How Map AI measures a location
Every number on Map AI starts at street level, in the smallest Census area its publisher releases: a block group or a tract. Our geographic weighting algorithm then fits that data to the exact area you ask about, whether it is a radius around an address, a 10-minute drive or a one-mile hexagon.
This page explains the method in plain terms. Which metric comes from which source, and how fresh each one is, is listed on our Data sources page linked below.
240,000+
Census block groups behind the numbers
1-mile
hexagon cells covering the whole US
Any area
radius or drive time around any address
Built from the finest data available
Each metric comes from the most detailed geography its publisher releases. A Census block group usually holds 600 to 3,000 people, so the numbers reflect the streets inside your area, not the county around it.
Census block groups
Age and sex, race, household income, employment, home ownership, home value, rent, household size, vehicles and commute mode.
Census tracts
Education, personal income, health insurance, travel time to work, CDC health measures and life expectancy - wherever the publisher does not go below tract level.
Roads and places
Traffic volumes per road segment, road length, points of interest and schools are counted exactly where they are, with no averaging.
A few datasets, such as crime and health workforce, are published only per county and are shown at that level.
How we fit data to your area
Census borders rarely match the area you care about. A 3-mile radius or a 10-minute drive cuts through dozens of block groups. Our geographic weighting algorithm turns them into one accurate profile in four steps.
1. Count what is inside
Every block group or tract that touches your area contributes the share of its people and households that falls inside. Fully inside counts in full, half inside counts half.
2. Weight by people, not land
Ratios and averages are computed from the people and households actually counted, so a block group with 3,000 residents weighs thirty times more than one with 100.
3. Rebuild full distributions
Income, age, rent and home value move bracket by bracket, and medians are calculated from your area's own distribution - never by averaging the medians of other areas.
4. Verify against the source
Automated checks reject impossible values while the data is built, and rebuilt totals are compared with the official figures each source publishes. Mismatches are traced to their cause and corrected.
In these checks, county and state totals rebuilt from block groups match official Census figures with a median difference of 0.04% to 0.24%.
One method behind every tool
The map, the explore pages and every analysis tool use the same weighting. A radius around an address is directly comparable with a city, a county or a hexagon cell.
Demographics by radius
Pick an address and a radius to get population, income, age, housing and commute for exactly that circle, built from the block groups and tracts inside it.
Try demographics by radiusDrive-time trade areas
Get the same granular profile for the area reachable in a set number of minutes by car, along the real road network.
Try demographics by drive time1-mile hexagon map
Browse the whole US as equal one-mile cells, so neighboring areas compare like for like - unlike ZIP codes or counties, which vary widely in size.
Open the interactive mapSources and freshness
Map AI works with official and open data - the U.S. Census Bureau American Community Survey and TIGER/Line boundaries, CDC, FBI, NCES, OpenStreetMap and Mapbox - plus traffic volumes from TrafficZoom, and maps it onto the area you choose.
Exact sources for every metric, and how current each one is, are listed on our Data sources page.
Try demographics by radius
