Predictive User Segmentation Model for Targeted Advertising
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing online advertising technologies face challenges in targeting specific user segments effectively due to the small size of qualified user populations and the inability to immediately identify users likely to convert, as existing methods struggle with predicting future user behavior.
Innovation Solution
A computer-implemented method and system for generating expanded user segments by receiving and analyzing online activity data, building models to predict user associations with target features, and scoring users based on their online behavior to create larger, similar user segments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If advertisers target specific user segments with precise characteristics, then advertising relevance and conversion quality improve, but the size of the targetable audience decreases
Solution Approach 1:
The patent applies segmentation by dividing the user population into distinct segments based on predicted characteristics. The system segments users into those likely to have target features (e.g., high-value customers, engaged users) and those who are not, enabling precise targeting while maintaining adequate segment sizes through predictive modeling rather than direct observation
Solution Approach 2:
The patent implements preliminary action by predicting user characteristics before they manifest or before advertisers need the data. The system proactively identifies users likely to have target features using machine learning models trained on historical data, allowing advertisers to target users in advance rather than waiting for explicit confirmation of their characteristics
2Measurement precision
If advertisers wait for users to explicitly demonstrate target characteristics, then measurement accuracy improves, but the time available for advertising campaigns decreases
Solution Approach 1:
The system performs preliminary identification of users with target characteristics using predictive models before the characteristics are explicitly demonstrated. By training on historical data and predicting future behavior patterns, the system can identify likely candidates in advance, giving advertisers immediate actionable data rather than requiring users to first exhibit the characteristics
Solution Approach 2:
The patent employs feedback mechanisms where the system continuously learns from user behavior data and refines its predictions. Historical data about user actions and outcomes is fed back into the modeling system to improve accuracy over time, allowing the system to make more precise predictions about future user characteristics without requiring explicit demonstration
3Quantity of substance
If advertisers use existing user data to create segments, then segment availability improves, but the ability to predict future user behavior decreases
Solution Approach 1:
The system performs preliminary predictive analysis using machine learning models that are trained on historical data but designed to predict future behavior. Rather than simply segmenting existing data, the models proactively identify users likely to exhibit target characteristics in the future, combining the availability of segment data with predictive capability through sophisticated algorithms
Solution Approach 2:
The patent applies parameter changes by transforming static user data into dynamic predictions. The system changes the nature of the data from recorded historical facts to predicted future probabilities, using machine learning models that process existing parameters to generate new predictive parameters about future user behavior and characteristics
Data Source
AI summary
Systems and methods are disclosed for generating an expanded user segment based on a target segment of users associated with a specified target feature. In one implementation, a method is provided that includes receiving information about online activity by a set of users, the information including a unique user ID associated with each of the set of users; extracting from the received information, data about the set of users who are associated with the specified target feature; specifying one or more independent features relevant to a user association with the specified target feature; building a model that represents the probability of any user being associated with the specified target feature, based on the extracted data and the independent features; and using the model to score a network population of users relative to the specified feature, based on information received about online activity by the network population of users.