Proactive customer churn prevention system with emotion-based risk scoring

Abstract

The present invention relates to a system (1) which enables customer interactions in the form of support requests, emails, social media posts, phone calls and chatbot correspondence to be analyzed in real time by using large language models (LLM); signals of negative emotion or disappointment to be detected; a dynamic customer churn risk score to be generated by evaluating customer behavior data in the form of reduced usage, fewer logins, late payments, cancellations of automatic bill payments, high withdrawals / transfers and security transfers; and actions towards regaining customers to be taken by triggering a personalized and proactive feedback mechanism as a result of integrated analysis.

Description

[0001] DESCRIPTION

[0002] PROACTIVE CUSTOMER CHURN PREVENTION SYSTEM WITH EMOTION-BASED RISK SCORING

[0003] Technical Field

[0004] The present invention relates to a system which enables customer interactions in the form of support requests, emails, social media posts, phone calls and chatbot correspondence to be analyzed in real time by using large language models (LLM); signals of negative emotion or disappointment to be detected; a dynamic customer churn risk score to be generated by evaluating customer behavior data in the form of reduced usage, fewer logins, late payments, cancellations of automatic bill payments, high withdrawal s / transfers and security transfers; and actions towards regaining customers to be taken by triggering a personalized and proactive feedback mechanism as a result of integrated analysis.

[0005] Background of the Invention

[0006] Current customer churn prevention methods are often limited to estimation models based on historical data and manual interventions. For example, customer service teams offer promotions or support by manually identifying customers who have been inactive for a certain period of time. However, this approach has a delayed and limited impact as it fails to monitor customer behavior in real time. Customer satisfaction surveys and feedback mechanisms are also mostly in static form and cannot provide real-time experience evaluation. Furthermore, the lack of advanced methods such as emotion analysis and dynamic risk scoring makes it difficult for early identification of customers when they are at risk of loss, which hinders timely and effective intervention.For this reason, it is understood that there is a need for a system which enables customer interactions in the form of support requests, emails, social media posts, phone calls and chatbot correspondence to be analyzed in real time by using large language models (LLM); signals of negative emotion or disappointment to be detected; a dynamic customer churn risk score to be generated by evaluating customer behavior data in the form of reduced usage, fewer logins, late payments, cancellations of automatic bill payments, high withdrawal s / transfers and security transfers; and actions towards regaining customers to be taken by triggering a personalized and proactive feedback mechanism as a result of integrated analysis.

[0007] The United States patent document no. US2023230100, an application included in the state of the art, discloses a system and method of predicting and preventing customer chum in service businesses. In the said invention, a system and method for minimizing customer chum in device service businesses commences with the execution of a customer service contract. Ongoing customer data capture is made for each contract. Customer data includes contract events, environmental events, service events, device usage analytics and personnel events. Machine learning is applied to customer data collected based on the state of customer data at the time of contract determination. Customer data is assigned weights, and aggregate data for each customer is compared to a preselected threshold level. Customers above the threshold level are considered happy, while customers below it are considered at risk. Remedial measures related to at-risk customer data generate automatic remediation levels with recommendations to an administrator when not sufficiently successful.

[0008] Summary of the Invention

[0009] An object of the present invention is to realize a system developed with the aim of enabling customer interactions in the form of support requests, emails, social media posts, phone calls and chatbot correspondence to be analyzed in real time by using large language models (LLM); signals of negative emotion ordisappointment to be detected; a dynamic customer churn risk score to be generated by evaluating customer behavior data in the form of reduced usage, fewer logins, late payments, cancellations of automatic bill payments, high withdrawal s / transfers and security transfers; and actions towards regaining customers to be taken by triggering a personalized and proactive feedback mechanism as a result of integrated analysis.

[0010] Detailed Description of the Invention

[0011] “Proactive Customer Churn Prevention System with Emotion-Based Risk Scoring” realized to fulfil the objective of the present invention is shown in the figure attached, in which:

[0012] Figure l is a schematic view of the inventive system.

[0013] The components illustrated in the figure are individually numbered, where the numbers refer to the following:

[0014] 1. System

[0015] 2. Database

[0016] 3. Server

[0017] The inventive system (1) developed with the aim of enabling customer interactions in the form of support requests, emails, social media posts, phone calls and chatbot correspondence to be analyzed in real time by using large language models (LLM); signals of negative emotion or disappointment to be detected; a dynamic customer churn risk score to be generated by evaluating customer behavior data in the form of reduced usage, fewer logins, late payments, cancellations of automatic bill payments, high withdrawal s / transfers and security transfers; and actions towards regaining customers to be taken by triggering apersonalized and proactive feedback mechanism as a result of integrated analysis comprises

[0018] at least one database (2) which is configured to keep a record of customer data;

[0019] at least one server (3) which is configured to collect customer interactions, to analyze banking transactions of customers, to perform customer emotion and behavior analysis by using large language models and / or machine learning algorithms, to generate a dynamic customer chum risk score and to take actions customized for customers above a certain risk score.

