How did you build the Zip+4 dataset?

Step 1: Expanding the USPS Dataset

The USPS address dataset includes 48 million addresses. Some have a single Zip+4, while others use a Zip+4 range. We break these ranges down into individual Zip+4s, adding 21 million new unique Zip+4s—bringing our total dataset to 69 million.

Examples of Zip+4 Ranges

Step 2: Geocoding Each Zip+4

Each Zip+4 address is geocoded using the Census Geocoder, converting it into a precise latitude and longitude. If the geocoder doesn’t return a match, we retry. If there’s still no match, we’re a little sad, but we assign the Zip+4 to the centroid of its 5-digit ZIP as a fallback.

Example Geocodes 

Step 3: Mapping to Census Block Groups

The U.S. Census Bureau provides annually updated income and demographic data at the block group level—the smallest geographic unit available for these estimates. By overlaying our geocoded Zip+4 data onto the latest Census block groups, we accurately assign each Zip+4 to its corresponding block group.

Step 4: Assigning Census Demographics to Zip+4s

Once we have the Zip+4 associated with a Census block group, we associate its demographic estimates—including income and race statistics—based on the Census data for that block group. This step ensures that you receive the most precise, Census-backed demographic estimates for every Zip+4. All Zip+4s within the same Census block group receive the same demographic estimates.

Step 5: Deliverables

Now how do you get this data into your databases? You have 2 options.

1. One large table (5 GB) with all zip+4s and the demographics in the same table

2. Two smaller tables with one table for the Zips+4s (2GB) and one table for the demographics

You’d use the demographics id to look up between the 2 tables.

Samples, pricing and other details are on our Zip+4 page: https://www.incomebyzipcode.com/zip-plus-4