This report reveals the foundational characteristics of your locations’ visitors, including household level demographic profiles such as age, income, race, and educational level.
Table of Contents
👋 Do you have questions about anything on this page or in our help center? Contact our Support team!
Overview
The demographic report is based on the latest available census data from governmental agencies in the supported countries. For example, the US data is based on US Census Bureau’s ACS dataset. The demographics are paired through the census ID associated with a device’s Common Evening Location.
There are 3 files delivered with this report. The summary file gives a holistic picture of the household level demographics of visitors to all locations using a weighted average methodology. The input file gives a breakdown by the input passed to Vista—in most cases this input is a set of polygons, thus the input file shows a breakdown of visitor demographics per polygon. Finally the census_ID file gives a listing of each census ID found to contain a visitor’s CEL and the associated demographics of that census ID.
Methodology
Using historical data from a 3-month window, Azira assigns a Common Evening Location (CEL) to each ID seen within the Vista platform. This is the point/area where the device is most often seen in the evenings (6pm-6am) and/or on the weekends.
To create demographics, Azira identifies the geographic boundaries used by the country’s census data (in US = Census Block Groups, in AUS = Statistical Areas 1) and maps the CELs to those boundaries. A weighted average of demographics for a set of devices is calculated based on the number of devices with a CEL in the census geographic boundaries.
The data in the report are household-level data, not device-level data. Thus, information like gender or age or income of the device holder cannot be calculated. Rather, statistics are given at the household level, reflecting, say, the income of the entire household.
Output and Descriptions
The format of each file is as follows:
By Rollup
| Input | Category | Sub_Category | Value | Devices Count | Sub_Cat Sort Order | Unit |
The By_Rollup file provides a demographic breakdown of each study location in the job. This allows the analysis of the demographics of those who go to McDonald’s 1234 and separately, the demographics of those who go to McDonald’s 5678, for example. This breakdown also allows for a comparative analysis between these two locations. If McDonald’s 1234’s visitors have a household income of $150,000 and the McDonald’s 5678 across town draws visitors with a median household income of $50,000, it is valid to conclude that McDonald’s 1234 draws proportionally from upscale neighborhoods.
Census ID Summary
| Category | Sub_Category | Value | Devices Count | Sub_Cat Sort Order | Unit |
In contrast to the By_Rollup file, the By_Census_ID_Summary file would provide one overall census metric for all locations included as study polygons. This is provided by using a weighted average methodology. So carrying through the above example of two McDonald’s, a simplistic analysis would assume the median HHI of all visitors was $100,000 (averaging the $150k from McDonald’s 1234 and $50k from McDonald’s 5678). However, using a weighted average will weight the singular demographics metric toward the household statistics associated with the majority of visitors. So if 90% of the visitors come from the $50k neighborhood, but only 10% of the visitors come from the $150k neighborhood, the median reported value will accurately be reported as much lower than the simplistic $100k figure.
Census ID
| Census ID | Category | Sub_Category | Value | Devices Count | Sub_Cat Sort Order | Unit |
Finally, the non-aggregated By_Census_ID file provides a breakdown of demographics by census id. This file contains the raw data that feeds into the Summary file discussed above. So in the US, the report provides all the Census Block Groups that visitors had CELs in and the associated demographics. The report is most useful for diving deeper into the visitor profile. If the data is skewing higher income, as in the example above, you can use this file to identify the specific Census Block Groups that are higher income or to analyze which Census Block Groups most visitors were coming from.
Interpreting the Contents of the File
| Column Headers | Descriptor | File Notes |
|
rollup_name |
Name of the input polygon |
Only in by_rollup.tsv |
|
census_id |
Geographic census boundary ID |
Only in by_census_id.tsv |
|
category |
Name of the census attribute |
Example: Education |
|
sub_category |
Census attributes of the category |
Example: High School, Bachelor Degree |
|
value |
Calculated weighted average |
|
|
devices |
Number of unique visitors with a CEL seen within the polygon |
|
|
sub_category_sort_ order |
Level of demographic focus for the selected attribute |
Enables the logical presentation of the data in the order outlined in this column, such as least education to most or youngest to oldest. |
|
unit |
Unit type of the value |
Indicates whether the metric is in dollars or percentage |
Use Cases
Azira’s Demographics Report provides another insight into the profile of the visitors to a location. Depending upon the country, the following demographic attributes are reported:
- Income
- Age
- Education
- Ethnicity
- Median Home Value
- Median Income
The report answers questions such as:
- Am I drawing customers from low or high-income households?
- What is their education level?
- What is the median income of visitors to my location?
- How do the demographics of one location compare to another?
- Has the demographic profile of my visitors changed after COVID-19?
In which countries is the Demographics Report available?
Where does Azira source their census data from?
The American census data is sourced from the US Government with the latest-available American Community Survey. The Australian census data is sourced from DataPacks provided by the Australian Bureau of Statistics.
Why do my demographic percentages not sum to 100?
The census data has some census areas that have a documented negative number. This negative number should be interpreted as the census bureau not having enough data (or any data at all o the specific statistic) for that census area. In that case, if a visitor is from one of these areas, that area wouldn't have been included in the final output of that specific bucket (like a household education level). However, the total device count is still used to create the percentages, leading to a situation where sometimes, a total across buckets is not equal to 100% of devices.
👋 Do you have questions about anything on this page or in our help center? Contact our Support team!