Listed below are common questions about Esri Location Data. As noted, these questions may apply to Esri Demographics and ArcGIS Places.
General
- How is data calculated for user-created areas?
- What are the sources for Esri Location Data?
- For which countries does Esri offer Esri Location Data?
- Why is Türkiye included in the European region?
- Why is Greenland included in the North American region?
- How can I find out what data variables are offered by Esri?
- Can I download Esri Location Data?
- Which Esri products include Esri Location Data?
- How frequently is Esri Location Data updated?
ArcGIS Places
- What are places and points of interest (POI) data and how are they used?
- What types of places and POI data are supported by Esri?
- Where can I access ArcGIS Places data?
- How often does Esri update ArcGIS Places data?
- Where do I find information about licensing restrictions related to ArcGIS Places?
- Why am I limited to adding 5,000 locations to a map?
- What ArcGIS Places data is available for my area and what's the difference between the data sources?
- Why does my sales volume value seem different than the expected value?
- Why don't some businesses have employee counts or sales volumes in Data Axle?
- Why is Esri moving away from SIC codes?
- I found a mistake in the ArcGIS Places data—for example, a location's coordinates are inaccurate. How do I report it?
- Why is a location missing or a closed location listed in an ArcGIS Places dataset?
- Why is contact information missing from ArcGIS Places datasets?
Esri Demographics
- What are standard demographics and advanced demographics?
- On what time period or date are Esri Demographics based?
- What demographic topics are covered by Esri Demographics?
- What is the data source for standard demographics?
- What is Population per Mill?
- Is spending data in Canada reported as Canadian dollars or U.S. dollars?
- What is the difference between income (constant) and income (current) in Canada?
- What currency does income and spending data represent?
U.S. data
- Why are Esri projections for five years?
- How accurate is the Esri Updated Demographics data for the United States, and what distinguishes it from other vendors?
- Why do Esri’s U.S. demographic update figures sometimes differ from recent census numbers?
- How is Esri's Diversity Index different from the Census Bureau's Diversity Index?
- Does Daytime Population represent population counts for the workweek only, or does it also include the weekend population?
- How are the percentages of White Collar, Blue Collar, and Services employment categories defined and calculated?
- What type of occupations are included in the Population by Industry and Occupation table?
- What is ACS data and why is it labeled with a five-year range?
- How are college students and military barracks accounted for in households?
- Do Household Income, Income by Age, Disposable Income, Net Worth, or Per Capita Income include the income of people living in group quarters?
- Is Hispanic Population included in the Total Population estimates and forecasts?
- Are centroids located at the geographic center or are they weighted?
- How can I learn more about certain Consumer Spending items?
- What do the three different Consumer Spending measures with each item mean?
- What do the Market Potential values represent—people or households?
- Why does Esri’s U.S. unemployment rate differ from other sources of unemployment?
- Why does Esri Updated Demographics not include margins of error (MOE) like those included in ACS data?
- Why are the total number of ZIP codes available from USPS different from the total number of ZIP codes in Esri Updated Demographics?
U.S. data—ArcGIS Tapestry
- What is the update cadence for Tapestry?
- Why do some Tapestry segments remain the same across generations, while others are newly created?
- Why are there 60 residential segments in the ArcGIS Tapestry system as opposed to 67 in the Esri Tapestry Segmentation system?
- Which variables are used to build Tapestry? Is race and ethnicity included?
- What happened to the Urbanicity groupings present in the Esri Tapestry Segmentation system?
- Why is there only one segment at the block group geographic level?
- Is there a direct comparison between ArcGIS Tapestry and Esri Tapestry Segmentation?
- Why has the Tapestry segment coding structure changed? Are codes the same for the segments that have persisted?
- Do all people and households in a segment share the same characteristics?
- How are college students and military barracks accounted for in the ArcGIS Tapestry database?
U.S. data—Esri Tapestry Segmentation
General
Data summarization provides estimates in areas that don't match up exactly to common geographic areas, such as census block groups in the United States. User-created areas of interest—such as rings, drive times, hand-drawn polygons, or imported shapefiles—frequently cut through typical geographic boundaries. To account for population that may not be evenly distributed geographically in user-created regions, the application identifies where the population is distributed using the smallest geographic level (such as a block level in the United States), and uses this information to summarize the data more accurately. Read more about the data allocation method in Data enrichment.
