Eric Hughes, advertising sales manager at Regional Broadcast Network (RBN), needs to avoid a takeover by increasing revenue from ad sales. Currently, ad plans are created for advertisers by combining ad spots from a fixed inventory of shows, making an effort to meet requirements such as a preferred split of prime/non-prime shows and views (impressions) in target demographics. Ad plans are priced using rate cards (RCs) based on industry norms, and are often discounted to meet budget requirements. Revenue is not usually optimized using this system because the RCs do not accurately reflect the value of inventory. In this case, Hughes first creates a model that allocates available inventory (i.e., 30-second ad spots) across 10 representative plans. He performs an optimization calculation that recreates the sequential allocation that his salespeople generate when advertisers approach the network one at a time. Next, he creates a model that optimizes the allocation of ad spots across all plans, assuming all customers request their plans at the same time. Hughes realizes that the bid prices revealed in the sensitivity table of the second model can also be interpreted as the opportunity cost associated with one incremental ad spot in each show. When the sequential allocation model replaces RCs with bid prices, which are dynamically adjusted for remaining inventory and expected future demand as each new customer arrives, the resulting revenue is closer to the value produced by optimizing all plans simultaneously. In the B case, "TV Advertising Pricing at Regional Broadcast Network (B)," Hughes uses the full historical sales dataset to conduct a multivariable regression analysis and better understand what drives the price of a plan. Students are challenged to create their own analysis and rationale, and to develop a guide for pricing each advertiser's plan. This case set presents emerging best practices in maximizing revenues in the ad industry.
Eric Hughes, advertising sales manager at Regional Broadcast Network (RBN), needs to avoid a takeover by increasing revenue from ad sales. Currently, ad plans are created for advertisers by combining ad spots from a fixed inventory of shows, making an effort to meet requirements such as a preferred split of prime/non-prime shows and views (impressions) in target demographics. Ad plans are priced using rate cards (RCs) based on industry norms, and are often discounted to meet budget requirements. Revenue is not usually optimized using this system because the RCs do not accurately reflect the value of inventory. In this case, which builds on "TV Advertising Pricing at Regional Broadcast Network (A)," Hughes uses the full historical sales dataset to conduct a multivariable regression analysis and better understand what drives the price of a plan. Students are challenged to create their own analysis and rationale, and to develop a guide for pricing each advertiser's plan. This case set presents emerging best practices in maximizing revenues in the ad industry. Students are given supplementary Excel workbooks containing sample data and use the Solver module and regression analysis to complete the assignment. This case set is suitable for use when teaching pricing analytics, decision analysis, statistics, quantitative analysis, or operations management. It may be used with a graduate, undergraduate, or executive education audience.
This technical note examines two common methods for estimating customer demand using historical data observations that are constrained by availability. First, this note explains how to leverage the averaging method for unconstraining the demand of airline ticket bookings. Next, the expectation-maximization algorithm is introduced using historical product sales data. Both examples rely on fictious data, and an accompanying Excel workbook provides example calculations for both methods highlighted in this technical note.