[0020] The database (2) included in the inventive system (1) is configured to establish communication and perform data exchange with the server (3) by using any communication protocol. The database (2) is configured to be managed by the server (3).

[0021] The server (3) included in the inventive system (1) is configured to establish communication and perform data exchange with the database (2) by using any communication protocol. The server (3) is configured to manage the database (2). The server (3) is configured to operate in integration with customer interaction agents and different interaction centers in the form of social media, phone calls, chatbot correspondence, surveys and communication channels of the respective bank, to enable customer interactions to be collected and the collected data to be recorded in the database (2). The server (3) is configured to enable customer behavior analysis to be performed by using variables in the form of frequency of use of the communication channels of the respective bank, number of logins, payment irregularities, automatic bill cancellations, asset drops, large withdrawal s / transfers and securities transfers of the customer. The server (3) is configured to enable customized feedback to be generated for customers with high customer churn risk scores and to be automatically transmitted to the customer through communication channels. The server (3) is configured to enable actionsconsisting of processes in the form of re-contacting the customer, apology email, loyalty reward, discount, advantageous prices to be performed for customers with high risk scores.

[0022] Industrial Application of the Invention

[0023] By means of the inventive system (1), it is enabled to analyze customer interactions in the form of support requests, emails, social media posts, phone calls and chatbot correspondence in real time by using large language models (LLM); to detect signals of negative emotion or disappointment; to generate a dynamic customer churn risk score by evaluating customer behavior data in the form of reduced usage, fewer logins, late payments, cancellations of automatic bill payments, high withdrawals / transfers and security transfers; and to take actions towards regaining customers by triggering a personalized and proactive feedback mechanism as a result of integrated analysis.

[0024] Within these basic concepts; it is possible to develop various embodiments of the inventive “Proactive Customer Churn Prevention System (1) with Emotion-Based Risk Scoring”; the invention cannot be limited to examples disclosed herein and it is essentially according to claims.

Claims

CLAIMS1. A system (1) which enables customer interactions in the form of support requests, emails, social media posts, phone calls and chatbot correspondence to be analyzed in real time by using large language models (LLM); signals of negative emotion or disappointment to be detected; a dynamic customer churn risk score to be generated by evaluating customer behavior data in the form of reduced usage, fewer logins, late payments, cancellations of automatic bill payments, high withdrawal s / transfers and security transfers; and actions towards regaining customers to be taken by triggering a personalized and proactive feedback mechanism as a result of integrated analysis; comprising;at least one database (2) which is configured to keep a record of customer data; and characterized byat least one server (3) which is configured to collect customer interactions, to analyze banking transactions of customers, to perform customer emotion and behavior analysis by using large language models and / or machine learning algorithms, to generate a dynamic customer chum risk score and to take actions customized for customers above a certain risk score.

2. A system (1) according to Claim 1; characterized by the database (2) which is configured to establish communication and perform data exchange with the server (3) by using any communication protocol.

3. A system (1) according to Claim 1 or 2; characterized by the database (2) which is configured to be managed by the server (3).

4. A system (1) according to any one of the preceding claims; characterized by the server (3) which is configured to establish communication and perform data exchange with the database (2) by using any communication protocol.

5. A system (1) according to any one of the preceding claims; characterized by the server (3) which is configured to manage the database (2).

6. A system (1) according to any one of the preceding claims; characterized by the server (3) which is configured to operate in integration with customer interaction agents and different interaction centers in the form of social media, phone calls, chatbot correspondence, surveys and communication channels of the respective bank, to enable customer interactions to be collected and the collected data to be recorded in the database (2).

7. A system (1) according to any one of the preceding claims; characterized by the server (3) which is configured to enable customer behavior analysis to be performed by using variables in the form of frequency of use of the communication channels of the respective bank, number of logins, payment irregularities, automatic bill cancellations, asset drops, large withdrawals / transfers and securities transfers of the customer.

8. A system (1) according to any one of the preceding claims; characterized by the server (3) which is configured to enable customized feedback to be generated for customers with high customer churn risk scores and to be automatically transmitted to the customer through communication channels.

9. A system (1) according to any one of the preceding claims; characterized by the server (3) which is configured to enable actions consisting of processes in the form of re-contacting the customer, apology email, loyalty reward, discount, advantageous prices to be performed for customers with high risk scores.

Images