Esri Demographics and ArcGIS Places data are supplied by a variety of data source providers:
- The Esri standard demographics datasets are sourced from data supplied by Michael Bauer Research GmbH.
- The Esri advanced demographics datasets are based on data supplied by a variety of demographic data content providers. See the data source providers topic for a list of data source providers for advanced demographics or the Esri Location Data vintages topic for a list of the update frequency.
- The ArcGIS Places datasets are based on place and point of interest data supplied by a variety of location data providers. See the ArcGIS Places data source providers topic for a list of the location data sources by region.
Esri offers Esri Demographics and ArcGIS Places data for the more than 170 countries listed below:
- Africa—Algeria, Angola, Benin, Botswana, Burkina Faso, Cameroon, Cabo Verde, Central African Republic, Chad, Côte d'Ivoire, Egypt, Equatorial Guinea, Ethiopia, Eswatini, Gabon, The Gambia, Ghana, Guinea, Guinea-Bissau, Kenya, Lesotho, Liberia, Madagascar, Malawi, Mali, Mauritania, Mauritius, Mayotte, Morocco, Mozambique, Namibia, Niger, Nigeria, Réunion, Republic of the Congo, Senegal, Sierra Leone, South Africa, Sudan, Tanzania, Togo, Tunisia, Uganda, Zambia, Zimbabwe
- Asia—Armenia, Azerbaijan, Bahrain, Bangladesh, Brunei Darussalam, Cambodia, China, Georgia, Hong Kong, India, Indonesia, Iran, Israel, Japan, Jordan, Kazakhstan, Kuwait, Kyrgyzstan, Lebanon, Macao, Malaysia, Maldives, Mongolia, Myanmar, Nepal, Oman, Pakistan, Philippines, Qatar, Saudi Arabia, Singapore, South Korea, Sri Lanka, Taiwan, Tajikistan, Thailand, United Arab Emirates, Uzbekistan, Vietnam
- Europe—Albania, Andorra, Austria, Belarus, Belgium, Bosnia and Herzegovina, Bulgaria, Croatia, Cyprus (Greek part), Czech Republic, Denmark, Estonia, Faroe Islands, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Kosovo, Latvia, Liechtenstein, Lithuania, Luxembourg, Malta, Moldova (without Transnistria), Monaco, Montenegro, Netherlands, North Macedonia, Norway, Poland, Portugal, Romania, Russia, Serbia, Slovakia, Slovenia, Spain, Sweden, Switzerland, Türkiye, Ukraine, United Kingdom
- North America—Anguilla, Aruba, Bahamas, Barbados, Belize, Bermuda, Canada, Cayman Islands, Costa Rica, Dominican Republic, El Salvador, Greenland, Guadeloupe, Guatemala, Haiti, Honduras, Jamaica, Martinique, Mexico, Montserrat, Nicaragua, Panama, Puerto Rico, Saint Lucia, Trinidad and Tobago, Turks and Caicos Islands, United States, U.S. Virgin Islands
- Oceania—Australia, Fiji, French Polynesia, New Caledonia, New Zealand, Papua New Guinea
- South America—Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador, French Guiana, Guyana, Paraguay, Peru, Uruguay, Venezuela
The levels of detail and number of attributes or locations vary by region.
Türkiye is geographically divided into two parts: a European part called the Thracian region and an Asian part called Anatolia. The Thracian region contains the capital city, Istanbul, the country's most populous urban area, while Anatolia is larger geographically. For purposes of Esri Demographics, the decision to categorize Türkiye within Europe is based on both internationally recognized classifications established by organizations such as the United Nations and regional demographic trends.
Although Greenland is a Danish territory, it is located within the North American continent. Esri Demographics includes Greenland in the North American region because this aligns with common geographic and demographic practices. This categorization helps to maintain geographic consistency and provide a comprehensive overview of the region's demographic trends.
