Context-Aware Customer Feedback System Using Bayesian Question Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for collecting customer feedback are not contextually aware of the customer and transaction details, leading to generic questions that do not effectively engage customers, resulting in low response rates and inaccurate feedback.
Innovation Solution
A system that analyzes transaction data, including the customer's basket contents and anonymous customer information, to determine a single contextually aware survey question, prioritizing relevant questions using Bayesian probability, and dynamically adjusts question weightings to optimize response rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If generic feedback questions are used for all customers, then the system is simple to operate, but the response rate is low and feedback accuracy is poor
Solution Approach 1:
The system performs preliminary analysis of transaction data, basket contents, and customer information before generating feedback questions. This advance preparation enables the system to customize questions based on customer context without adding complexity during the actual feedback collection process
Solution Approach 2:
The patent introduces an intermediary processing layer that analyzes transaction data and customer information, then uses Bayesian probability to select appropriate feedback questions. This intermediary layer handles the complexity of data analysis while presenting only simple, relevant questions to customers
2Loss of information
If multiple detailed questions are asked to gather comprehensive feedback, then the information quality improves, but the time required for customers to respond increases
Solution Approach 1:
The system applies local quality by customizing feedback questions based on specific customer contexts such as transaction type, basket contents, and purchase history. Instead of asking all customers the same comprehensive set of questions, each customer receives a tailored subset of questions relevant to their specific experience
Solution Approach 2:
The patent dynamically changes the parameters of feedback collection by adjusting which questions are asked based on customer segmentation and transaction analysis. The Bayesian probability model continuously updates question selection parameters based on observed response patterns and contextual data
3Measurement precision
If feedback is collected immediately during or after purchase, then the feedback accuracy is higher, but customers may be rushed and less willing to engage
Solution Approach 1:
The system dynamically adjusts the feedback collection process based on customer behavior and transaction context. It can adapt the timing, format, and content of feedback requests to match the customer's current state, making the process more flexible and less intrusive while maintaining timing advantages
4Reliability
If personal details are required for feedback collection, then the feedback can be validated and linked to specific customers, but this discourages potential reviewers from providing feedback
Solution Approach 1:
The patent extracts and utilizes existing transaction data and anonymous customer information that is already available from the purchase process. Instead of requiring additional personal details, the system leverages data already collected during the transaction, such as purchase history and basket contents, to validate and contextualize feedback
Data Source
AI summary
There is provided a system operable to collect, from a customer, a response to a question relating to a transaction, the system comprising: a requesting device configured to send, to a questions server, data relating to the transaction and a request for a question; the questions server, comprising: a processor; one or more databases storing questions; and a rules engine, the questions server configured to: receive the data relating to the transaction and the request for a question from the requesting device; select a question from the stored questions based on the transaction data; and send the selected question to the requesting device; wherein the requesting device is further configured to: present the question to the customer or instruct a display device to present the question to a customer; wherein the system is further configured to: receive a response to the question; and send data related to the response to the questions server or a response database. Corresponding methods, questions server and requesting device are also provided.