Proactive Hyperlink Bidding for Online Ad Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current real-time bidding methods in online advertising often delay webpage loading due to the time-consuming bidding process, which reduces computing and network resources and can result in improper bidding decisions due to insufficient analysis of user information.
Innovation Solution
An advanced real-time bidding system that proactively initiates bidding before a user requests to load a webpage by determining the likelihood of user interaction with hyperlinks, shifting resource utilization from client-side to server-side and allowing for more in-depth analysis of user data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If real-time bidding is performed after user requests webpage loading, then bidding decisions can be made based on user information, but webpage loading is delayed due to time-consuming bidding process
Solution Approach 1:
The system performs bid requests in advance by identifying hyperlinks on the current webpage and proactively initiating bid requests for the webpages those hyperlinks point to, before the user actually navigates to those pages. This preliminary action allows bidding decisions to be prepared beforehand, reducing the time delay when the user actually clicks and loads the target webpage.
2Adaptability or versatility
If bid requests are initiated for all hyperlinks, then comprehensive advertising coverage is achieved, but computing and network resources are wasted on low-probability links
Solution Approach 1:
The system applies different treatment to different hyperlinks based on their individual characteristics. It calculates a probability score for each hyperlink representing the likelihood of user selection, and only initiates bid requests for hyperlinks whose probability exceeds a predetermined threshold. This selective approach ensures computing and network resources are focused on high-value opportunities rather than being wasted on low-probability links.
Solution Approach 2:
The system changes the parameter of bid request initiation from a binary all-or-nothing approach to a probability-based selective approach. By introducing a probability threshold parameter, the system dynamically determines which hyperlinks warrant bid requests, optimizing the balance between advertising coverage and resource consumption.
3Speed
If bidding process is accelerated to reduce loading delays, then webpage loading speed improves, but bidding accuracy decreases due to insufficient analysis time
Solution Approach 1:
The system performs the time-consuming bid request and analysis process in advance, before the user actually navigates to the target webpage. This allows sufficient time for thorough analysis of user information and bidding decisions to be made with high precision, while the actual webpage loading occurs quickly since the bidding is already complete.
Solution Approach 2:
The system shifts the bidding process from the time dimension of webpage loading to a different temporal dimension - performing bids in advance during the user's current webpage viewing time. This dimensional shift allows bidding analysis to occur without interfering with the speed of actual webpage loading when the user clicks a hyperlink.
Data Source
AI summary
A computer-implemented method includes identifying a plurality of hyperlinks on a first webpage. The computer-implemented method further includes determining, for each hyperlink in the plurality of hyperlinks, a degree of likelihood that a user will select a given hyperlink in the plurality of hyperlinks. The computer-implemented method further includes initiating a first bid request for a first ad space for a second webpage that is accessible by a first hyperlink in the plurality of hyperlinks based on the first hyperlink having a degree of likelihood above a given threshold level. The computer-implemented method further includes determining a selection of the first hyperlink by the user. The computer-implemented method further includes displaying a first advertisement associated with a first winning bid on the second webpage. A corresponding computer system and computer program product are also disclosed.