Marketing 469: Strategy and Tactics of Pricing

Marketing 469: Strategy and Tactics of Pricing

Exercise 1: Estimation and Use of Market Response Functions

Purpose: This is an exercise in estimating own-price and cross-price elasticities, and understanding its implications for market structure and pricing decisions.

Price elasticities provide one possible way to understand the competitive structure of a market. We can also use cross-price elasticities to understand how much a brand can attract sales from competing brands (also called a brand's competitive clout), and how much a brand can lose sales to competing brands (also called a brand's competitive vulberability).

Problem: The sales data of 12 SKUs of orange juice in one supermarket for 110 weeks are available in an EXCEL file. The file name is:

OJData.xls

The entries are:

Column 1 SKU name (Tropicana 64oz., Tropicana 96oz., Florida National 64oz., Tropicana 64oz., Minute Maid 64oz., Minute Maid 96oz., Citrus Hill 64oz., Citrus Hill 96oz., Tropical Fresh 64 oz., Florida Gold 64 oz., Store Brand 64oz., Store Brand 96oz.)

Column 2 Serial number of brand (1, 2, ....12)

Column 3 Week (1, 2, 3, .....110)

Column 4 Natural log of weekly sales (in equivalent units).

Columns 5-16. Natural log of the prices of the 12 SKUs in the week.

Columns 17-28. Intensity of feature promotion in the week

(a) Prepare summary statistics for the 12 SKUs (including average weekly sales, average price, average feature intensity). Discuss what you can discern from these summary statistics.

(b) Plot ln(sales) against ln(price) for each of the 12 SKUs. Please state in a few sentences what you are able to learn from a visual analysis of the 12 graphs.

(c) Using these data, the following model was estimated for each of the 12 SKUs/brands:

See the attached SAS output files. Describe the fit of the model for each brand. Note that all analysis in this exercise are done at the SKU level (that is, for each specific brand-size combination). The parameter estimates can be seen in the SAS output file.

The notation is:

i = SKU number (1, 2, 3, or 4)

Sit = Sales of SKU i in week t

Pit = Price of SKU i in week t

PIit = Feature intensity of SKU i in week t

= parameters of the model

Note that this model is a Log-Log model. The own-price elasticity for Brand i is and cross-price elasticities are .

(d) Prepare a 12x12 matrix of own and cross-price elasticities. Enter in the 12x12 matrix only coefficients that are statistically significant.

1. Using this matrix, identify which SKU(s) is (are) most vulnerable to promotion activities. Also identify which SKU(s) has (have) most promotional clout in the market.

In other words, consider arranging the beta coefficients from the regression analysis done

...