Video Engagement Scoring via Length-Normalized Completion Metrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video engagement metrics, such as completion rate, are misleading due to variations in video length, making it unfair to compare the engagement of videos of different durations, and there is a need for a metric that accounts for these differences.
Innovation Solution
A content score system that generates scores indicative of video engagement, using viewing factors like long and short attention span viewers, affinity, and replay factors, which consider the percentage of viewers watching a video to a predetermined duration, adjusted based on video length and audience demographics, to provide a fair comparison across videos of varying lengths.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If completion rate is used as the engagement metric, then the metric is simple to calculate, but the metric becomes misleading when comparing videos of different lengths
Solution Approach 1:
The patent transforms the engagement metric from a simple completion rate (binary: watched/not watched) to a weighted engagement score that incorporates multiple parameters including watch time duration, video length normalization, and engagement weighting factors. This allows the system to maintain computational simplicity while achieving accurate cross-video-length comparisons by changing the parameters used in the metric calculation.
2Measurement precision
If a standardized engagement metric is developed to account for video length differences, then the measurement accuracy improves, but the metric calculation complexity increases
Solution Approach 1:
The engagement metric is segmented into distinct computational components: (1) base engagement calculation from viewing data, (2) video length normalization factor, (3) engagement weighting adjustments. Each segment is calculated separately and then combined, making the overall complex metric manageable and computationally efficient while maintaining high measurement accuracy.
Solution Approach 2:
The system calculates engagement metrics with varying degrees of detail based on needs - using full weighted calculations when precision is required and simplified versions when speed is prioritized. This partial action approach allows the system to balance accuracy and complexity dynamically without requiring the full complex calculation in all scenarios.
3Reliability
If multiple viewing factors are considered to assess engagement accurately, then the engagement assessment becomes more comprehensive, but the data processing complexity increases
Solution Approach 1:
Multiple viewing factors (watch time, completion status, replay behavior, pause patterns) are merged into a single unified engagement score through weighted aggregation. This combining approach maintains comprehensive assessment reliability while simplifying data processing by consolidating multiple data streams into one integrated metric rather than requiring separate analysis of each factor.
Data Source
AI summary
Systems and methods for generating a content score for a video that is indicative of how engaging the video is to viewers makes use of a modified completion rate factor. The modified completion rate factor indicates the percentage of viewers that played the video and that watched at least a predetermined number of second of the video. Other viewing factors may also be used to help generate the content score of the video.