The marketing department of a food retailer wants to analyze the profile and behavior of mineral water drinkers in London. It buys a sample survey from a market-research agency that describes the profile of U.K. inhabitants (i.e., the reference population) because data for the London area was not available.
However, the company has a database that characterizes Londoners in the same way except for their water consumption (i.e., the target population). Is it possible to reveal the Londoners' drinking behavior?
ISoft's AC(2) sets up a combination of criteria that best filter the reference population and breaks it down according to the filter. It turns out that gender is the first criterion to separate the refere
nce population. On the next level, filters can be different in different branches. It separates the female branch according to the population's age and filters the socioprofessional characteristics of the male branch. This process is repeated until groups of exclusive drinkers of one brand are isolated and the decision tree is completed.
To segment the mineral water drinkers in London, AC(2) applies the same combination of filters to the target population and breaks it down accordingly.
Decision Tree
illustration_link (9 Kbytes)
Market Segmentation