Each Esri Demographics country or region topic has a Related links section containing a variable list itemizing the available dataset for that country. Countries or regions with advanced demographics or ArcGIS Places may have more than one variable list. You can also explore the data categories and individual variables available for each country using the data browser.
For information about purchasing Esri datasets as stand-alone content, contact datasales@esri.com or Buy Esri Data directly from the store. For information about purchasing reports, see Esri Reports.
See Where to find Esri Location Data and Introduction to ArcGIS Places for information about using Esri Demographics and ArcGIS Places in various ArcGIS products.
Esri Location Data is updated at various times:
- Standard demographics datasets sourced from data supplied by Michael Bauer Research GmbH are updated approximately once a year for Europe and once every two years for Africa, Asia, North America, Oceania, and South America.
- Advanced demographics datasets are updated approximately every two years, but this varies by data source.
- ArcGIS Places datasets are updated at various intervals. See Introduction to ArcGIS Places for more details.
ArcGIS Places
Places, also known as points of interest or POIs, are businesses and geographic locations that you can discover around the world. The level of detail available for a place can vary by place type, region, and the data source that supplied the place details.
Places or points of interest are typically destinations, such as shopping centers, hotels, banks and ATMs, gas stations, beaches, parks, and more. The kind of data you can get about the place typically includes the name, location address, business category information, and related public details. Different data sources provide different levels of detail.
Not all places have comparable levels of detail available. For example, a place may not include a postal code if it is in a country that does not use postal codes. An address may be missing if the source of the information didn't provide the address. In other cases, a feature may not be associated with a street address—examples include natural features, such as a pond or a forest; built features, such as bridges or statues; or nonpermanent places, such as weekend street markets.
See Introduction to ArcGIS Places and Where to find Esri Location Data to learn about the Esri products where you can find ArcGIS Places data.
Esri updates ArcGIS Places datasets as frequently as three times per year and as infrequently as every few years, depending on the data source and dataset. See Data source vintages and update frequency for a list of all datasets, locations, and data sources.
Please refer to your organization's signed agreement with Esri or see the Esri E204 Legal Master Agreement Products and Services document. Section 3 specifically addresses data and associated use restrictions.
Esri has a contractual obligation with certain ArcGIS Places data source providers that caps the number of locations a user can export or download to 5,000 points per request. This helps protect our vendors' intellectual property. Further, user requests that exceed 5,000 locations or points of interest could potentially impact system performance. Therefore, this cap has been generally implemented to meet both contractual obligations and maintain optimal system speeds.
- If you are interested in accessing larger numbers of locations or points of interest in an ArcGIS project, Esri offers ArcGIS Business Analyst Bundles for ArcGIS Pro and ArcGIS Enterprise. This includes ArcGIS Demographic and Business data that does not have a cap on the number of records that can be used. This dataset is only available for Canadian or United States customers. Speak with your Esri account team for more information about ArcGIS Business Analyst Bundles.
- If you are a developer with an Esri Developer Bundle subscription, you can use the ArcGIS Places service on the ArcGIS Location Platform. Users of the ArcGIS Places service have access to location information from more than 170 countries, with no limit to how many business locations can be visualized. Speak with your Esri account team rep to discuss the Esri Developer Bundle options.
What ArcGIS Places data is available for my area and what's the difference between the data sources?
Different data sources typically specialize in data content for a country or smaller region. The data sources topic lists each data source provider and the countries they specialize in; the Attribute comparison table in the Introduction to ArcGIS Places provides a general overview of the types of attributes the data source focuses on. One data source, Foursquare, provides open source, global ArcGIS Places content for ± 170 countries. In the U.S. and Canada, there are three data source providers—Data Axle, SafeGraph, and Foursquare. Data Axle specializes in business information—such as employees and sales, while the other providers feature locations that include both business and non-business POI (such as hiking trails). Refer to the Attribute comparison table to decide which data source is better for your use case.
The sales volume attribute is a value generated based on a proprietary sales model developed by Data Axle. Esri does not have direct access to this propriety methodology for generating sales volume estimates. The value can fluctuate over Esri's three annual location data release cycles. If there are any concerns, Esri will reach out to the data supplier for a better understanding of changes.
