Session-Based Purchase Intent Prediction for Anonymous Retail Users
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Websites struggle to accurately determine user satisfaction and purchase intent, particularly for anonymous users, which hinders effective marketing strategies and conversion optimization.
Innovation Solution
An experience analytics system utilizing machine learning models to analyze user behavior on retail websites, predicting purchase intent based on session data and identifying a missed purchase segment, even for anonymous users, by employing a 1-class classification model to recognize patterns in user interactions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional analytics methods are used to track user behavior, then user satisfaction and purchase intent can be monitored, but anonymous users cannot be accurately identified and analyzed
Solution Approach 1:
The patent introduces session data as an intermediary element that bridges the gap between anonymous user identification and purchase intent analysis. By using session-based tracking instead of user-based tracking, the system can analyze behavior patterns of anonymous users through their interaction sequences, page views, and conversion actions without requiring personal identification information.
Solution Approach 2:
The system changes the fundamental parameter of user identification from persistent user identifiers to transient session identifiers. This parameter change enables the analysis of anonymous users by focusing on behavioral patterns within session contexts rather than relying on user identity, thereby improving both measurement precision for purchase intent and adaptability for anonymous user analysis.
2Measurement precision
If comprehensive user data is collected to improve prediction accuracy, then purchase intent can be better predicted, but user privacy and data security concerns increase
Solution Approach 1:
The patent extracts only the essential behavioral data elements needed for purchase intent prediction while excluding personal identifying information. By taking out only the necessary session-based interaction data (page views, clicks, conversions) and leaving out user identity information, the system maintains prediction accuracy while reducing privacy risks.
Solution Approach 2:
The system uses transient session data that is automatically discarded after analysis rather than storing persistent user profiles. This approach uses 'cheap short-living objects' (temporary session records) to achieve prediction goals without creating long-term privacy vulnerabilities, as the data has inherent limited lifespan and scope.
3Productivity
If real-time analysis is implemented to enable timely promotional interventions, then conversion rates can be improved, but system complexity and computational resources increase
Solution Approach 1:
The system performs preliminary classification of session data into converted and non-converted categories during or immediately after user sessions. By conducting this classification action in advance, the system prepares prediction-ready data structures that can be quickly queried for real-time promotional interventions without requiring complex on-the-fly analysis during critical decision moments.
Solution Approach 2:
The patent segments user sessions into distinct converted and non-converted groups based on conversion status. This segmentation simplifies the analytical complexity by creating discrete, well-defined categories that can be analyzed with simpler models, reducing the overall system complexity while enabling targeted real-time interventions for non-converted segments.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The subject technology receives session data from a set of users associated with a retail website. The subject technology generates, using at least one machine learning model, a purchase intent prediction based at least in part on the session data. The subject technology determines a set of sessions that correspond to conversions on the retail website based on the session data. The subject technology generates a combination of data using first data from the purchase intent prediction with second data based on the set of sessions that correspond to conversions. The subject technology generates information related to a missed purchase segment based at least in part on the combination of data. The subject technology provides the information for display on a client device.