Minimize revenue loss from churn by using proactive trigger

Only a few customers suddenly want to close a card due to reasons that can neither be predicted nor can be actioned upon. Most customers show a gradual journey towards attrition. Of the 20-30% customers frequently provide early indicators of changing engagement through their financial behavior, online interactions, and contact center communications.

In many cases problems like Fraud, Disputes, or Transaction Decline have a major impact on customer engagement, leading them to rethink the relationship. In many cases, very prompt and effective customer service can soothe the customer and bring them back to engagement.

Unfortunately, many businesses lack the appropriate data intelligence framework to detect these subtle changes promptly, often noticing them only when they become more pronounced.

If these subtle changes are monitored over time and appropriate measures are implemented to mitigate them, businesses can achieve a longer lead time to identify potential attrition and take steps to prevent customers from leaving.

This solution is a highly proactive approach to identifying changes in customer engagement and sentiment through three types of triggers

1. Decline in financial transactions
2. Changes in contact pattern or complaints
3. Changes in customer sentiment

It employs an advanced network model to analyze the relationships between various events in a customer’s lifecycle and correlate them with attrition probability. It classifies these events with a severity score, ranging from low to critical, based on the customer’s profile, lifecycle, and associated events.

These triggers can then be used to immediately attend to such customers to offer them help and solve their problems.

This solution is designed for businesses that:

1. Aim to enhance customer engagement and servicing by being attuned to customer needs and changes.
2. Intend to adopt an attrition strategy that offers a substantial edge over competitors.
3. Have targeted revenue objectives to meet by strategically reducing attrition.

Revenue: Achieve increased long-term revenue through higher customer engagement over a 2-5 year period.

Cost: Expect lower attrition and, consequently, reduced acquisition costs. This solution is designed to enhance customer loyalty over time.

The model incorporates the following characteristics:

1. Category-specific spending data
2. Credit limit utilization
3. Attrition data
4. Payment behavior patterns
5. Risk profiling
6. Contact information
7. Website navigation data

The following types of models can be used:

1. Classical Logistic regression models
2. Gradient Boosting
3. Xtreme Gradient Boosting
4. Support Vector Machine
5. Random Forest
6. Deep learning techniques – CNNs, RNNs
7. Ensemble Models

The best approach is to try multiple models and pick the one with the best performance.

1. On-prem ( on customer systems or on DeepQ-AI Environment)
2. On hosted cloud space ( Customer or DeepQ-AI Environment)
   ( deployment is subject to data availability in the same environment, or feasibility of seamless data transfers within secured environments)

1. Final outputs could include a list of triggers with their impact and level of severity.
2. PowerPoint summaries or automated Excel files.
3. Interactive dashboards.