Geo-demographic segmentation is based on the premise that people of similar backgrounds, interests and means tend to gravitate into communities. Marketers have been able to harness this sociological phenomenon by using models, which take abstract distributions of demographic characteristics, and synthesize them into a very descriptive, compact, easy-to-understand neighborhood lifestyle typology.
Commercially available geo-segmentation systems are highly descriptive marketing tools. The segments describe the neighborhood as a whole by combining demographics with behavioral data. During analysis, an index is calculated. This calculation compares a particular clusters demographics or behaviors against other clusters. An index indicates a degree of propensity (based on previous behavior) for a cluster or segment to behave a particular way compared to other clusters.
While geo-segmentation systems are most definitely descriptive, they are somewhat less effective at being predictive. That is, they can be used to describe which clusters purchased products at higher rates than others however, to estimate potential sales or response rates they should be used in conjunction with other variables to produce a more robust model.
In other words, just because a particular cluster has a high propensity to purchase a product or service does not mean that one can accurately predict future sales using clusters alone. There is rarely a direct correlation between the propensity to purchase and the actual purchase rate. In the highly competitive retail environment a number of factors contribute to the sales & profitability of a particular location.
Having said that, these tools "out of the box" can be used for a variety of purposes – and can help retailers make more measurable and actionable decisions.
Geodemographics can be used to:
profile existing store locations
compare profiles of several stores
evaluate the resident population in a proposed trade area
describe current customers
identify sales trends over time by clusters
profile customers by distinct segments such as sales, products purchased, etc.
identify areas which have a large concentration of people similar to your best customers
– Can be done with minimal orientation and training
– Requires geocoding software, Profiler software & knowledge of database software (Access, FoxPro, etc.) for preparation of the database files
-Can be done with minimal orientation and training
-Requires basic knowledge of mapping concepts
– Requires the database to be analyzed over time be geocoded on a regularly scheduled basic. There after, database reporting tools can be set to summarize sales / units / customers by cluster. After a period of time it will become clear that some clusters are consistently high / low performers for particular product segments
Evaluation of Target Households / Population
– Site location reports can include summarized counts of best performing clusters. Requires mapping software. Can easily be generated by a novice user but may require some assistance with interpretation.
Data-driven models that incorporate sales factors, location factors, individual customer demographics and geodemographic clusters can be created to specifically tailor the descriptive and predictive capabilities of the system for a particular client. This is something best left to someone with a strong background in statistical modeling. Catenate works with several statistical consultants with various industry expertise to get the appropriate solution in place. Pricing for custom modeling is dependent upon data inputs as well as methodology and deliverables.