There are a few reasons. Some businesses may not have an employee count because the business may have a single owner/operator and no employees. Some businesses—such as school district offices, non-profit organizations, and government agencies—do not have sales volumes. Kiosks (such as ATMs) do not have sales volumes or employee counts. Broadly speaking, sometimes there is just not enough information about a business to create an estimate for these attributes.
I found a mistake in the ArcGIS Places data—for example, a location's coordinates are inaccurate. How do I report it?
Esri works to ensure the quality of the data, but mistakes can and do happen. Please use the feedback form in the Esri app, map, or help topic where you found the problem to report errors. The feedback form link is usually at the lower right. If you can't find a feedback form link, use the Contact us page. Please provide as much specific information about the application, dataset, and function where you observed the error to help solve the problem.
There is no substitute for local knowledge when it comes to locations opening and closing. Esri data sources work at a national or even global scale, which can be slower than a local scale. New businesses may be identified by social media signals, registration with local governments, utility activations, and so on, but getting this into a global dataset takes time. Closures can be even more difficult to identify because a business may close a location but not deactivate their business license. Esri is happy to share any opening/closure information provided by customers with data vendors. You can always report incorrect location information to help improve the content.
Tip:
We often hear "But it's in Google." Our vendors are not allowed to copy Google's proprietary data.
Esri Demographics
Standard demographics is a basic set of comparable variables for each country, sourced from data supplied by Michael Bauer Research GmbH (MB-Research). MB-Research is a globally recognized developer of demographic data and boundaries at small scales for more than 170 countries or regions and has been in operation since 2004.
Advanced demographics are available for select countries. These provide a more detailed set of demographic variables and, in some cases, a more detailed level of geography. Data sources vary by country or region.
Vintages vary by country or region. See the Esri Location Data vintages topic for a list of the vintages for each country or region.
Current date information is provided for Canada and the United States:
- Canada
- Demographic Statistics are current as of March 31 of the applicable year.
- Household Spending are current as of June 30 of the applicable year.
- CensusPlus are current as of March 31 of the applicable year.
- United States
- Updated Demographics are current as of July 1 of the applicable year. For example, 2024 U.S. estimates are current as of July 1, 2024.
- 2020 Census are current as of April 1 of the census year.
- U.S. Census American Community Survey (ACS) data uses a continuous measurement, or rolling sample, in which a small percent of the population is sampled every month. The ACS data is updated and released more frequently than the decennial census—every year instead of every 10 years. For example, a five-year range for Esri’s ACS data is 2019-2023. See What is ACS data and why is it labeled with a five-year range? in the U.S. FAQ.
The attributes and variables available for each country vary; however, the following data categories are generally available for the ±170 countries or regions for which Esri provides standard demographics: Population counts, Population by Gender counts, Population by 15-year Age bands, Household counts, Households by Income Quintiles, Households by Type, Educational Attainment, Marital Status, Unemployed Population, Purchasing Power, and Consumer Spending.
Advanced demographics datasets include a broader range of data categories in addition to those included with standard demographics. These include more granular details within the data categories, such as Population by Religious Affiliation or by Generational Cohort. There are also a number of data categories that are available only with advanced demographics, such as Housing, Health, Business, Race or Ethnicity, Retail and Centrality, Behaviors and opinions, Psychographics, Poverty-metrics, Language-Immigration, Segmentation, At Risk, Retail Demand, Supply and Demand, data that impacts government Policy, and Crime data.
Standard demographics are sourced from data supplied by Michael Bauer Research GmbH, a globally recognized developer of demographic data and boundaries at small scales for more than 170 countries that has been in operation since 2004.
The precise answer is that it is a count of the total regional population divided by the total national population and then multiplied by 1,000. A less formulaic answer is that population per mill is a common statistic used to express proportions—or ratios—of population per thousand. Per mill values show how a nation's population is distributed: densely populated areas have larger per mill values, sparsely populated areas have smaller per mill values.
All dollar amounts in the Household Spending dataset from Environics Analytics are in Canadian dollars.
