Ad Inventory Allocation with Uncertain Supply
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems for optimizing display advertising struggle to efficiently allocate advertisement impressions between guaranteed and non-guaranteed demand, often resulting in inefficiencies due to uncertain supply and conservative projections, which can lead to unsold or underutilized ad inventory.
Innovation Solution
A data processing system that uses an optimizer to generate an allocation plan for advertisement impressions, dividing them between guaranteed and non-guaranteed demand based on uncertain supply, employing stochastic programming with recourse to determine projected supply and adjust allocations dynamically, thereby optimizing the proportion of impressions committed to guaranteed contracts versus those sold on a spot market.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conservative projections are used to estimate supply, then the reliability of meeting guaranteed obligations is improved, but the productivity of ad inventory utilization deteriorates
Solution Approach 1:
The system performs preliminary actions by pre-committing to guaranteed demand allocations before actual supply is known, using stochastic programming to simulate multiple supply scenarios. This allows the system to prepare allocation plans in advance that account for uncertainty while maximizing expected inventory utilization.
Solution Approach 2:
The allocation plan is made dynamic by continuously adjusting the proportion of guaranteed versus non-guaranteed demand based on actual supply realizations. The system recalculates optimal allocations as supply information becomes available, transitioning from static conservative projections to adaptive response.
2Ease of operation
If a fixed allocation proportion is used between guaranteed and non-guaranteed demand, then the ease of operation is improved, but the adaptability to uncertain supply deteriorates
Solution Approach 1:
The system replaces fixed allocation proportions with dynamic allocation that adapts to actual supply conditions. The optimizer continuously adjusts the split between guaranteed and non-guaranteed demand based on realized supply, making the system versatile while maintaining operational simplicity through automated decision-making.
Solution Approach 2:
The system implements feedback mechanisms where actual supply information feeds back into the allocation optimization process. This closed-loop approach allows the system to learn from actual supply conditions and adjust future allocations, improving adaptability without complicating operation.
3Reliability
If all advertisement impressions are allocated to guaranteed demand, then the reliability of contract fulfillment is improved, but the loss of revenue from unsold inventory increases
Solution Approach 1:
The system changes the allocation parameter from 100% guaranteed demand to an optimized mix that includes non-guaranteed demand. By adjusting this parameter dynamically based on supply uncertainty and revenue considerations, the system balances contract fulfillment reliability with revenue maximization.
Solution Approach 2:
The system applies partial action by allocating only the optimal proportion of impressions to guaranteed demand rather than all available impressions. This partial allocation to guaranteed demand allows the remaining inventory to be sold on the spot market, reducing revenue loss from unsold inventory.
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, the optimizer to establish a relationship between the first portion of advertisement impressions and the second portion of advertisement impressions. The relationship defines a range of possible proportions of allocation of the first portion of advertisement impressions and the second portion of advertisement impressions. The optimizer generates a solution in accordance with an uncertain supply of advertisement impressions, in which a projected supply of advertisement impressions is determined with a model from the uncertain supply of advertisement impressions. The data processing system outputs the allocation plan including the solution to control serving of the advertisement impressions in determined proportions.