FINAL EXAM FOR MKT 555 – SPRING 2023

Please submit your answers electronically by uploading on Canvas by 1 PM on April 28 (Friday), 2023.

1.    Suppose that you are the owner of a brand of energy drink, called ACTIVE (which competes with brands such as Red Bull, Monster etc.). Suppose that the price elasticity of ACTIVE is ‐2.5 (as revealed by the analysis of historical data on sales and prices of ACTIVE over the past 2 years).

a.    You are contemplating a price cut of 10 % on ACTIVE. What is the minimum profit margin on ACTIVE at which this price cut will be profitable?

b.    Assume that ACTIVE’s profit margin is equal to the minimum profit margin calculated under question 1 (a). What is the maximum percentage price cut that you can offer on ACTIVE without decreasing your profit?

c.    Suppose it costs $1 to produce a can of ACTIVE. What is the optimal price that you must charge for ACTIVE per can?

d.    Suppose that you introduce a second brand of energy drink, called HIGH ENERGY. Now, you own a portfolio of two brands, ACTIVE and HIGH ENERGY. In this scenario, will you find it optimal to maintain the  price of ACTIVE at the optimal  price  calculated  under question 1 (c), or increase it or decrease it? Why? Explain the rationale for your answer (no calculations are necessary).

2.    What is the rationale for brand managers to offer trade deals? Clearly explain the difference between an off‐invoice trade deal and a scan‐back trade deal. Why does the retailer prefer the off‐invoice trade deal? How would a brand manager offer a suitably redesigned scan‐back trade deal in order to make the retailer accept it (and not demand an off‐invoice trade deal)? When going from an off‐invoice trade deal to the redesigned scan‐back trade deal, the retailer’s profit does not decrease, however, the brand manager’s profit still increases. Why? In other words, how does the  redesigned scan‐back trade deal yield a “win‐win”  for the brand manager and the retailer?

3.    Using the weekly advertising spending (thousands of GRPs), prices ($) and unit sales (in hundreds of thousands) of a brand of perfume in the United States using 104 weeks of data from 2019 and 2020, a brand manager has estimated the following Scan*Pro model:

ln (Unit Sales) = 3.6 – 2.55 * ln (Price) + 0.16 * ln (Ad)

a.    What is the advertising elasticity of this perfume brand? Explain its meaning in words.

b.    The total annual revenues for this perfume brand for 2020 were $4.2m. Use the Dorfman‐ Steiner rule to figure out the optimal advertising spending (in $) for the perfume brand for 2021.

c.    Suppose Adpp1 and Adpp2 refer to the weekly advertising spending (thousands of GRPs) in the previous week, and in the previous‐to‐previous week, respectively. Assume that the brand manager has estimated the following extended Scan*Pro model, which accounts for carryover effects of advertising, for this perfume brand:

ln (Unit Sales) = 3.8 ‐ 2.5 * ln (Price) + 0.15 * ln (Ad) + 0.1155 * ln (Adpp1) + 0.0866 * ln (Adpp2).

What is the total advertising elasticity for the brand of perfume?

d.    Use the results reported under question 3 (c) above, as well as the Dorfman‐Steiner rule, to figure out the optimal advertising spending for the brand of perfume (in $) for 2021. How different is this answer compared to that obtained under question 3 (b)? Which one would you trust? Why?

4.    Barry Nalebuff is an entrepreneur who has recently launched a bottled organic tea brand called Honest Tea (which can be bought online). He has decided to bid on the Google keyword, “Healthy Beverage,” and would like to know how much to bid on this keyword. For this purpose, he decides to use the PROSAD decision support system. Barry learns from Google that 25,000 daily searches happen, on average, on the Google keyword, “Healthy Beverage.” Suppose that among the people who both search on the keyword “Healthy Beverage” on a given day and also then click on Barry’s sponsored ad, 0.5 % are successfully acquired (i.e., become customers of Honest Tea). Suppose further that the annual profit contribution to Barry per acquired customer is $500.

a.    Barry Nalebuff wants to invest in tracking online clickstream data using Google Analytics and, therefore, understanding the impact of various ranks on clicks. Right now, before having gathered such online data, what would be a conservative bid for Barry Nalebuff to make for the Google keyword, “Healthy Beverage”? Explain your answer.

