Transcript Business Applications - University of Maine System

GIS for Business and
Service Planning
M. Birkin, G.P. Clarke, & M. Clarke
Jim Campbell
SIE 510
Feb. 17, 2004
Business Applications
• Outline of this presentation
– summarize article
– comment on article
– look briefly at some other current strategic
or planning business application areas
– examples
Caveats
• Published in 1999, latest citation is 1997
– out of date in many ways
• Focus is on "bricks and mortar" retailing
– minimal mention of B2B
– no mention of electronic retailing, Internet
– minimal mention of non-location based
services, e.g, brokerage, insurance, etc.
– that said...
Article Organization
•
•
•
•
•
Introduction
The Corporate Growth Model
Geography of Service Provision
GIS and Models in Retail Analysis
Conclusions
Author’s Main Points:
• GIS are weak at predictive modeling for
strategic business purposes
• There is great potential for increasing
the use of modeling tools within a GIS
environment to create “Intelligent GIS” i.e, “spatial decision support systems”
Decision Support Systems:
•
•
•
•
data to
information to
intelligence to
action
GIS are Good at ”What is”
Not at ”What if"
• Authors' solution: customize GIS so that
modeling tools become central for data
analysis
Modeling Corporate Growth
• Little use of GIS in modeling growth
– managers and decision makers are not
familiar with GIS
– companies feel they have a growth plan:
expand from HQ, or tiered focus on
population centers
– decision makers feel market is saturated
and there is no point to further location
research
Modeling Corporate Growth
• In fact, there is often significant variation
in market share, per store sales, market
saturation, etc. across and even within
regions
• In what little location research does
exist, the effect of competition in an
area is seldom factored in
Geography of Service
Provision
• Companies often have a good idea of
national market share but know much less
about regional and local share
• Data exists for many industries but does not
rise to level of information: most analysis is
"eyeball level"
• Analysis needs to happen at different levels,
e.g., not only locations of customers but types
of customers - Wal-Mart vs. Target
Geography of Service
Provision
• Need to understand combinations of
customers, sales, markets
• Banking example: who contributes most
to profits?
– suppose it is the customer who has a
mortgage, credit card, & large deposits
• Strategic marketing decision: target more
customers like those - where are they?
Geography of Service
Provision
• Advantage of GIS is storage of a large
amount of data that can be accessed
and placed in an easily understood
framework
"Intelligent GIS" Models in
Retail Analysis
• Authors recognize there is some
modeling capability in GIS software but
assert it is not flexible enough, e.g.,
health care center location vs. car
dealer location
• Solution: integrate GIS with modeling
systems to produce actionable
intelligence
"Intelligent GIS" Models in
Retail Analysis
Low
Market Share
High
Low market share
High market share
High sales per branch
High sales per branch
Action: Extend branch network Action: Maintain status quo
Sales
per
Branch Low market share
Low sales per branch
Action: Reconfigure network
High market share
Low sales per branch
Action: Rationalize network
GIS and Models in
Retail Analysis
•
•
•
•
Marketing and store revenue predictions
Launch of a new product
Mergers and Acquisitions
Optimizing retail networks
Marketing and Store
Revenue Predictions
• Identify existing or potential locations
• Establish buffer (e.g., driving time,
physical distance)
• Analyze characteristics of consumers in
area
• Take competitor locations into account
• Project potential sales and income
Launch of a New Product
• Thoroughly understand the nature of the
product
• Understand characteristics of potential
customers
• Match store placement with locations of
potential target customers, e.g., model
of car, type of furniture
Mergers and Acquisitions
• Analyze regional market share of
potential merger/acquisition candidates
• Match with acquiring company's market
shares
• ”Maximize opportunity, minimize impact"
Optimizing Retail Networks
• GIS itself cannot provide all necessary "what
if" information
• Need to use techniques such as linear
programming or non-linear OO programming
• Supply chain optimization is good use of GIS
if distance minimization is criterion
• Supply location choice often more
complicated, e.g., locate new car dealer in an
area without cannibalizing sales from existing
dealers
Conclusions:
• GIS are weak at predictive modeling for
strategic business purposes
• There is great potential for increasing
the use of modeling tools within a GIS
environment to create ”Intelligent GIS"
i.e.,"spatial decisions support systems".
Conclusions:
• The solution is to design highly flexible
information systems "which combine
database manipulation and high-quality
map and graphical output with spatial
modeling techniques."
Comments
• Article is light on detail but two books
expand on these ideas, again with a
retail emphasis:
– Intelligent GIS (1996)
– Retail Geography and Intelligent Network
Planning (2002) (which has significant
discussion about e-commerce)
Comments
• Article is out of date at this point in time:
the authors' plea has largely been
answered:
– built-in GIS capabilities can do many of the
things authors hoped for 5 or 6 years ago
– developments in GIS extend modeling
capability, e.g., ESRI BIS products
“Neighbourhood classifications characterize
small areas by different types of
neighbourhoods according to the household,
housing and socio economic characteristics of
their residents. Such systems have been built
by commercial organisations in some 20
different countries around the world, often
using census statistics, and have proved very
practical tools for organisations, whether in the
private or the public sector, that wish to target
their communications or activities at particular
population groups.”
Geodemographics
www.vespucci.org
as of 2/15/02
Comments
• Rudimentary (or greater) GIS capability
is built into Oracle and other enterprise
database systems
• Many commercial data providers are
now offering some level of geographic
information as part of their packages,
e.g., D&B (B2B) www.zapdata.com,
InfoUSA (consumer) www.infousa.com
Business Applications
Using GIS in 2004
• We've seen many applications of GIS to
business in previous presentations, e.g.,
network design, location based
services, etc.
• These are mostly operational
applications. Focus here is on decision
support
Marketing Decision Support
• Finding more customers like your best
current customers in a given area, or
finding where they are in any area
• Deciding whether to provide service in
an area: logging & mapping requests for
service - often an automated function,
e.g. cable TV, high speed Internet, dsl
Marketing Decision Support
• Expansions of service areas
– analyzing demand for a product or mix
– analyzing competition
– assessing risk
– assessing Return On Investment (ROI)
Marketing Decision Support
• Advertising
– data mining - combining information
sources - CRM and GIS
•
•
•
•
•
who customers are
where they are
what they're like
what they want
what they can buy
Marketing Decision Support
• Profiling current or potential customers
– psychometrics - (radio antenna example)
•
•
•
•
Cultural Creatives
Blue Collar Acquirers
LOHAS (www.lohasjournal.com)
etc.
Marketing Decision Support
•
•
•
•
•
•
•
•
Strugglers
Believers
Makers
Strivers
Fulfilled
Actualizers
Achievers
Experiencers
SRI VALS System
Marketing Decision Support
• “Experiencers appreciate the
unconventional. They are active and
extroverted, and they like stimulation by
the new, offbeat, and risky. Their
lifestyle focuses on fashion, exercise,
socializing, and sports.”
SRI VALS System
Network Planning and
Reconfiguration
• Airline scheduling - topographical
algorithms applied to routing, hub and
spoke modeling
• Telecommunications planning and
outage response (AT&T example)
Real Time Planning for
Resource Reallocation
• Long distance trucking
– fuel savings
– customer satisfaction
– regulatory compliance
• Cab or limo scheduling
– minimize down time
Some GIS Business
Application Issues
• Privacy concerns
• Artificial intelligence will be increasingly
embedded in decision making systems
– knowledge management/expert systems
as component of decision support systems
– human interaction with and/or override of
system recommendations?