Main Ad Server Video Header Bidding Latency Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In online advertising, video header bidding is challenging for publishers due to communication difficulties between video players and ad servers, leading to delays and inefficiencies in ad delivery, as existing methods are unstable and resource-intensive, and often result in suboptimal ad selection.
Innovation Solution
A main ad server is introduced to simplify communication among network ad servers, collect campaign information, and prioritize demand tags based on publisher preferences, facilitating video header bidding by initiating an auction process among willing ad servers to reduce latency and resource consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If display header bidding methodologies are implemented in the video player, then video header bidding can be enabled, but the implementation becomes unstable and difficult to scale
Solution Approach 1:
The patent introduces a header bidding module as an intermediary component that sits between the video player and ad servers. This module standardizes communication protocols and abstracts the complexity of video header bidding, making the implementation stable and scalable across different video players and ad servers without requiring direct integration between them.
2Productivity
If the video player submits bids to ad servers, then ad revenue can be generated, but communication difficulties arise between the video player and ad servers
Solution Approach 1:
The patent segments the ad serving system into distinct functional modules: a header bidding module in the video player, intermediary bid servers, and ad servers. Each module handles specific tasks independently, simplifying communication by defining clear interfaces and responsibilities. The header bidding module manages bid submissions, the bid servers relay information, and ad servers process requests, reducing overall system complexity.
3Adaptability or versatility
If the video player restarts the bidding process to obtain another video, then another ad can be obtained, but time is consumed and additional processing resources are required
Solution Approach 1:
The patent implements preliminary actions by pre-fetching and caching bid information during video playback. The header bidding module maintains an active bidding state and pre-loads available ads, so when a video ends or needs replacement, the system can quickly retrieve pre-prepared bid results without restarting the entire bidding process, significantly reducing latency and resource consumption.
4Productivity
If multiple ad networks arbitrage ad space through sequential ad calls, then ad space can be resold, but significant delay occurs before the ad is available
Solution Approach 1:
The patent enables continuous ad delivery by implementing a waterfall mechanism that processes multiple ad network requests in parallel rather than sequentially. The system continuously maintains bid states and ad availability information, allowing ad space to be filled from multiple sources simultaneously without waiting for each ad call to complete, thereby eliminating delays while maintaining full ad space utilization.
Data Source
AI summary
Systems and methods are provided for supplying a video advertisement to a web page or application. A main ad server provides an ecosystem to network ad servers from different ad networks. The main ad server can collect and store information about each of the demand tags associated with each of the network ad servers in a database. The main ad server can then review the database in response to an ad call from a publisher to generate an ordered list of the demand tags that correspond to the ad call. The main ad server can then determine if any of the demand tags in the ordered list has a corresponding video header bidding tag and initiate an auction process for all of the demand tags in the ordered list that have a video header bidding tag. The ordered list of demand tags can then be updated based on the results of the auction process.