Control Account Identification via Prediction Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for measuring the impact of events on spend behavior in loyalty programs lack accuracy and efficiency in determining control accounts exposed to specific events, often relying on comparative analysis of transaction profiles which is resource-intensive and less precise.
Innovation Solution
A system and method that determines a control account corresponding to an exposed account by creating a combined account pool table, aggregating transaction data, and using prediction models to identify cohort groups and segment levels, thereby enhancing the accuracy and efficiency of identifying accounts affected by events.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional comparative analysis of transaction profiles is used to determine control accounts, then measurement of event impact can be performed, but accuracy and efficiency are reduced and network resource usage increases
Solution Approach 1:
The system performs preliminary actions by pre-establishing prediction models that incorporate multiple variables (account attributes, event characteristics, temporal patterns) before the actual measurement process. These pre-trained models enable rapid, accurate identification of control accounts without requiring resource-intensive real-time comparative analysis of transaction profiles, thus improving both accuracy and efficiency simultaneously
2Measurement precision
If traditional comparative analysis of transaction profiles is used to determine control accounts, then measurement of event impact can be performed, but network resource usage increases
Solution Approach 1:
The patent replaces the mechanically intensive process of real-time comparative analysis of transaction profiles with a computational model-based approach. Prediction models trained on historical data and account attributes perform the identification function through algorithmic processing rather than exhaustive computational comparison, significantly reducing network resource consumption while maintaining or improving measurement precision
Data Source
AI summary
Provided is a method including determining a combined plurality of accounts, determining, for each account of the combined plurality of accounts, aggregate transaction data associated with a plurality of transactions involving the account; determining a first cohort level group including a group of exposed accounts and controls accounts; determining a first segment level group of accounts from the first cohort level group of accounts; generating a prediction model based on a plurality of control accounts that are included in the first segment level group of accounts; identifying a first exposed account of a plurality of exposed accounts; determining a first control account of the plurality of control accounts that corresponds to a first exposed account using the prediction model; and outputting a report comprising data associated with the first control account that corresponds to the first exposed account. Systems and computer program products are also provided.