In one embodiment, a promotional subsystem generates promotional plans that include promotionals, where each promotional targets one or more pieces of scheduled content. First, the promotional subsystem generates a statistical model based on historical respondent viewership data. The statistical model model maps a respondent viewing of a promotional that targets a piece of scheduled content to a probability of the respondent viewing the piece of scheduled content. Subsequently, the promotional subsystem generates a proposed promotional plan based on the statistical model, a schedule that includes the piece of scheduled content, and a risk tolerance. Advantageously, the promotional subsystem may be configured to generate different proposed promotional plans based on different risk tolerances associated with different viewership growth strategies. Automatically generating proposed promotional plans based on probabilities reduces the time required to identify an acceptable promotional plan compared to current techniques that generate a single promotional plan based on deterministic strategies.
BACKGROUND OF THE INVENTIONField of the Invention
Embodiments of the present invention relate generally to computer processing and, more specifically, to techniques for generating promotional plans to increase viewership.Description of the Related Art
Many media publishers (e.g., broadcast or cable networks) generate promotional plans that use some of their own non-programming time to promote their own pieces of scheduled content, such as episodes of television shows, movies, etc. These types of promotional plans can include any number of creatives, where each creative includes material designed to advertise one or more pieces of scheduled content. Examples of creatives include television commercials, dynamic graphics that scroll across a portion of a display screen, and static graphics that occupy a portion of the display screen, to name a few. Because manually generating promotional plans may be prohibitively time consuming, media publishers often use promotional scheduling applications that implement semi-automated flows for generating promotional plans.
One drawback to using promotional scheduling applications is that these applications typically implement deterministic algorithms. As is well-known, deterministic algorithms compute a single output based on input data that is assumed to accurately and comprehensively specify any possible conditions that influence the output. Because effective business strategies and associated input data for generating pr
1. A computer-implemented method for generating promotional plans for different pieces of scheduled content, the method comprising: generating a statistical model based on historical respondent viewership data, wherein the statistical model maps a respondent viewing of a first promotional that targets a first piece of scheduled content to a probability of the respondent viewing the first piece of scheduled content; performing, via at least one processor, a first set of optimization operations on a schedule that includes the first piece of scheduled content based on the statistical model and a first risk tolerance to generate a first proposed promotional plan that includes a first scheduled promotional, wherein the first scheduled promotional is associated with the first promotional and a first point in time that lies between a start point of time associated with a second piece of scheduled content and an end point of time associated with the second piece of scheduled content; and displaying or transmitting for further processing the first proposed promotional plan to specify that the first promotional is scheduled at the first point in time.
11. A computer-readable storage medium including instructions that, when executed by a processor, configure the processor to perform the steps of: generating a statistical model based on historical respondent viewership data, wherein the statistical model maps a respondent viewing of a first promotional that targets a first piece of scheduled content to a probability of the respondent viewing the first piece of scheduled content; performing a first set of optimization operations on a schedule that includes the first piece of scheduled content based on the statistical model and a first risk tolerance to generate a first proposed promotional plan that includes a first scheduled promotional, wherein the first scheduled promotional is associated with the first promotional and a first point in time that lies between a start point of time associated with a second piece of scheduled content and an end point of time associated with the second piece of scheduled content; and displaying or transmitting for further processing the first proposed promotional plan to specify that the first promotional is scheduled at the first point in time.
20. A system comprising: a memory storing a promotional application; and a processor that is coupled to the memory, wherein, when executed by the processor, the promotional application configures the processor to: generate a statistical model based on historical respondent viewership data, wherein the statistical model maps a respondent viewing of a first promotional that targets a first piece of scheduled content to a probability of the respondent viewing the first piece of scheduled content; perform a first set of optimization operations on a schedule that includes the first piece of scheduled content based on the statistical model and a first risk tolerance to generate a first proposed promotional plan that includes a first scheduled promotional, wherein the first scheduled promotional is associated with the first promotional and a first point in time that lies between a start point of time associated with a second piece of scheduled content and an end point of time associated with the second piece of scheduled content; and display or transmit for further processing the first proposed promotional plan to specify that the first promotional is scheduled at the first point in time.