Consumer Demographics

From geoTribes

Consumer Demographics is a synthetic reconstruction of the 2021 Australian Census that’s optimized for database profiling

geoTribes AUS Consumer Demographics

Data Overview

geoTribes Consumer Demographics is a synthetic reconstruction of the 2021 Australian Census that’s optimized for database profiling. What makes our Consumer Demographics special is its high level of match resolution.

Instead of matching databases at SA1-level with only 60K levels, our match-key of SA1 + Dwelling Type + Age Band + Gender matches at 50 times the precision in a secure, compliant, virtual and instantaneous manner with over 3.1M match levels.

This means that within a particular SA1, people in different age and gender cohorts will have different financial, cultural and life stage profiles. This higher level of accuracy really pays off in:

  • Cluster Analysis,
  • Predictive analytics and
  • Benchmarking database cultural composition against the relevant population

This unique capability, based on our Synthetic Population Models, means that the demographic profiles on customer databases are much more accurate and responsive to an individual customer’s age and gender rather than just the SA1 they live in, leading to improved business analytics and better training of machine learning (ML) models for data science.

A trial product is provided at no charge for a 90-day evaluation.
The trial dataset includes data for the SA4 (Statistical Area Level 4) of 125 – Sydney – Parramatta.
A data dictionary for the geoTribes: Consumer Demographics can be found at

The data is built in a sophisticated process based on Synthetic population models that accurately represent the total household and person populations of each country, with many millions of georeferenced individual synthetic records and deep sociodemographic profiling.

Expected Workflow:

  1. User requests this data from geoTribes (via the Request button);
  2. geoTribes and user agree to commercial terms;
  3. geoTribes provides user with details of the share via UDTF;
  4. User appends geoTribes Consumer Demographics to their database through the Snowflake share;
  5. geoTribes Consumer Demographics available on user database for analytics, visualization and other use cases.

geoTribes: Consumer Demographics (AUS) are supplied as a Snowflake UDTF that accepts a database with the following schema as input:
– RECORDID (unique record ID)
– MATCHKEY (Either 11-character 2021 SA1 code or 4-character postcode)
– ADDTYPE (2-level coding of dwelling type)
– AGEBRK (13-level coding of age in 5-year bands)
– GENDER (2-level coding of gender)

The UDTF will return the following as additional fields for each RECORDID (expressed as probabilities to 3 decimal points):
– Household Demographics – 39 Variables
– Person Demographics: Cultural – 49 Variables
– Person Demographics: Work and Life – 85 Variables

Cloud Region Availability

Asia Pacific (Sydney)






Available on request

Geographic Coverage

Australia (by Address)


Business Needs

Population Benchmarking

Cluster Analyses

Predictive Analytics & Machine Learning

Population Benchmarking



After appending the Consumer Demographics variables to your database, create a summary profile by averaging the values. Next create a parallel profile for a random population sample provided by RDA. Then index the customer profile against the random profile to see how the extent to which the two profiles differ.

For example, this is particularly useful for determining the relative cultural diversity of a customer database and the proportion of the database that comprises high income earners.

Cluster Analyses



After appending the Consumer Demographics, define an attribute vector for clustering, for example using K-means. This may also include indigenous (behavioural) measures from the customer database.

This type of hybrid segmentation is able to offer real insights into the underlying segments on a customer database.

Predictive Analytics & Machine Learning



After appending the Consumer Demographics, select a target variable and define a vector of predictors, that may also include indigenous (behavioural) measures from the customer database.

An example of this type of model would be predicting Funds Under Management (FUM) and then comparing actual vs predicted values to identify value gaps.