Predictive User Segmentation for Digital Advertising
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Brands and advertisers face challenges in effectively targeting potential new or reengaged customers due to difficulties in identifying the most likely candidates for their ad campaigns, leading to inefficiencies in reaching users who are likely to react positively to their advertisements.
Innovation Solution
The implementation of predictive user segmentation modeling and browsing interaction analysis, which involves generating customer scores in real-time based on browsing behaviors and historical data to identify users likely to belong to specific consumer segments, allowing for targeted advertising to users who are more likely to engage with the brand.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional ad campaign targeting methods are used, then ad campaigns can be launched, but they fail to reach the most likely potential customers due to inability to identify high-probability users
Solution Approach 1:
The system performs preliminary actions by proactively identifying and scoring potential customers before ad campaigns are launched. Customer service representatives use the likelihood scores to pre-identify high-probability targets, ensuring that when ad campaigns run, they immediately reach the most receptive audiences without wasting resources on low-probability users.
Solution Approach 2:
The patent replaces traditional mechanical/manual methods of identifying potential customers (such as manual market research, surveys, and guesswork) with an automated electronic system that uses machine learning models and browsing interaction data to automatically calculate and update customer likelihood scores in real-time.
2Quantity of substance
If ad campaigns target broad audiences, then more users are reached, but the effectiveness decreases because resources are wasted on users unlikely to convert
Solution Approach 1:
The system applies local quality by treating each customer individually with a customized likelihood score based on their specific browsing interactions, demographic data, and engagement history. Instead of applying a uniform targeting approach to all users, the system assigns different probability scores to different customers, allowing advertisers to focus resources locally on high-scoring individuals rather than spreading efforts thin across a broad audience.
Solution Approach 2:
The system dynamically changes the targeting parameter from static demographic categories to dynamic likelihood scores that are continuously updated based on real-time browsing interactions. This parameter transformation allows the system to identify high-probability customers across diverse demographic groups, maintaining quantity of reach while improving conversion productivity through data-driven segmentation.
3Measurement precision
If real-time customer score generation is implemented, then targeted advertising to high-probability users is achieved, but system complexity increases
Solution Approach 1:
The system segments the complex task of customer scoring into distinct functional modules: a machine learning model generation component that creates scoring algorithms, a browsing interaction data collection component that captures user behaviors, and a score calculation component that combines these elements to generate likelihood scores. This segmentation allows each module to be developed and maintained independently, reducing overall system complexity while maintaining measurement precision.
Solution Approach 2:
The patent introduces intermediary components that simplify the interaction between complex systems. A scoring service acts as an intermediary layer between the machine learning models and the advertising platform, translating complex model outputs into simple likelihood scores that can be easily used for targeting decisions. This intermediary abstraction shields users from the underlying complexity while preserving measurement precision.
Data Source
AI summary
Systems, methods, and computer-readable media are disclosed for predictive user segmentation modeling and browsing interaction analysis for digital advertising. In one embodiment, an example method may include identifying a target consumer segment, generating a predictive user behavior model for the target consumer segment based at least in part on selected user data, and receiving an indication that a user is browsing a website. Example methods may include generating a first customer score for the user using the predictive user behavior model in real-time, and determining that the first customer score meets a consumer segment modification threshold for the target consumer segment.