Income (constant) reports values in a constant-year price. Current income reports values in the current-year price. For example, Median Household Income (Constant Year 2024 $) reports the median in 2024 dollars, whereas Median Household Income (Current Year $) reports the median in current-year dollars.
Income and spending data is displayed in the local currency. This information is available for each country and is shown in the Release notes.
U.S. data
Esri uses a five-year projection interval to ensure that projected data remains applicable to multiple use cases. The further into the future a projection goes, the more uncertain it becomes. Esri's projection methodology has three main components. First, past and present trends are used to predict future trends. Second, a cohort component approach is used to model expected population change based on demographic characteristics. Third, pipeline data for future residential construction, as well as building plans for developments currently under construction, give Esri a unique view of the future landscape. Variability of all three input components increases as the projection horizon lengthens. For these reasons, a five-year projection horizon reduces uncertainty while maximizing the data's usability.
How accurate is the Esri Updated Demographics data for the United States, and what distinguishes it from other vendors?
Esri has a team of demographers, statisticians, and economists who use a variety of public and private data sources to develop Esri Updated Demographics. It is a detailed representation of local population, economic, housing, and business characteristics in the United States. Using GIS tools and demographic estimating experience, Esri captures the effects of significant events that occur between decennial censuses.
Esri uses the decennial census to evaluate the accuracy of their annual demographic estimates compared to census results. In 2011, Esri commissioned an independent study of their data accuracy compared to other major data vendors and the United States 2010 Census at four geography levels. The panel included credentialed researchers and experts in small-area forecasts and measures of forecast accuracy. The resulting 2010 Vendor Accuracy Study provides some insight into Esri's ongoing commitment to data accuracy.
For the 2020 Census, Esri refined estimate and forecast accuracy between decades, particularly at lower levels of geography. Esri's 2020 estimates were also assessed using comparisons to five-year estimates from the 2016-2020 American Community Survey (ACS) data, resulting in closer data representations of ground truth. The Census 2020 Accuracy Analysis provides an assessment of how accurately Esri's data compares to the latest decennial census.
Esri provides demographic updates annually, representing July 1 of the current year.
Census numbers can refer to a variety of counts, estimates, and projections from the U.S. Census Bureau.
- Counts—Census counts refer to the data reported from a decennial census of the U.S. population. The most recent census count was in 2020.
- Estimates—In addition to the census count, the Census Bureau provides both independent estimates based on demographic analysis and survey-based estimates. The timing for each estimate type the U.S. Census provides varies.
- The latest independent county estimates from the Census Bureau usually run a year behind Esri’s current-year estimates, and these estimates are released after Esri’s updates are completed.
- The annual time series of estimates from the Census Bureau are also subject to revision annually. Survey-based estimates, such as the American Community Survey (ACS), predate Esri’s current-year updates.
In addition, the ACS data for all areas is a five-year average, while Esri’s updates are point data estimates, representing July 1 of the reference year.
Race and ethnicity are separate concepts. For this reason, race and ethnicity questions are asked separately by the decennial census. Everyone counted by the decennial census falls into one of seven race groups (White, Black, American Indian, Asian, Pacific Islander, Some Other Race, or Multiple Race) and one of two ethnicities (Hispanic or Non-Hispanic). By treating race and ethnicity as two independent variables, Esri's Diversity Index compounds racial diversity with ethnic diversity to provide a more inclusive understanding of both race and ethnicity. The Census Bureau's Diversity Index does not account for racial diversity within the Hispanic population. For these reasons the Esri Diversity Index is often slightly higher than the Census Bureau Diversity Index. However, the two measures are highly correlated (r=0.95 at the county level).
Does Daytime Population represent population counts for the workweek only, or does it also include the weekend population?
Esri’s Daytime Population is an estimate of the population that covers both residents and workers during the day. Esri’s data source for calculating daytime population does not distinguish if workers are in an area predominantly between the hours of 9:00 a.m. and 5:00 p.m. or during some other time of the day. However, while there is no way to identify specific workday hours, workday activities typically occur Monday through Friday between 9:00 a.m. and 5:00 p.m., unless the data user knows otherwise and takes that into account in their analysis. To learn more about Esri’s Daytime Population methodology, see Esri Updated Demographics.
