Understand the market before you size the opportunity

A transparent market-analysis process

1. Define the marketSpecify the customer, product category, geography, channel, price point, and time horizon. “The U.S. market” is rarely a useful unit by itself.
2. Build demand from observable unitsStart with relevant people, households, businesses, workers, or transactions. Document each filter and avoid presenting the entire population as addressable demand.
3. Measure supply consistentlyChoose industry codes and business definitions deliberately. Establishments, firms, locations, employees, and online sellers are not interchangeable.
4. Compare levels and changeUse absolute size, per-capita or per-business density, multi-period growth, and appropriate benchmarks. One large annual change can be revision or rebound rather than a trend.
5. Turn the market into scenariosCalculate low, base, and high cases for reachable customers, conversion, purchase frequency, and revenue, rather than publishing one false-precision estimate.
6. Identify disconfirming evidenceLook for shrinking customer segments, falling real income, rising supplier counts, weak employment, or category-specific saturation.

Worked-example framing: a regional service launch

Imagine a service business choosing among three metros. The first pass would define its actual buyer - for example, households within a stated income band or businesses within specified industries and size classes. ACS estimates or Census business data can count those units, while QCEW can show whether the relevant local industries are expanding or contracting.

Next, the analysis would approximate the provider base using a consistent industry definition. The comparison should show both the number of providers and a normalized measure, such as providers per 10,000 relevant households. A metro with many competitors may still be less saturated if its reachable demand base is much larger.

Finally, a scenario model would make commercial assumptions explicit. The output might show how many customers and how much purchase frequency are required at the proposed price to reach a target, then compare those requirements with the measured addressable base. Public data can constrain the story; it does not reveal a company’s future conversion rate.

Representative primary sources

• People and householdsCensus ACS data tables for demographic, income, housing, and workforce estimates.
• Local industriesBLS QCEW for industry establishments, jobs, and wages by area.
• Business structureCensus County Business Patterns for subnational business activity by NAICS category.
• Economic benchmarkBEA GDP by county for broad regional economic output and industry context.
• Consumer spendingBLS Consumer Expenditure Surveys for expenditure patterns at supported geographies and classifications.
• Price contextBEA Regional Price Parities for regional price-level comparisons.
• Food-market contextUSDA ERS Food Environment Atlas for county food access and store-availability measures when evaluating grocery or food-service markets. Variables have different vintages, and access measures do not prove profitable demand.
• Health-market contextCDC PLACES, HRSA AHRF, and the CMS Provider Data Catalog for aggregate prevalence, workforce, and provider signals. Modeled and area-level measures are not individual diagnoses, patient-flow records, or complete quality measures.
• Education and talent contextNCES IPEDS and College Scorecard for institutions, completions, fields, and selected outcomes. Completions are not current applicants or guaranteed local workers, and outcome cohorts can lag.
• Financial-market contextFDIC BankFind Suite, CFPB complaints, and HMDA for branch presence, reported complaints, and mortgage applications. Snapshots, allegations, and unadjusted lending outcomes do not establish market health, misconduct, or discrimination.

Limits that belong in the answer

Industry codes may not match a new category cleanly. Public establishment data can be suppressed, lagged, or unable to identify direct competitors. Consumer expenditure data may not support the requested local geography. Market growth does not imply an individual entrant will gain share, and a correlation between local indicators and category activity does not establish causation.