Response Attribution Valuation Using Linear Regression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current marketing techniques struggle to accurately attribute credit to multiple promotions across marketing campaigns for customer responses, especially in cases where direct response attribution is not feasible, and existing methods fail to provide data-driven fractional credit assignment.
Innovation Solution
A computer-implemented method using a linear-weighted regression model to calculate response attribution values by retrieving and iteratively updating response rate and power law parameters, based on time delay distributions, to determine the effectiveness of each promotion in eliciting a response.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If direct response attribution is used to determine promotion effectiveness, then each response can be easily associated with identifying information about the promotion, but it fails to provide fractional credit assignment when multiple promotions contribute to a response
Solution Approach 1:
The patent segments the credit assignment by dividing the total response credit into fractional portions allocated to multiple promotions based on their respective contributions. Instead of assigning full credit to a single promotion, the system calculates partial credit values for each promotion that contributed to the response, enabling accurate multi-promotion attribution.
Solution Approach 2:
The patent changes the parameter of credit assignment from binary (full or no credit) to continuous fractional values. By introducing response attribution values that can take any value between 0 and 1, the system enables nuanced credit distribution across multiple promotions, transforming the attribution model from discrete to continuous.
2Measurement precision
If iterative parameter updating is performed to improve response attribution accuracy, then the model convergence improves measurement precision, but the computing resources and time required increase
Solution Approach 1:
The patent implements feedback by using the generated response attribution values to update the power law parameters in subsequent iterations. The updated parameters are then used to recalculate response attribution values, creating a closed-loop system that progressively refines the estimates until convergence is achieved or maximum iterations are reached.
Solution Approach 2:
The patent introduces dynamics by making the power law parameters adaptive rather than static. The parameters are continuously updated based on the response attribution values from previous iterations, allowing the model to dynamically adjust to better fit the observed data patterns and improve attribution accuracy over time.
Data Source
AI summary
A computer-implemented method is described for determining a response attribution value that represents a credit assignment to a communication, for an associated response received from a recipient during a marketing campaign.