b.    Over the first quarter of launch, Barry subjectively revises his bids on the Google keyword, “Healthy Beverage” on a daily basis. As his bids change, his sponsored ad’s rank changes and, therefore, the clicks on his sponsored ad change as well. Based on the data collected during the first quarter (using Google Analytics software), Barry estimates the relationship between click‐through rate (CTR) and sponsored ad rank, and it is as follows: ln (CTR) = ‐ 1.5 ‐ 0.5 * RANK. Barry also estimates the relationship between cost‐per‐click (CPC) and sponsored ad rank, and it is as follows: ln (CPC) = 1 – 0.2 * RANK. Which increases at a faster rate going from an inferior rank (say, X+1) to a superior rank (say, X): click‐through‐ rate or cost‐per‐click? Does this imply that Barry’s optimal bid for the keyword “Healthy Beverage”  would be higher than, or lower than the conservative bid, calculated in question 4 (a) ? Why?

c.    Suppose Barry Nalebuff bids the optimal bid for the Google keyword, “Healthy Beverage” as recommended by the PROSAD decision support system. What rank would he obtain for such a bid?

d.    What would be the transactional profit that Barry Nalebuff obtains from the above bid?

5.    The CVS store at Clayton and Big Bend ran a 40 % off price promotion on Haagen Daaz ice cream during the week of Thanksgiving. The store observed a 400 % sales increase (“gross  lift”) for Haagen Daaz ice cream that week. That is, weekly sales of Haagen Daaz during the week of promotion were 5 times the usual amount of weekly sales observed for Haagen Daaz ice cream at the store. However, not all of the gross lift can be called incremental to CVS. The part of the gross lift that can be called incremental to CVS (“incremental lift”) is a sum of 6 components:

a.    Consumers who would have bought Haagen Daaz at nearby competing stores (Schnucks, Walgreens etc.) during the week of Thanksgiving ended up buying HaagenDaaz ice cream at CVS instead,

b.    Consumers who would have bought other ice cream brands (Ben & Jerry, Breyer etc.) at nearby competing stores  (Schnucks, Walgreens etc.) during the week of Thanksgiving ended up buying Haagen Daaz ice cream at CVS instead,

c.    Consumers who would have bought Haagen Daaz at nearby competing stores (Schnucks, Walgreens etc.) during the week following the Thanksgiving  week ended up buying Haagen Daaz ice cream at CVS instead during the week of Thanksgiving (i.e., “accelerated” their ice cream purchase by a week on account of the promotion),

d.    Consumers who would have bought other ice cream brands (Ben & Jerry, Breyer etc.) at nearby competing  stores  (Schnucks,  Walgreens  etc.) during  the  week  following  the Thanksgiving week  ended up buying Haagen Daaz ice cream at CVS instead during the week of Thanksgiving (i.e., “accelerated” their ice cream purchase by a week on account of the promotion),

e.    Consumers who would have anyway bought Haagen Daaz at CVS during the week of Thanksgiving (even if the promotion were absent) ended up buying more quantity of Haagen Daaz than usual because of the promotion (without decreasing their future purchasing), and

f.     Consumers who do not otherwise buy ice cream (either at CVS or elsewhere) ended up buying Haagen Daaz ice  cream  at CVS during  the  week   of  Thanksgiving   (without decreasing their purchase of other products at CVS).

Among the above 6 components, it is impossible to figure out the first 4 since CVS does  not observe consumer purchasing behavior. at Schnucks, Walgreens etc. However, CVS has its internal data warehouse, which contains sales transactions at the store‐ level (collected using checkout scanners), as well as individual purchasing activity (collected using loyalty cards). How would CVS use  its  data warehouse to  effectively  estimate the  incremental  lift,  even though  CVS  cannot

directly estimate the incremental lift as the sum of the above 6 components? Explain clearly.

6.    In the dataset, “FinalExam‐SpecialtyPractices.xls” is data from Syntex Corporation on their specialty practices’ salesforce across 9 divisions (General Practice, Family Practice, Internal Medicine etc.). In the worksheet titled, “Data” are the judgment  inputs  obtained from  sales managers to five survey questions, in each of the 9 divisions.

a.    Using these judgment inputs, estimate two parameters – min, max – of the AdBudg model for each of the 9 divisions.

b.    Using the worksheet titled,  “Parameter Estim. Specialty” estimate two parameters ‐‐ c and d ‐‐ of the AdBudg model for each of the 9 divisions.

In the worksheet titled, “Optimal Salesforce Specialty” are given the profit calculations across the 9 divisions.

c.    Evaluate the optimal salesforce allocation across the 9 divisions, assuming that the maximum salesforce across the 9 divisions cannot exceed 1000. How does this allocation compare to the existing allocation i.e., 91.2 in General Practice, 79.4 in Family Practice etc.?

d.    What is the resulting total profit across the 9 divisions combined? How does it compare to the existing profit of $226.07m?