Ad Delivery Indexing and Yield-Based Ranking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional ad delivery systems face issues with slow response times and pricing structures that discourage higher bids, leading to inefficient matching and limited ad placement effectiveness.
Innovation Solution
An improved ad delivery system that extracts an index file alphabetized by keyword, matches query strings to advertisements, ranks them based on estimated yield, and prices them lower than the bid price to maintain ranking, while filtering out ads with low click-through rates or below minimum thresholds, using incremental updates and multiple matching algorithms to optimize ad placement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If conventional ad delivery systems are used, then ad placement can be achieved, but response time is slow
Solution Approach 1:
The system pre-processes and indexes advertisements by keywords before queries arrive. The index file is built in advance with advertisements organized by keywords, allowing rapid retrieval during query processing without needing to search through the entire ad database from scratch.
Solution Approach 2:
The ad database is segmented into an index file structure that separates keyword information from full advertisement data. The index file contains only keyword-to-ad-reference mappings, while the full ad database stores complete ad details, enabling efficient segmented access during queries.
2Adaptability or versatility
If conventional pricing models are used, then ad placement can be secured, but higher bids are discouraged and matching is limited
Solution Approach 1:
The pricing model dynamically adjusts ad prices based on real-time factors including bid amounts, keyword competitiveness, and predicted click-through rates. The system can modify prices during the ad delivery process rather than using static pre-set prices, allowing flexible adaptation to changing market conditions.
Solution Approach 2:
The system incorporates feedback loops that monitor actual click-through rates and user engagement metrics, using this information to refine future pricing and matching decisions. Performance data from delivered ads feeds back into the system to optimize subsequent ad selection and pricing strategies.
3Quantity of substance
If all advertisements are indexed, then matching coverage is maximized, but processing time and system complexity increase
Solution Approach 1:
The system extracts only the necessary keyword information from the full ad database to create a separate index file. This index file contains minimal data (keyword-to-ad-reference mappings) rather than complete advertisement details, reducing the amount of data that needs to be processed and stored while maintaining full matching capability.
Solution Approach 2:
The index file acts as an intermediary structure between the query processing system and the full ad database. It serves as a lookup table that translates query keywords into relevant ad references, eliminating the need to scan through the entire ad database during queries.
Data Source
AI summary
Advertisements over a network, such as the Internet, by extracting an index file of suppliers of search result sets for a plurality of keywords and/or an index of ads to be placed. A keyword in a received query may be matched to the suppliers of related search result sets and/or the ads to be placed in the index file(s). The search results may be ranked based on various criteria such as yield, click through rates and history of dealings. Search result sets for the keywords from the matching suppliers and the ads to be placed may be combined to produce a search result set ranked in accordance with estimated yields for advertisements associated with each result in the search result set and delivered in response to the received query. The search results in the combined set may be reranked and/or repriced.