A Bayesian choice model for home video entertainment digital demand

Home video electronic sell-through (EST) demand growth is slow and shows signs of a tipping point in 2019. Point-of-sale time series data shows that consumers are sensitive to high-definition (HD) EST price, and unit sales spike when price drops.

Violeta Hennessey shares a proactive approach to enhance EST growth through price—a Bayesian choice model that quantifies the unit sales lift for a $15 price point versus a $20 price point typically used by the Warner Bros. Marketing Group. The solution enables Warner Bros. to simultaneously model HD EST, standard definition (SD) EST, Blu-ray, and DVD sales—the totality of consumer choices for the same movie—to examine the impact of a lower HD EST price on the other formats.

Violeta compares this model with the classic discrete choice models in terms of quantifying unit sales lift. The classic approach models the shifting of consumer choice based on product characteristics. In other words, it is modeling the market share of the formats. The limitation of this approach is that it assumes equilibrium on the total units sold. The ability to influence nonbuying consumers to buy at the lower price is ignored. The proposed model does not have this limitation, and the Bayesian framework can easily handle the complexity of the joint model.

Violeta illustrates the application of the model with an example using Fox Studios’s Deadpool, since Fox Studios tends to use a $15 price point for HD EST versus the $20 price point used by Disney and Warner Bros. Violeta uses the Bayesian choice model to quantify the lift Fox Studios experienced by using the lower price point, which resulted in a unit sales lift of 656,504 units for HD EST.