size data?
Complete data with 5-year forecasts, geographic breakdowns, and competitive analysis. Every data point sourced and cited.
Top-Down vs Bottom-Up Market Sizing #
The two primary market sizing methods differ in where they start — and both have legitimate uses. The choice depends on data availability, the maturity of the market, and how precise the output needs to be.
Top-Down Market Sizing
Top-down starts with a large aggregate — total industry revenue, national payroll, or consumer spending — and carves out a relevant slice using assumed percentages. It is fast and useful for early-stage scoping, but it compounds assumption errors quickly. If your top-level figure is wrong, every subsequent filter magnifies the error.
Example logic: US restaurant industry revenue (NAICS 722) × share attributable to fast-casual format × share in target metro = SAM estimate.
Bottom-Up Market Sizing
Bottom-up builds the estimate from the ground up: count the potential customers, multiply by average purchase frequency, multiply by average transaction value. It requires more data but produces a more defensible number because each assumption is explicit and independently verifiable.
Example logic: Number of NAICS-classified establishments in target geography (Census CBP) × average annual spend per establishment on target category (BLS or FRED benchmarks) = demand-side TAM.
For US industries, bottom-up analysis is significantly strengthened by BLS Quarterly Census of Employment and Wages (QCEW) data, which provides employment counts by industry and geography — a reliable proxy for establishment-level demand when revenue data is unavailable.
Using Federal Data Sources for Market Sizing #
The United States federal statistical system produces more granular, higher-frequency industry data than any private research firm. It is free, methodologically transparent, and updated on consistent schedules. The five sources below cover the full range of market sizing needs.
- Census Bureau — County Business Patterns (CBP): Annual establishment counts, employment, and payroll by NAICS code at the national, state, county, and metro level. Primary source for bottom-up sizing and geographic segmentation. (Census CBP, 2024)
- Census Bureau — Statistics of US Businesses (SUSB): Establishment counts and receipts by firm size and NAICS code. Essential for sizing markets segmented by business scale — e.g., SMB vs. enterprise. (Census SUSB, 2022)
- BLS — Quarterly Census of Employment and Wages (QCEW): Quarterly employment and wage data by industry and geography, covering 95%+ of US jobs. Strong proxy for demand intensity when direct revenue data is suppressed. (BLS QCEW, 2024)
- BLS — Occupational Employment and Wage Statistics (OES): Wage rates and employment counts by occupation and industry. Useful for labor cost benchmarking and workforce sizing within a market. (BLS OES, 2024)
- FRED — Federal Reserve Economic Data: Macro and sector-level economic series from the St. Louis Fed. Over 800,000 data series, including industry output, personal consumption expenditures, and credit conditions. (FRED, 2024)
A critical limitation: Census suppresses revenue data when fewer than three firms operate in a cell, to protect respondent confidentiality. When suppression occurs, QCEW employment data combined with industry-average revenue-per-employee ratios from SUSB provides a reliable workaround.
A Worked Example: Calculating Market Size with Census Data #
Let's build a TAM estimate for a B2B software company targeting HVAC contractors (NAICS 238220) in the United States. The goal: estimate total addressable annual spend if every qualifying establishment purchased the product.
Step 1 — Establish the Universe
Pull Census CBP data for NAICS 238220. The most recent release shows approximately 68,000 HVAC contractor establishments operating in the US. (Census CBP, 2022)
Step 2 — Segment by Firm Size
The product is priced for firms with 5–49 employees. Census SUSB shows roughly 28,000 establishments in NAICS 238220 fall within that employee band. (Census SUSB, 2022) This narrows the universe from 68,000 to 28,000 — your SAM establishment count.
Step 3 — Apply Revenue or Spend Per Unit
At a hypothetical annual contract value of $3,600 per establishment (a figure the selling company would set based on pricing), the SAM calculation is straightforward:
Step 4 — Define SOM
If the company has sales capacity to close 500 accounts in Year 1, SOM = 500 × $3,600 = $1.8M. That represents approximately 1.8% of SAM — a realistic early-stage capture rate for an unproven product entering a fragmented trade contractor market.
Every figure in this model traces back to a named federal source. That is what makes it defensible. VantaInsights pre-calculates establishment counts, payroll, employment, and firm-size breakdowns for 1,000+ NAICS codes — eliminating the data assembly step for analysts who need to move fast.
Common Pitfalls in Market Size Estimation #
Most market size errors are not mathematical — they are structural. The following mistakes recur across pitch decks, investment memos, and strategy documents regardless of the analyst's experience level.
- Citing market research reports without auditing the methodology: Many published market size figures are modeled estimates, not counts. When a report says "the global X market will reach $Y billion by 2030," the underlying methodology is often opaque. Always ask: what is the base data source? If it is not a named federal or statistical agency, treat the number as directional at best.
- Confusing TAM with SAM: Pitching a $500B TAM when your product only serves a $2B SAM destroys credibility faster than any competitive weakness. Define the boundary of each tier explicitly before presenting any number.
- Using national figures for a regional play: A product launching in three states cannot claim the national TAM. Census CBP data is available at the county and metro level — use it.
- Ignoring firm-size distribution: A market with 50,000 establishments sounds large until Census SUSB reveals that 80% are sole proprietors with no budget authority. Always filter by relevant employee band before applying a per-unit value.
- Static sizing in dynamic markets: A market size calculated on 2019 data may be structurally wrong in 2024. BLS QCEW updates quarterly. Use the most recent available vintage and document the date.
Avoiding these pitfalls requires both methodological discipline and clean data. VantaInsights reports are built entirely on sourced and cited federal data — Census CBP, SUSB, BLS QCEW, OES, and FRED — with pre-calculated metrics for over 1,000 NAICS-classified US industries. For analysts who need a verified starting point rather than a blank spreadsheet, reports start at $239 for a single-industry profile.