Methodology · 4 min read

How Demografics Calculates Trade-Area Demographics (and Where the Numbers Have Limits)

Published October 10, 2026 by the Demografics team

Every number in a Demografics report comes from a public data source and a documented calculation. This article explains both, including the places where the numbers are approximations, so you can judge how much weight to give them.

Data sources

Demographics: US Census Bureau American Community Survey (ACS) 5-year estimates for 2018–2022, at the census tract level.
Trade-area boundaries: 2020 census tract boundaries from the Census Bureau's TIGERweb service.
Five-year growth: county-level ACS 2013–2017 compared with 2018–2022.
Nearby competitors: Google Places listings for the selected store category.
Geocoding: Google's geocoding service, with OpenStreetMap as a fallback.

From a radius to census tracts

Census data is published for fixed geographic units, not for circles. To estimate a 3-mile trade area, Demografics finds every 2020 census tract whose internal point (a representative point inside the tract chosen by the Census Bureau) lies within 3 miles of the address, then combines the data for those tracts. Tracts are designed to hold 1,200 to 8,000 people (about 4,000 on average), so the edge of the circle is approximate: a tract that is mostly inside the circle but whose internal point is just outside will be left out, and vice versa. For very small radii, if no tract's internal point falls inside the circle, the single nearest tract is used.

Combining tracts

Counts (population, households, homes by tenure, people by age, households by income bracket, workers by industry) are added together.
Medians cannot be added. Median household income is estimated as the average of the tract medians weighted by each tract's number of households; median age uses population weights. This is a common approximation; the true median of the combined area can differ.
Percentages are calculated from the combined counts.

A worked example: 3 miles around Houston City Hall

Using the method above, the 3-mile circle around Houston City Hall contains 62 census tracts with internal points inside it. Together they have 189,288 residents. The household-weighted median household income is $102,832, the population-weighted median age is 35.6, and 61% of occupied homes are rented. Density is the population divided by the circle's area (3.1416 × 3² ≈ 28.3 square miles), or 6,695 residents per square mile.

Applying the prepaid wireless fit rules to those figures gives 80 (A). Each step can be checked against the published Census tables and the rules listed in this article.

Survey estimates and margins of error

The ACS is a sample survey. Each tract figure has a margin of error, and for small tracts or small groups (for example, a single income bracket in one tract) the margin can be large relative to the estimate. Summing many tracts reduces the relative error, which is one reason trade-area totals are more reliable than single-tract figures. As a rule of thumb, treat differences of a few percentage points between two places as noise.

The 5-year estimates describe the average over 2018–2022, not a single year. Areas that have changed quickly since then, such as neighborhoods with large new apartment buildings, will be under-represented.

Residents, not visitors

Census data counts where people live. It does not measure office workers who commute in, students who are counted at a family address, tourists or shoppers passing through. Downtown business districts and malls can be busy while having few residents. Demografics does not measure foot traffic.

Competitors

Competitor counts come from a Google Places search for the store category within 1 mile of the site. Google returns at most 20 places per search, so 20 means "20 or more". Listings can be out of date, and category matching is imperfect: for the prepaid wireless category, results are filtered to carrier and wireless store names and exclude repair shops and kiosks.

How the scores work

General viability score: starts at 60 and moves up or down with median household income and population density, within a range of 10 to 98.
Industry fit score: category-specific rules that add points when density, income, renter share and age match the typical customer profile of that kind of business (published for prepaid wireless in our article on 25 downtowns).
Opportunity index: the industry fit score minus 5 points per competitor within 1 mile, up to a 40-point deduction.

The scores are transparent rules of thumb to help compare locations consistently. They are not sales forecasts and do not account for rent, visibility, access, parking or local knowledge. Always visit a site before deciding.

AI-written summaries

Where a report shows an AI-written summary, the AI (Google's Gemini) is given only the figures from the report and instructed not to add numbers of its own. The underlying figures and scores are always calculated directly from the data described above.

Sources

U.S. Census Bureau, American Community Survey 5-year estimates, 2018–2022 (tables B01001, B01002, B01003, B11001, B19001, B19013, B25003, C24050).
U.S. Census Bureau, TIGERweb 2020 census tract boundaries (used to find the tracts inside each trade radius).
U.S. Census Bureau, American Community Survey: Accuracy of the Data and margins of error documentation.
Google Places API (Nearby Search) for competitor listings.

Want these figures for a specific address? Run a free trade-area report on Demografics for any US location, at 1, 3, 5 or 10 miles.