Dynamic Ad Bidding for Search Session Relevance Ranking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Marketplaces face challenges in displaying relevant products to users due to the large number of products and varying user search sessions, hindering effective advertising strategies.
Innovation Solution
A system and method for determining a dynamic bid for a ranking system that includes receiving user search queries and session information, calculating first and second bids based on campaign input, and combining these bids to generate a final bid for advertisement positioning, which is then used to reposition listings in a graphical user interface.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If participants bid on advertising space to display products, then advertising revenue is improved, but the ability to display relevant products to relevant users deteriorates due to the large number of products and constant variation of user search sessions
Solution Approach 1:
The system implements dynamic bid adjustment by continuously updating bid amounts based on real-time user session characteristics, product relevance scores, and campaign parameters. The bid amount is not static but adapts dynamically to match user intent and product suitability, resolving the contradiction between maximizing revenue and maintaining relevance.
Solution Approach 2:
The system changes multiple parameters simultaneously including bid amount, product ranking position, and display timing based on user session analysis. By adjusting these parameters dynamically according to user behavior patterns and product relevance, the system optimizes both advertising revenue and display relevance.
2Quantity of substance
If the system displays more products to handle the large number of products offered for sale, then product availability is improved, but the difficulty of finding desired products through search increases
Solution Approach 1:
The system applies different quality levels of product display to different users based on their session characteristics. Instead of uniformly displaying all products, it tailors the product set, ranking, and prominence based on user intent, making the large product catalog manageable and discoverable while maintaining comprehensive availability.
Solution Approach 2:
The system uses feedback from user search behavior, click patterns, and session characteristics to continuously refine product display decisions. This feedback loop enables the system to navigate the large product catalog efficiently by learning from user interactions and improving product discoverability over time.
Data Source
AI summary
Systems and methods including one or more processors and one or more non-transitory storage devices storing computing instructions configured to run on the one or more processors and perform: receiving a search query from a user and user session information; receiving input information corresponding to one or more campaigns to display to the user; determining a first bid for the one or more campaigns based on the search query and the user session information; determining a second bid for the one or more campaigns based on the input information; determining a final bid for the one or more campaigns based on the first bid and the second bid; transmitting the final bid to a ranking system to generate a listing of advertisements, the listing of advertisements including positioning information; and enabling a graphical user interface to be modified to reposition the listing of advertisements based on the positioning information. Other embodiments are disclosed herein.