
Segmentation Analysis
Segmentation or clustering refers to a family of statistical methods that
are used to identify relatively homogeneous groups of cases/individuals
based on selected characteristics. Why segmentation? Driver modeling
generates a strategic understanding of the opinion landscape and thereby
prioritizes key messages/characteristics/drivers. Segmentation analysis is
crucial for the tactical implementation of messaging strategies.
Classification algorithms show who to target and how best to target them
based on selected demographic, geographic, and other factors.
Consider a scenario... a political advocacy group has used driver modeling
to prioritize the top message that will be communicated to likely voters in
a key battleground state. A narrowcast communications campaign of mail and
phone contacts will be initiated, but the question is: which individuals
should be contacted in order to generate maximum impact. Geo-demographic
segmentation is used to determine classify the population into a core
mobilizeable group consisting of married women in 5 counties and a
persuadable group comprised on hispanic men under the age of fifty-five. The
advocacy group customizes its mailings and phone contact messages to suit
each audience and thereby achieve the greatest effect on the opinion
environment in the state.
|




|