In some large and accurately sampled primary surveys, such as Mediamark's Survey of American Consumers, a comprehensive selection of data items comparable to and consistent with census items (commonly termed "demographics") are included among the questions asked and the respondents are geocoded (have census area codes derived from their detailed address information added to their records). In such cases, data derived from respondents' actual small area contexts can be appended to their survey records and geodemographic inferences can be constructed and extrapolated to all areas based on a model relating the census surrogates to the dependent variables measured in the survey. This is a powerful methodology in the arsenal of geodemographic research methods which greatly extends the value of primary data resources in the form of ad hoc small sample research.
Nevertheless, geodemographic targeting is often accomplished with excellent results given only secondary data summarized to small geographic areas along with a dependent variable measured in the same small area geographic framework. Counts of direct mail promotions mailed divided by orders received from the mailing within each of the 35,000 five-digit ZIP Codes provide a typical example of such a dependent variable (percent response). Others include similar relative measures derived from: customer files, administrative records, sales transactions, subscriber files, credit purchase receipts, warranty card returns, coupon redemptions, point of sale customer address surveys, membership records, address compilations by interest, title or past ordering actions as well as many other geographically comprehensive individual data resources.
about geodemographic data resources...