Ad Inventory Allocation Optimizer for Guaranteed and Non-Guaranteed Demand
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing advertising systems face challenges in optimizing the allocation of advertisement impressions between guaranteed and non-guaranteed demand, leading to inefficiencies in revenue generation and under-delivery penalties, as they struggle to balance representativeness and revenue maximization.
Innovation Solution
A data processing system that uses an optimizer to generate an allocation plan for advertisement impressions, allocating a portion to satisfy guaranteed demand and another portion to maximize non-guaranteed revenue while minimizing under-delivery penalties, by establishing a relationship between these proportions and controlling their serving proportions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If advertisement impressions are allocated to satisfy guaranteed demand, then demand fairness and representativeness are improved, but non-guaranteed revenue is reduced
Solution Approach 1:
The patent segments advertisement impressions into two distinct portions: a first portion allocated to satisfy guaranteed demand and a second portion allocated to satisfy non-guaranteed demand. This segmentation allows the system to independently optimize each portion for its specific objectives, resolving the contradiction between ensuring guaranteed demand fairness and maximizing non-guaranteed revenue.
Solution Approach 2:
The patent introduces a dynamic optimization mechanism where the proportions of the first and second portions are not fixed but are determined through an optimization process that maximizes guaranteed demand fairness, non-guaranteed revenue, and minimizes under-delivery penalties. This dynamic adjustment allows the system to adapt to changing conditions and resolve the contradiction between guaranteed demand satisfaction and revenue maximization.
2Productivity
If advertisement impressions are allocated to maximize non-guaranteed revenue, then revenue is improved, but under-delivery penalties increase
Solution Approach 1:
The patent separates the allocation of advertisement impressions into guaranteed and non-guaranteed portions, allowing the system to optimize each portion independently. This segmentation prevents the system from prioritizing non-guaranteed revenue at the expense of guaranteed demand, thereby reducing under-delivery penalties while maintaining revenue optimization.
Solution Approach 2:
The patent incorporates under-delivery penalties as a feedback mechanism in the optimization process. The optimizer considers the potential under-delivery penalties when determining the optimal proportions of guaranteed and non-guaranteed allocations, creating a feedback loop that adjusts the allocation to minimize penalties while maximizing revenue.
3Ease of operation
If the allocation plan is simplified, then ease of operation is improved, but optimization precision is reduced
Solution Approach 1:
The patent divides the complex allocation problem into two simpler sub-problems: allocating the first portion to satisfy guaranteed demand and allocating the second portion to satisfy non-guaranteed demand. This segmentation maintains optimization precision for each portion while simplifying the overall allocation plan structure, making it easier to operate and manage.
Data Source
AI summary
An advertisement impression distribution system includes a data processing system operable to generate an allocation plan for serving advertisement impressions. The allocation plan allocates a first portion of advertisement impressions to satisfy guaranteed demand and a second portion of advertisement impressions to satisfy non-guaranteed demand. The data processing system includes an optimizer to establish a relationship between the first portion and the second portion. The relationship defines a range of possible proportions of allocation of the first portion and the second portion. The optimizer generates indicia in accordance with maximizing guaranteed demand fairness or representativeness, maximizing non-guaranteed revenue, and minimizing under-delivery penalties, where the indicia indentifies a determined proportion of the first portion to serve and a determined proportion of the second portion to serve. The data processing system outputs the allocation plan including the indicia to control serving of the advertisement impressions in the determined proportions.