The occupations described in the Population by Industry and Occupation table are defined by the Standard Occupational Classification (SOC) from the Bureau of Labor Statistics (BLS). See the 2018 SOC System for more information.
ACS, or the U.S. Census Bureau's American Community Survey, is the replacement for sample data from the decennial census. The 2010 Census replaced the long form that previously provided data about income and poverty status, school enrollment, journey to work, household type and relationships, languages spoken, migration, citizenship, disability, health insurance, ancestry, military service, or housing characteristics with rolling surveys from the ACS. ACS data uses a continuous measurement, or rolling sample, in which a small percent of the population is sampled every month. The data is reported as a 60-month average. The ACS dataset is updated and released more frequently than the decennial census—every year instead of every 10 years. To learn more about ACS data, see the American Community Survey topic.
Yes. Total Population is the sum of Hispanic Population and Non-Hispanic Population.
Note:
Total Population is not the sum of all Races and Hispanic Population added together.
The U.S. Census Bureau considers race and ethnicity to be two separate and distinct concepts. Hispanic Origin/Population refers to ethnicity, not race, and is viewed as the heritage, nationality, lineage, or country of birth of the person or the person's parents or ancestors before arriving in the United States. Persons are either Hispanic or Non-Hispanic. Those persons that identify as Hispanic may be of any race.
To learn how Total Population is calculated, see the example pie charts below. The first pie chart represents the percentage of people who are non-Hispanic, while the third pie chart represents the percentage of people who are Hispanic. Each of these when summed separately sums to a total for non-Hispanic population and a total for Hispanic population. The result, the middle pie chart, shows that each of these two totals summed together (Non-Hispanic Pop + Hispanic Pop) equals Total Population by Ethnicity (or Total Population).
Total Population is also defined as the population by race. Race is reported as White, Black or African American, Asian, American Indian and Alaska Native, Native Hawaiian/Other Pacific Islander, Some Other Race, and Multiple Races. To learn how Total Population by Race is calculated, see the example pie chart below. Each pie segment represents the percentage of the population comprising one of the seven races. The population by race segments, when summed together, equal Total Population by Race (or Total Population). 
The following scenarios show how to accurately calculate race and ethnicity percentages:
- Scenario 1—Percentage of Race
compared to Total Population
- Example: (Black/African American Race) / Population by Race Base
- (2,500) / (5,000) = 50%
- Scenario 2—Percentage of Hispanic
Population by Race compared to Total Hispanic Population
- Example: (White Hispanics / Hispanic Population)
- (1,258/6,500) = 19.4%
- Scenario 3—Percentage of White
Hispanics compared to Total Population
- Example: [Hispanic White Population / (Sum of Hispanic Pop + Non-Hispanic Pop)]
-
(3,000/40,000) = 7.5%
Note:
You can also use the Population of Race base to calculate this percentage.
In the data browser, hover over the U.S. Spending information button 
while browsing the available data. The Consumer Spending variables include descriptions about what is included in the value.
Total Spending | The total amount spent by all households in this region annually on the product or service. |
Average Spending | The average amount spent by all households in this region annually on the product or service. |
Index | A comparison of average annual regional spending on the product or service to average national spending. The national index value is 100. An SPI of 115 indicates that local consumers spend 15 percent more per year on this product or service than the national average; an SPI of 85 indicates that local consumers spend 15 percent less than average. |
A letter h in the 10-character variable name indicates that the variable is household based. In the Total Tapestry Households (THHBASE) data, for example, MP01001h_B Household owns/leases any vehicle is a household-based variable.
A letter a in the variable name indicates that the variable is adult based. In the Total Tapestry Adults (TADULTBASE) data, for example, MP01066a_B Member of auto club/roadside assistance program is an adult-based variable.
Esri’s U.S. labor force estimates can differ from unemployment data published by the Current Population Survey (CPS) program of the Bureau of Labor Statistics (BLS) for the following reasons:
- Small area data source—Esri’s labor force updates use American Community Survey data for its forecast base because it is the only source available by block group for this socioeconomic information. While ACS uses the same labor force definitions (for example, how working-aged persons are classified as employed, unemployed, not in labor force, and so on), their rates of unemployed persons will differ (and often run higher) than the CPS—the source often quoted in the news. Ultimately, ACS and CPS differ because the surveys deploy different questions, samples, and data collection methods. More detail can be found in the Comparison of ACS and CPS Data on Employment Status working paper.
