Churn Early Warning System

Company: Azuga Telematics Pvt. Ltd.

Duration: 1 Year.

Role: Sr. Business Analyst
Project Leadership: Data Analytics, Process Re-Engineering, SOP Documentation, Cross Functional Team coordination, System and Process Integrations

The​‍​‌‍​‍‌​‍​‌‍​‍‌ Churned Early Warning System aimed to use any available means to intercept data trends showing commonly that a customer would be churned in near future, figure out what was going on and ultimately retain the customer. Therefore, the main objective was to use data insights to assign the probability of churn for each customer and thus be able to intervene in time, thereby retaining customers, obtaining their loyalty and sustaining revenue streams.

Project Scope and Objectives

• Utilisation of the data sources and more with as much granularity and freshness as possible as the main input for the analysis.

• Creating a churn risk segmentation model rendering four groups of customers: "High Risk", "Moderate Risk", "Low Risk", and "No Risk".

• Provision of customer health information to Salesforce CRM.

• We target customer retention and the biggest leverage of ARR (Annual Recurring Revenue) thus is most certainly the goal, i.e., customer subscription base growth or increase.

Leadership and Project Management Approach

• Project Manager and chief data analyst’s interlocutor; the working relation was established with the Data Science team regarding the alignment of tasks concerning data quality, model development, and deployment time frames.

• The project was broken down into discrete milestones that spanned from assessing data to validating the model and finally to integrating the system—all employing agile methodologies to benefit from the iterative feedback.

• Made sure that all the workshop participants (data scientists, CRM administrators, sales, and retention teams) were prepared and that the execution was flawless by organising communication among these groups, confirming understanding of requirements, and coordination.

• By anticipating problems such as data scarcity and issues with system integration early in the game, they orchestrated risk mitigation moves to retain project momentum and thus managed risk well.

• The updates they shared with the stakeholders and the progress they made towards KPIs were very helpful in terms of communicating with the stakeholders, thus encouraging openness and participation.

Data Analysis and Insights

• They studied a huge amount of data from many different angles concerning customer behaviour and company operations, and this gave them an accurate picture of what was causing customer churn.

• They came up with a predictive scoring system utilizing behavioural and transactional indicators with which the score of churn intent is assigned and customers are grouped into risk segments accordingly.

• They compared the model with the occurrence of actual churn cases in the past to confirm the model's accuracy and at the same time, they perfected the model features to enhance the correctness and lessen the false positives.

• They fixed the refresh intervals for the scores of the churn risk to keep up with the ever-changing customer behaviour.

Business Process Improvements

• The implementation of the Salesforce CRM system makes it possible to directly integrate scores of the churn risk, automating health updates of customers and bringing visibility that is almost in real-time to the sales and retention teams, thus making the system more interactive.

• By defining target groups and using their risk levels to do prioritization, retention activities were made more efficient as the time of outreach and thus, customer engagement, were tailored.

• A sales input system was introduced to continuously update the predictive model with the latest in sales indicators and thus refine churn indicators and update thee predictive model.

• The enhancement of the following billing processes was done by communicating that it is the task of proactive intervention to take care of overdue accounts with a high risk of churn, hence, enabling the early solving of the problem.

Project Outcomes and Achievements

• Over the first year since its enactment, customer retention has increased by 12%, and this improvement can be very directly linked to the targeted retention measures enabled by the churn scoring model, i.e., the latter’s success.

• The ongoing customer attrition reduction efforts have also led to additional Annual Recurring Revenue generation to the amount of around US$ 400,000. 

• It has become possible for the company to make decisions and set priorities when it comes to customer retention by using data which in turn has led to an increase in operational efficiency and the level of customer satisfaction.

• In fact, the project worked as a stepping stone not only to a system less bound to certain conditions but also to future predictive analytics projects that could be aligned with this system ​‍​‌‍​‍‌​‍​‌‍​‍‌integration.