GeodemographicAnalysis

Download Report

Transcript GeodemographicAnalysis

Geodemographic Analysis
Claire M. Palmer
GIS 3130 Advanced Spatial Analysis
March 19, 2008
Outline





What is Geodemographic Analysis?
History
Geodemographics today
Case Studies
Conclusion
What is Geodemographic
Analysis?
12 geodemographic “groups” or “neighborhoods”
in Bristol, United Kingdom
Geodemographics…
The analysis of people by where they live
(Sleight, 2004)
Where you are says something about who you are
Linking people to places
Geodemographics
is based on two simple principles…
1.
Two people who live in the same area are
more likely to have similar characteristics
than two people selected at random


Geodemographics
is based on two simple principles…
2.
Two areas can be identified in terms of the
characteristics of the populace they contain,
using demographics and other measures.
Geographical areas can then be placed in the
same segment even though they are
geographically distant

Geodemographics…
 Analysis of socio-economic and behavioral
data about people
 Investigates the geographical patterns that are
structured by the forms and functions of
settlements
 Efficient discriminator of consumer behaviors
and aids market analysis
 Effective predictive tool for decision support
Where did this concept
come from
?
TOBLER’S LAW:
Things near each other are more alike
than things far apart
The ESRI Guide to GIS Analysis, Volume 2 p. 104
TOBLER’S LAW:
(examples)
 Climate of nearby areas
 House values
 High crop yields of neighboring farms
(same soil characteristics)
 Ethnic communities within cities tend to settle in
same neighborhoods
(relatives and others of same ethnicity tend to live
near each other)
The ESRI Guide to GIS Analysis, Volume 2 p. 104
TOBLER’S LAW:
(exceptions)
 Climate of two cities far apart on one side of a
mountain are more similar than to a closer city on
the other side of a mountain range
 Neighborhoods can change abruptly if separated
by a highway or a river
The ESRI Guide to GIS Analysis, Volume 2 p. 104
Geodemographic Analysis
is not new!
Chicago School of Urban Sociologists
1920-1930
Concentric Zone Theory (1925)
Ernest W. Burgess
(1886-1966)
Robert E. Park
(1864-1944)
Concentric Zone Theory
(1925)
Concentric zone theory was one of the earliest models
developed to explain the spatial organization of urban areas.
 maps social problems such as
unemployment and crime in certain
districts
 reveals the spatial distribution of social
problems and permits comparison
between areas
Present-day
Geodemographic Analysis
Geodemographics today…
 Cluster Analysis (Bailey & Tyron, 1970) – four
decades of census data & election results in the SF
Bay Area revealed that the aggregate political
behavior of the tracts stayed the same
 rapid growth in the amount of geographic
information collected about people and places
 geographic information handling technologies
such as GIS
 development of Geodemographic (GD)
classification systems
Geodemographic classification
systems…characteristics
 Private sector household-level databases prove
more relevant than census data
 Cluster analysis to identify similar
neighborhoods
 Three dimensions to households:
1) life-cycle needs
2) buying power
3) spending power
Geodemographic classification
systems…uses
Business
 Retail Management / Market Analysis
 Site Location – where’s the best places to open/close/re-brand
 Target Marketing – who are my prospects and where can I find them?
 Media Analysis – which Newspapers/TV stations/Radio/Web sites are
most cost effective?
 Market Size Estimation – what is the local market size for my
product/service
 Recruitment & Retention – which customers are most like to
stay/churn
Government
 Resource Allocation / Facility Planning
 Health, education, law enforcement, social regeneration
 Justify the appropriateness for sought allocation
Commercially available GD systems...
USA
UK
Claritas

62 clusters define each
neighborhood in the US

15 social groups within
each cluster, by the degree
of urbanization
ACORN
A Classification Of Residential Neighborhoods
Mosaic interface...
Case Study:
Geodemographics & Recycling in Surrey, UK
 Evaluate demographic and geographic factors
in recycling motivation
 Survey households (2 urban, 2 rural)
urban study zones:
rural study zones:
Worcester Park (deprived)
Longmead Estates (least deprived)
Middle Burne (deprived)
Upper Hale (least deprived)
Case Study:
Geodemographics & Recycling in Surrey, UK
 ArcGIS Network Analyst for road networks
Case Study:
Geodemographics & Recycling in Surrey, UK
Case Study:
Geodemographics & Recycling in Surrey, UK
Case Study:
Geodemographics & Recycling in Surrey, UK
Findings
 higher recycling in affluent zones
 retired residents recycled most
 People in rural areas travel further to recycle but
travel time was not significantly different between
urban and rural groups
 People in urban areas tend to seek out the nearest
bring-site whilst those in rural areas choose more
of a variety
Case Study:
Geodemographics & the
Financial Service Industry in the UK
GIS & geodemographics for a competitive edge




Profiling & finding customers
Credit scoring
Branch location
Fuzzy geodemographics
Case Study:
GD & the Financial Service Industry in the UK
Profiling & finding customers
1. Produce a list of account holders of various types
(current account, mortgages, savings, etc.)
2. Assign account holder to a census tract by their address
3. Use the GIS to find new customers by searching for
areas that contain the same geodemographic mix as the
existing customer profile, preferably where existing
market share is low
Overlay
Case Study:
GD & the Financial Service Industry in the UK
Credit Scoring




Increased cases of bad debt
One response to this phenomenon has been for many
banks to close branches in less affluent parts of U.K. cities
Geodemographics could be used to identify potential
market areas where mortgages and loans might be more
difficult to recover
Consumers could be rated on the likelihood of their ability
to repay based on existing knowledge of the
geodemographics of past defaulters
Overlay
Case Study:
GD & the Financial Service Industry in the UK
Branch Location
Case Study:
GD & the Financial Service Industry in the UK
Customer Profiling & Fuzzy Geodemographics




Geodemographic products are general purpose systems
with often limited data sets
‘‘Smarter’’ (or ‘‘fuzzy’’) geodemographic systems are not as
reliant on the usual single descriptor
Fuzziness in attribute space – a locality may differ by only a
very small amount in the geodemographic classification
from its neighbors but still be assigned to a very different
cluster
Fuzziness in geographical space – ecological fallacy
problem and/or MAUP. Two neighboring census tracts may
have very different classifications, but often people who live
in neighboring tracts still demonstrate characteristics and
economic behavior similar to that of their neighbors
Solution
Display all clusters, esp., clusters similar to dominant cluster
Case Study:
GD & the Financial Service Industry in the UK
Customer Profiling & Fuzzy Geodemographics
Case Study:
GD & the Financial Service Industry in the UK
Summary
 GIS is a useful support tool for
geodemographics – data storage and display,
overlay of non-census data
 Key target groups can be identified, enabling
focused marketing and credit scoring
 Buffer & overlay analysis key for catchment
area analysis
Conclusions / Critiques
 Geodemographic data is more robust than census data
(lifestyles & behaviors)
 GIS is a useful support tool for geodemographics
BUT…
 Problem of decay & inaccurate data
 Represents relatively crude averages of the population
 Ecological fallacy (the fallacy of homogeneity across
the neighborhood)
 "The 'strategic intent' of geodemographic systems is
replete with metaphors of vision, insight, omniscience,
prediction, manipulation, and control. " (Pickles, 1994)
 Concern for individual privacy rights
 Serendipity factor lost?
Thanks for your patience!
comments or tomatoes welcome