Question 1

1. Using the demand data in Demand.csv, estimate a log-log demand curve for the existing subscription service. Use whichever software you see fit (e.g. R, Excel etc). Specifically, regress log quantity on log price. Write the mathematical equation for this model. In a welllabeled table, report your regression estimates. Also be sure to report standard errors and tstatistics for each parameter as well as an R-square for the regression. Interpret your parameter estimates. What concerns (if any) do you have with this regression? Do you feel comfortable making a recommendation about pricing based on the evidence? If so, make a recommendation. If not, explain any concerns you have.
2. Rerun the above regression while accounting for the possibility of having different base demand for each season. Write the mathematical equation for this model. In a well-labeled table, report your regression estimates. Also be sure to report standard errors and t-statistics for each parameter as well as an R-square for the regression. What concerns (if any) do you have with this regression? Do you feel comfortable making a recommendation about pricing based on the evidence? If so, make a recommendation. If not, explain any concerns you have.
3. Would your conclusions in (2) change if you also included controls for "region" in your regression? Explain why (or why not).

Question 2

One particular area of interest for the firm is whether changing the specific content offered in the subscription would have a material impact on a consumer's willingness-to-pay. To form a basis for your quantitative analysis, you recently conducted a large-scale conjoint survey to a sample of 1000 subjects, selected to mimic your current target market of prospective customers. Currently, your firm offers a single monthly subscription plan consisting of content from the following television channels: NBC, CBS and ESPN. However, for the purposes of the study, several other channels were also tested. Each subject was asked to rate 15 different subscription profiles. In each profile, the subject was asked to evaluate the content packages on a 1-month basis (i.e. ratings are on a per-month of service basis). In addition, the study included the specific phone (device) model since different models have different displays, making the "viewing experience" device-specific. Three specific phone models were tested: Nokia, Samsung and LG. Subjects were given live demonstrations of the service on each of the devices before completing the survey. Currently, the firm only sells its subscription service for usage on the LG phone.

Preliminary results from this survey have already been tabulated by your marketing research team. Each row in the data file corresponds to a profile which was evaluated by the subject. In the preliminary analysis, the market research team has already clustered the subjects from the study into distinct groups (or "segments"). Your current task consists of using the data they have sent you to study demand for the subscription service and to assess pricing. The conjoint data are available on Canvas in the file conjoint.csv.

Analyzing the conjoint data using ex post segmentation.

1. Write down the regression model equation for your conjoint analysis and label everything clearly.
2. Using the data in conjoint.csv, estimate the relevant regression(s) and report results in a welllabeled table that includes parameter estimates, standard errors and t-statistics for each parameter. Use whichever software you see fit (e.g. R, Excel etc).
3. By cluster (i.e. segment), report in a well-labeled table the willingness-to-pay for each attribute. In words, provide a verbal profile of each cluster based on your calculations.

Demand Analysis:

4. Consider the current subscription plan offered by the firm and assume it is the only service offered: NBC, CBS, ESPN on the LG device. Plot the EVC demand curve for this subscription plan. Be sure to label your graph clearly.

1. Now suppose the firm also offers an alternative plan consisting of NBC, CBS, ABC, ESPN and STARZ for $50 per month. Plot the new EVC demand curve for the plan offered in 4(a).

Static Pricing Analysis:

5. Suppose the firm only offers its current plan, as in 4(a). What is the static, revenuemaximizing, monopoly price of the subscription for a month of service? Show your work.

6. Suppose the firm would be willing to unbundle the offering, allowing customers to customize part of their subscription. Under this plan, the firm sells the LG device bundled only with basic phone service. However, consumers could optionally choose to add the content bundle to the phone with basic service. The consumer cannot design the content bundle: the content bundle is set to consist of NBC, CBS and ESPN. An additional price would be charged for adding the content bundle to the monthly subscription. Naturally, a consumer would have to purchase the basic phone service in order to get access to the additional optional television content.

1. What are the optimal prices of the phone with basic service and the optional content bundle respectively? Show your work.
2. How do the revenues and prices from this scheme compare with the scheme in Q5? Does your answer make intuitive sense? Why or why not? 

Accounting for the Lifetime Value of the Consumer:

7. As before, suppose the firm only offers the subscription plan consisting of NBC, CBS and ESPN. Suppose also that it prices this plan at the rate you computed in question Using other existing field data on consumers who have already adopted the service, you know that an existing customer has a 70% chance of defecting in any given month when you charge the price computed in question 5. Assume that the size of your total customer market is 500,000 individuals and your current existing customer base is 150,000 individuals. From accounting, you also know that the cost of providing service to a customer (existing or prospective) is $8 per month^[1]. Moreover, the cost of contacting a prospective customer with an offer to sign up for the service is roughly $5. This "initial marketing contact" cost does not apply to existing customers. Finally, the firm's monthly borrowing cost of capital used to discount future revenues from a customer is 5%.

1. What fraction of the prospective consumer market would you predict will adopt the service at this price? This fraction constitutes the probability that a randomlyselected prospective customer would adopt at the given price level. If a consumer is indifferent between buying and not buying, assume that she buys.
2. Build a LTV model and calculate the net present value of a prospective and an existing customer respectively. Assume that the customers are homogeneous in their taste for the service and a prospective customer is contacted once per month. Show your work.
3. Recently, HBO has contacted your firm to negotiate a contract that would add HBO as part of the monthly subscription plan. The proposed contract would bill your firm an upfront fixed payment of $50 million irrespective of the number of subscribers and a monthly payment of $2 per subscriber. Suppose you would continue to charge the price calculated in question 5 even after adding HBO to the subscription. Would adding HBO to the subscription package be profitable in the long term? Show your work.