For the articles as a whole, be able to answer:

• What are these seven steps trying to accomplish?
• What are the seven steps?

Obviously, this analysis assumes that location – and more specifically, characteristics of the primary market area – determines the performance of an outlet.  What else determines the performance of an outlet?  Might location make up a higher proportion of the determinants of performance when performance is measured by the indicators that Thrall et al. used, versus other possible indicators?

In an earlier lecture, I suggested that it generally doesn’t make sense for retailers to open only one outlet in a metropolitan area, and gave several reasons, using department stores as an example.  However, in “Retail Location Analysis, Step Four,” Thrall et al. explain why some of Red Lobster’s better performing outlets are the only ones in a metro area.  Why?  Why the difference in recommendations?

Figures 1-5 in “Retail Location Analysis, Step Four,” compare the characteristics of Red Lobster markets and overall metropolitan-area (MSA) markets, by noting the distribution of ages, family sizes, house prices, rent levels, and African-American household incomes.  How did the authors get these data for Red Lobster?

DATABASE  MARKETING
How does Hughes define “database marketing”?

GEODEMOGRAPHICS

Birkin provides a narrow and a broad definition of geodemographics.  What are they?

What are the key uses of geodemographic analysis, as presented by Birkin (Section 6.2.2)?  Compare these with the marketing uses described by Mitchell in the second half of her article.

What are the key issues or problems that Birkin identifies in the use of geodemographic analysis?
How does Birkin define “geolifestyles” data?

AREAL  UNITS
In class, we've discussed the pros and cons of using Census tracts and ZIP Codes for defining trade or marketing areas.  The former conform to government data and are explicitly geographic.  The latter conform to much consumer data, and to businesses’ and agencies’ own databases, but are not necessarily geographic, and can change quite radically.

Carroll suggests that many peoples’ initial distaste for geodemographic cluster analyses results from their interpreting the clusters in a “univariate” way.  He argues that, when recognized as “a multivariate creation,” clusters make much more sense.  What does he mean by this?  Do you agree?

Provide two examples of how you might make use of geodemographic clusters.