Behavior Pattern Search System for KPI Improvement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional techniques fail to effectively discover changes in behavior patterns among customers with specific characteristics that could improve Key Performance Indicators (KPIs) by analyzing purchase behaviors.
Innovation Solution
A behavior pattern search system and method that classifies customers based on predicted characteristic purchase behaviors using a state transition model created from purchase history and attribute information, identifying state transition paths that cross predetermined state groups, thereby uncovering changes in behavior patterns contributing to KPI improvements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If purchase behavior of each customer is analyzed using conventional techniques, then individual customer purchase probability can be calculated, but it becomes difficult to discover changes in behavior patterns that contribute to improving KPIs among customers with specific characteristics
Solution Approach 1:
The patent segments customers into distinct groups based on their behavior patterns and characteristics. By dividing the customer base into segments with specific characteristics (e.g., frequent purchasers, seasonal buyers, new customers), the system can analyze behavior changes within each segment rather than treating all customers uniformly, thereby improving the ability to detect KPI-improving pattern changes.
Solution Approach 2:
The patent introduces behavior pattern models as intermediary representations that capture the essential characteristics and transition patterns of customer groups. These models serve as mediators between raw purchase data and KPI analysis, enabling the system to detect and measure behavior changes that lead to KPI improvements without analyzing every individual customer transaction in detail.
2Adaptability or versatility
If detailed analysis of each customer's purchase behavior is performed, then individual customization is enabled, but the complexity of the analysis system increases significantly
Solution Approach 1:
The patent merges individual customer behavior data with group-level patterns to create comprehensive customer profiles. By combining individual purchase history with aggregated behavior patterns from similar customers, the system achieves adaptability for individual customization while avoiding the complexity of analyzing every customer in isolation. The merged approach leverages both individual and collective insights efficiently.
Solution Approach 2:
The patent transforms detailed individual customer data into simplified behavioral parameters and state transition models. By changing the representation from raw transactional data to abstracted behavior states and transition probabilities, the system maintains adaptability for personalized marketing while significantly reducing computational complexity. The parameter transformation enables efficient analysis of behavior patterns without processing every individual customer detail.
Data Source
AI summary
A behavior pattern search system and method is disclosed. A change in a behavior pattern contributing to improving a Key Performance Indicator (KPI) is discovered from among changes in the behavior pattern arising in common among customers having a specific characteristic. The behavior pattern search system classifies transfiguration candidates that are customers satisfying transfiguration result information. The transfiguration result information indicates a characteristic purchase behavior predicted to occur by a change in a behavior pattern of each of the customers.