Data enrichment

Data enrichment is an ArcGIS service function that enhances your location data, such as an address, with information about the people and places nearby. It is performed by the GeoEnrichment Service in Esri products like ArcGIS Online, ArcGIS Enterprise, ArcGIS Pro, ArcGIS Business Analyst, ArcGIS Platform, and more.

The information provided from data enrichment can help you to perform location-based analysis and make decisions. For example, when you select an area on a map to enrich, the GeoEnrichment Service can retrieve information about that region, such as local spending habits, household wealth, or the average number of children per household. This might help you to produce projected results like sales territories, school districts, or voting precincts.

You can use data enrichment to:

  • Choose from over 15,000 demographic analysis variables in more than 170 countries.
  • Add location-based context about the people and places affected by your data.
  • Learn about consumer spending habits, identify markets, and explore demographic traits to understand communities.
  • Use location analytics to formulate business and policy decisions.

Typically, you perform data enrichment in an Esri product. Some products require that you sign in with an ArcGIS account that has GeoEnrichment privileges to perform data enrichment; this may consume credits or generate pay-as-you-go costs.

How data enrichment works

Data enrichment uses your input location data and selected analysis variables (also called attributes) and returns the requested information for that region.

When data enrichment is used in an ArcGIS mapping application, the result creates a map layer in the app. The GeoEnrichment Service uses the best available apportionment method to determine the value of each variable on the map layer.

To learn more about summarizing data and to visualize the data apportionment process, visit Data apportionment.

Settlement points

To apportion data correctly, the GeoEnrichment Service relies on settlement points, which are a weighted estimate of population density (population per square mile/km).

There are several types of settlement points. Not every country or region produces settlement points in the same way, and depending on the country or region, a settlement point is one of the following:

Settlement point typeDescription

Residential Address Settlement Points

  • Based on residential address points, these are building footprint centroids (centroid = geometric center), parcel centroids, or address points that represent residential housing units (HU). Many countries/regions track and make available the points that represent actual residential addresses.
  • Usually produced within the past two years.
  • The most accurate form of settlement. Esri works to acquire the most up-to-date, accurate HU points whenever possible.

Government-Sourced Grid Settlement Points

  • Based on government-provided small-grid population counts. Often derived from residential housing units with population counts.
  • Usually smaller than 100 sq. meters.
  • Typically produced within the past five years.
  • To simplify the data and increase performance, Esri may create these points by aggregating the count of HU points to an approximate 100-meter resolution grid.

Census Block Centroid Settlement Points

  • Produced as centroids from the most detailed tabulation areas in their source country or region, such as Australian census mesh blocks. Australia and the United States are examples of this settlement point type.
  • Usually equivalent to a city block.
  • In some areas, Esri has moved these points to be located within residential areas, rather than obviously industrial or other non-residential areas.
  • Roads, HU data, and land use data can help adjust centroid location to large rural block polygons.
  • Each point contains attributes for the count of people and households that live in the corresponding tabulation area.
  • While very accurate in urban areas, block centroids cannot represent multiple locations in rural areas because there is only one centroid per census unit.

Residential Generic Building or Parcel Footprints (GF) Settlement Points

  • GF are building footprint centroids, parcel centroids, or address points that represent residential housing units (HU).
  • Many countries and regions track and make available points that represent actual building and property addresses. Other third-party data providers offer building footprints for various countries or regions. Esri and others in the GIS community continue to develop automated Building Footprint Extraction tools using imagery; this work is ongoing.

All Building Footprints Settlement Points

  • These are building footprint centroids, parcel centroids, or address points that represent non-residential housing units.
  • Many countries and regions track and make available points that represent actual building and property addresses. Other third-party data providers offer building footprints for various countries or regions. Esri and others in the GIS community continue to develop automated Building Footprint Extraction tools using imagery; this work is ongoing.

Esri Raster Based Settlement Points

  • Produced by Esri based on a settlement likelihood model that uses Landsat8 satellite imagery and road intersections. Road intersections are particularly useful in areas where the dense forest canopy obscures dwellings.
  • Initially produced as a dasymetric raster surface, which means that places where people cannot live or where people do not live have been removed.
  • The raster surface is produced at a resolution of 75-meters, which is roughly the size of a city block.
  • This model assigns each cell or point a settlement likelihood score, representing the likelihood of people living there.
  • Created using a model. Accuracy varies between countries and regions due to variations in input data. Includes a reliability score with data GeoEnrichment for countries or regions that use these points.
  • Learn more about Reliability.

WorldPop.org Raster Based Settlement Points

  • WorldPop.org uses a Machine Learning (ML) algorithm to predict a settlement likelihood. See www.worldpop.org for more details.
  • Created using a model. Accuracy varies between countries and regions due to variations in input data. Includes a reliability score with data GeoEnrichment for countries or regions that use these points.
  • Learn more about Reliability.

Each country or region for which Esri provides demographic data uses settlement points to apportion analysis variables. The Data source section of each country or region page provides settlement points information, when available.

Example of settlement points from the Armenia Data source content.
Example of settlement points from the Armenian Data source content.

Data sources and reliability

Data enrichment is available in over 170 countries and regions. Data sources, enrichment variables, and data quality vary by country and region.

Some countries/regions, such as Australia, Canada, Japan, the Netherlands, and the United States, include variables that describe, in detail, the characteristics and consumer behaviors of their citizens. Other countries and regions have fewer variables based on coarser census or postal code geographies. Consequently, the reliability of data enrichment can vary by country or region and sometimes by location within a country or region.

When raster-based settlement points are used in an enrichment operation, the metadata returned includes a reliability statistic for each country or region. The ratings for reliability range from 1.0 to 5.0, where 1.0 is the most reliable. All other countries/regions that use different settlement point methods have scores of -1.

Learn more about raster-based settlement point apportionment reliability.