- Data vintage—Esri’s labor force data is current as of July 1 of the release year. Data from the BLS (for example, CPS and Local Area Unemployment Statistics (LAUS) programs) is more current monthly time series.
- Seasonal adjustment—Esri’s labor force data is not seasonally adjusted. CPS produces two sets of national unemployment estimates that are seasonally adjusted and not seasonally adjusted. The unemployment rate often quoted in the headlines is typically a seasonally adjusted estimate.
Additional information is available in the Labor Force Data section of Esri’s U.S. Updated Demographics methodology.
Why does Esri Updated Demographics not include margins of error (MOE) like those included in ACS data?
MOEs can only be calculated from a survey of an entire population. MOEs describe how close a survey value can be expected to be to the true value. Esri Updated Demographics data is not a survey and therefore MOEs cannot be calculated. Esri Updated Demographics values are forecasted estimates created by statistical modeling.
Why are the total number of ZIP codes available from USPS different from the total number of ZIP codes in Esri Updated Demographics?
Esri provides residential ZIP codes with land area and excludes point-level ZIP codes. This ensures that residential-based demographic and socioeconomic data are not erroneously appended to a point. Point ZIP codes include PO boxes, and other unique organizations including military and government buildings. Enclosing ZIP codes that surround and include the points are included in Esri's ZIP code inventory.
U.S. data—ArcGIS Tapestry
Tapestry is updated in three distinct phases following the release of decennial census data.
Approximately every 10 years, Esri's geodemographic segmentation system is rebuilt using newly available census inputs alongside other key datasets. While not the sole data source, the census provides variables—such as age by sex—that are only captured once per decade. After integrating all inputs, the Tapestry model undergoes iterative recalibration until a stable segmentation structure emerges. For this reason, users are advised to migrate from Esri Tapestry Segmentation to ArcGIS Tapestry to ensure alignment with current population insights.
A mid-cycle model refresh typically occurs three to five years post-rebuild. This update incorporates recent demographic and economic data to reflect major societal shifts, such as those in housing or the labor market. The segmentation algorithm is rerun without structural redesign, allowing for segment reassignments based on updated input characteristics.
With each annual release, the model undergoes review to detect significant local changes, such as new residential developments, major demolitions, or other transformative land use changes. Segment assignments for affected areas may be adjusted accordingly to keep the model aligned with rapidly evolving neighborhood dynamics.
Tapestry is rebuilt each decade to capture shifts in consumer markets. While many neighborhoods remain demographically stable, others undergo significant change. Segments that retain the same name across generations are persistent segments whose core behaviors and characteristics remain consistent despite minor demographic shifts, such as aging, changes in income, or geographic redistribution.
New segments emerge when neighborhood dynamics change substantially, often related to redevelopment, turnover, or new construction. In these cases, Esri identifies distinct consumer profiles that are not adequately represented by existing segment definitions, requiring the creation of new segment names and descriptions.
Why are there 60 residential segments in the ArcGIS Tapestry system as opposed to 67 in the Esri Tapestry Segmentation system?
The Esri Tapestry Segmentation system included 68 segments (67 classified, 1 unclassified). The ArcGIS Tapestry system includes 61 segments (60 classified, 1 unclassified). Unclassified areas typically include unpopulated land (for example, parks and open space), institutional group quarters (for example, prisons and mental health facilities), or areas lacking sufficient data for classification.
The total number of segments is determined through statistical modeling, including cluster analysis, which groups similar geographic areas based on key variables. The goal is to minimize variation within clusters while maximizing differences between them. Various statistical tools and visualizations guide the selection of the optimal number of clusters.
The reduction in segments reflects evolving demographic and social trends. Increasing diversity across U.S. neighborhoods has blurred distinctions between some groups. The rise of remote work post-pandemic has reshaped residential patterns, decentralizing urban populations and shifting consumer behaviors. Broader cultural and technological integration has also contributed to the emergence of new segments while consolidating others. These dynamics likely influenced the reduction of neighborhood types in the most recent Tapestry update.
