Automated Churn Prediction For A Multinational Bank | Marlabs

Automated churn prediction for a multinational bank


The client is a leading multinational bank and financial services firm with global reach, offering products and services across personal, corporate, investment banking, and wealth management. The client was looking for a way to identify and mitigate the possibility of customer churn.

Business Challenge:

Customer churn has become a major problem within the banking industry and banks need to track customer interaction on a regular basis to detect early warning signs in customer behaviour such as decrease in transaction value and frequency, inactive accounts and late payments and take steps to prevent churn. Acquiring a new customer is 5X more expensive than retaining an existing customer. With increasing competition from new age data-driven fin-tech companies, retaining existing customers has become more crucial than ever.

Marlabs Solution:

This solution was built on mAdvisor, utilizing its Automated Prediction module. mAdvisor, an AI & Cognitive Computing Platform that aims to automate the entire process of data analysis and machine learning using cognitive approaches like Machine Reasoning, NLP, NLG etc. mAdvisor analyzes their current customer base, to evaluate and identify patterns to predict customer churn. It also identifies reasons and influencing factors of customer churn for the identified customer segment. The scored result was used by the marketing team to identify customers who are likely to churn and target them with personalized campaigns.

Key Features:

Ensemble Model – Alongside a wide range of algorithms that are available in mAdvisor, for this solution we used an ensemble model which optimized parameters from different algorithms.

Personalized Targeting – Based on the churn probabilities assigned in the churn model, micro segments were created and based on the historical usage patterns of these customers, personalized campaigns were designed by the marketing team.

The Impact:

The solution was deployed as an mAdvisor based cloud solution with a modular design in a private cloud instance of the client. The primary users for the solution was Business Analysts who used the scored model to improve their campaigns. The solution identified churn behaviour accurately which led to savings in terms of revenues. mAdvisor also helped the Client understand key influencing variables and relationship between them pertaining to customer churn rates.

  • Reduced the monthly customer churn rate from 2.05% to 1.5%
  • Saved $2.3 million a month in transaction value


About mAdvisor

mAdvisor is a patent pending AI & Cognitive Computing platform, which helps enterprises to translate data into meaningful insights and narratives without any manual intervention. Now enterprises can reduce the analytics timelines from weeks to minutes using mAdvisor. mAdvisor employs cognitive technologies like machine learning, machine reasoning, deep learning, natural language generation, natural language processing and expert rules systems, thereby enabling enterprises to identify revenue streams, enhance customer experience and productivity.

Marlabs Inc.

About Marlabs Inc.

Marlabs helps leading companies around the world make operations sleeker, keep customers closer, transform data into decisions, de-risk cyberspace, boost legacy systems, and capture novel opportunities and digital-led revenues. It provides digital-first strategy and advisory services, rapid solution incubation and prototyping, and agile digital solution engineering. Marlabs is headquartered in New Jersey, with offices in the US, Germany, and India. Its 2,500+ global workforce includes highly experienced technology, platform, and industry specialists from the world’s leading technical universities.

Recommended Reads

Ready to step into your digital future?
Let's go there together.

Contact Us
What’s Happening
close slider