User Identification via Device Fingerprinting in Ad Bidding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing online advertising systems face challenges in accurately tracking and targeting unique users due to the limitations of cookies, which can result in multiple user IDs for the same individual, difficulty in tracking exposure to marketing campaigns, and loss of user data due to cookie expiration or deletion.
Innovation Solution
The system uses fingerprint context data, including IP address, browser plugins, system fonts, time zone, and network parameters, to detect unique users and assign them identifiers, even without traditional user IDs, and clusters users based on their attributes and event data to group similar users for targeted advertising.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional cookies are used for user tracking, then user identification can be implemented, but user data is lost due to cookie expiration or deletion and multiple user IDs may be assigned to the same individual
Solution Approach 1:
The patent creates a fingerprint copy of the user's device characteristics (browser version, installed fonts, screen resolution, operating system, etc.) as an alternative to traditional cookies. This fingerprint serves as a persistent identifier that doesn't expire or get deleted like cookies, thereby resolving the data loss issue while maintaining reliable user identification
Solution Approach 2:
The patent introduces a fingerprinting mechanism as an intermediary between the user's device and the advertising system. Instead of relying on mutable cookies, the system uses this intermediary fingerprint derived from device characteristics to maintain persistent and reliable user identification across sessions
2Measurement precision
If fingerprint context data is collected for user identification, then user tracking accuracy improves, but system complexity increases due to multiple data collection points
Solution Approach 1:
The patent makes the fingerprinting system universal by integrating it into the existing ad serving infrastructure. The same fingerprint data serves multiple purposes: user identification, device characterization, and tracking across different contexts. This multi-functionality reduces the need for separate tracking systems and minimizes overall system complexity
Solution Approach 2:
The system automatically collects fingerprint context data from the user's device without requiring manual intervention or complex configuration. The fingerprinting process is self-service in nature, gathering necessary information through standard web technologies and automatically processing it to generate persistent identifiers
3Productivity
If user data is collected for targeted advertising, then advertising effectiveness improves, but user privacy concerns increase
Solution Approach 1:
The patent uses device fingerprint data as a disposable alternative to traditional persistent cookies and personal identifiers. The fingerprint is derived from transient device characteristics rather than personal information, enabling effective advertising targeting without collecting sensitive user data that would raise privacy concerns
Solution Approach 2:
The patent converts potentially harmful detailed user tracking into beneficial anonymized fingerprint data. By using device characteristics rather than personal information, the system maintains advertising effectiveness while actually improving user privacy - the harm of detailed tracking is transformed into the benefit of anonymous, non-personal identification
Data Source
AI summary
Methods and apparatus for finding detecting unique on-line users for advertisement or content targeting are disclosed. A bid request with initial fingerprint context data for a user and for content that is requested by such user is received at a bid management system, and the bid request is received with or without a user identifier. At the bid management system, an optimum bid for the bid request is determined based on the initial fingerprint context data for the user and the content. The optimum bid with a reference to an advertisement and a script for obtaining additional fingerprint context data from a device of the user is sent from the bid management system. The additional fingerprint context data is received and a unique user identifier is assigned in association with a combination of the initial and additional fingerprint context data.