For more information about Tapestry inputs and methods, see ArcGIS Tapestry methodology.
Tapestry is built using a selection of variables that distinguish neighborhood types. Variables span three core categories: people (such as age, sex, race, household composition, education, and income), housing (such as housing type, age, value, and vacancy), and behaviors (such as consumer spending, media habits, and psychographics).
Select race and ethnicity data are included, as testing confirms their importance in identifying distinct market segments that reflect the diversity of communities across the U.S.
Data sources include Esri Updated Demographics, the U.S. Census Bureau, the U.S. Census Bureau's American Community Survey, and consumer surveys such as MRI-Simmons. Variable selection is informed by established segmentation methodologies and employs statistical techniques, including factor analysis, principal components analysis, and correlation modeling, to determine the most appropriate representations of the data, such as medians, averages, and ratios.
Esri has introduced a new Urbanicity Type system, replacing the previous six Urbanization groups. This classification defines land using 10 categories based on factors like density, proximity to urban centers, and socioeconomic ties to urban cores. Unlike the previous one-to-one mapping between segments and Urbanization groups, the new system provides a full distribution of urbanicity types within each ArcGIS Tapestry segment, providing additional neighborhood context. For example, the Metro Renters segment is approximately 85 percent Urban Core, 8 percent Suburb, and 7 percent Urban Vicinity.
For more information, see Esri Urbanicity Type.
Census block groups are the smallest geographic unit for which most input data are consistently available. They also offer the most accurate approximation of neighborhoods within the census hierarchy, as they typically contain 240 to 1,200 housing units.
Cluster analysis is performed at the block group level, evaluating multiple variables to assign each area a single segment classification. Smaller geographic units, such as blocks or individual households, are more error-prone and raise privacy concerns, making them less suitable for segmentation.
The ArcGIS Tapestry dataset is different from the Esri Tapestry Segmentation dataset in a variety of ways. Because of these differences, directly replacing Esri Tapestry Segmentation data with ArcGIS Tapestry data is not possible. See Migrate to ArcGIS Tapestry for more information.
Why has the Tapestry segment coding structure changed? Are codes the same for the segments that have persisted?
Both ArcGIS Tapestry and Esri Tapestry Segmentation use alphanumeric codes. However, Esri has intentionally introduced a breaking change in the coding structure. The Esri Tapestry Segmentation system used codes with a number first and a letter second (1A), whereas ArcGIS Tapestry uses codes with a letter first and a number second (A1). Existing workflows that explicitly reference these codes will require modification to migrate to the new syntax.
No. Segment descriptions and statistics reflect average or general traits within a segment type based on group-level data. The definitions and statistical values used to describe a segment are not representative of individual households or people. Households and people within a segment are unique and have characteristics that are divergent from summary-level data and descriptions. For example, while a segment may have a median income or age, actual residents span a range around that median. It is important to avoid making assumptions about individual traits based solely on segment-level data.
U.S. data—Esri Tapestry Segmentation
For example, ZIP code 19553 (Leesport, PA within Berks County) is a postal delivery area located northwest of Philadelphia. The 2022 dominant Tapestry Segment is 6A Green Acres (from LifeMode Group 6: Cozy Country Living). However, the dominant LifeMode Group is 5: GenXurban. How can this be?
The determination of a dominant value, whether for a Tapestry Segment or a LifeMode Group, is computed independent of one another. To illustrate this, consider the following three-segment Tapestry distribution for ZIP code 19553. Total households are the primary input used to determine a dominant Tapestry Segment or LifeMode Group. Hence, the dominant Tapestry Segment is 6A: Green Acres because most households belong to that residential market.
However, that does not imply the dominant Esri Tapestry Segmentation LifeMode Group for this area is 6: Cozy Country Living. Instead, the dominant LifeMode Group is 5: GenXurban, since the total number of households from this group (436 + 1,051 = 1,487) exceeds the total number of households in LifeMode Group 6 (1,198).
