Reward Impression Content Slotting via ML Conversion Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Identifying the ideal content item to present to users in a reward impression is challenging, as existing systems typically only present one video at a time, limiting user engagement and requiring multiple reward impressions for each content item.
Innovation Solution
A computing system using machine-learned models to select and optimize multiple content items for different time slots within a reward impression, merging or truncating content to fit allocated time, and presenting end cards for user interaction, thereby improving user engagement and reducing the need for multiple impressions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If only one video is presented at a time in a reward impression, then the system complexity is reduced and ease of operation is improved, but user engagement is limited and productivity decreases
Solution Approach 1:
The reward impression is segmented into multiple time slots, with each slot dedicated to presenting a different content item. This segmentation allows the system to display multiple videos within a single reward impression structure, thereby increasing user engagement without significantly complicating the overall system operation.
Solution Approach 2:
The system dynamically selects and presents different content items in different time slots based on machine-learned models that predict conversion rates. This dynamic approach optimizes user engagement by adapting content presentation to user behavior patterns while maintaining manageable system complexity through automated decision-making.
2Productivity
If multiple content items are presented in a single reward impression, then user engagement is improved and productivity increases, but the device complexity and difficulty of detecting and measuring increase
Solution Approach 1:
The system employs machine-learned models that automatically select optimal content items for each time slot based on predicted conversion rates. This self-service approach allows the system to manage multiple content items dynamically without requiring complex manual configuration, thereby increasing user engagement while keeping device complexity manageable through automation.
Solution Approach 2:
The system changes parameters such as content item selection, time slot allocation, and presentation timing based on machine-learned predictions. By dynamically adjusting these parameters, the system can optimize user engagement across multiple content items without requiring proportional increases in device complexity, as the changes are driven by automated algorithms rather than manual system redesign.
3Loss of time
If multiple content items are presented in a single reward impression, then the number of reward impressions required is reduced and loss of time is decreased, but calculating and selecting optimal content items increases computational complexity
Solution Approach 1:
The system performs preliminary calculations of conversion rates for multiple content items using machine-learned models before the reward impression is displayed. By pre-computing these predictions, the system can quickly select optimal content items for each time slot, thereby reducing the overall time required for multiple impressions while managing computational complexity through advance preparation.
Solution Approach 2:
The system maintains continuous optimization by continuously updating machine-learned models with new data and continuously selecting optimal content items for presentation. This continuous useful action ensures that the system efficiently presents multiple content items in a single reward impression, reducing time loss while managing computational complexity through ongoing automated optimization rather than repeated manual interventions.
Data Source
AI summary
The present disclosure provides techniques for presenting a reward impression. A computing system can receive, from a client device, a request to view the reward impression having a first time slot and a second time slot. The computing system can calculate a first conversion rate associated with a first content item being presented in the first time slot and a second conversion rate associated with a second content item being presented in the second time slot. The computing system can select, using the one or more machine-learned models based on the first conversion rate and the second conversion rate, the first content item and the second content item from a plurality of content items. The computing system can cause the presentation of the first content item in the first time slot and the second content item in the second time slot of the reward impression.