Pitney Bowes MapInfo Vertical Mapper - Manual (Page 65)

Chapter 4: Creating Grids Using Spatial Models
User Guide
63
Understanding the Trade Area Analysis Model
The Trade Area Analysis models enable you to generate trade areas around stores or other services
based on the probability of an individual patronizing a particular location. It is possible to identify
market islands in a network of competing stores using criteria such as the attractiveness of a store
and the distance one must travel to get to the store. Store attractiveness can be defined by
parameters such as total floor space or shelf space, number of parking spaces, age of the store, or
any combination of elements that defines its appeal. Using commercially available wealth and
market profile information for potential customers located within each trade area, it is possible to
estimate revenue for store locations and model the influence of competing stores or the effect of
adding or removing stores.
Trade Areas
Trade areas can be defined by simple circular boundaries around a store location. These trade
areas are easy to visualize and construct and provide a standard for comparing stores. Trade areas
determined in this way do not, however, account for the existence of competing stores and assume
that the store has a monopoly over customers within the area. There is the common sense concept
that, all other things being equal, a person is likely to shop at a closer store rather than a more
distant one. This concept of likelihood forms the basis for defining probabilistic trade areas.
Defining the likelihood or probability that a given customer will patronize a certain store requires the
use of a spatial interaction model. Examining movement over space for activities such as shopping
trips, commuting to work, or migration patterns, became popular with the use of gravity models in the
1960s and 1970s.
Gravity models are based on an analogy with Newton’s theory of gravitational attraction. The degree
of attraction between two objects is based on the size of the objects and the distance between them.
Objects that are closer together will exert a greater pull on each other compared with objects that are
farther away. Larger objects have a greater gravitational pull than smaller objects. Refinements to
the gravity model were made by Professor David Huff.
The Huff model is still one of the most popular models for predicting retail customer behaviour. It
enables the mapping of the neighbourhoods from which each store derives its patronage. The
answer is not a single circle or polygon but a probability surface or grid. This probability grid can be
contoured to produce regions of patronage probability. The key feature implicit in the probability
surface is that it accounts for overlapping trade areas.
Using a Huff model, you can calculate the probable trade area of a single store (next figure) or
compute the patronage probability values for every store and extract the maximum probability value
at each grid location. You can then use this information to determine areas where people are least
serviced or areas where there is great competition for a customer’s business, that is, there is no
preferred store location clearly winning a customer’s patronage.