Client background

Our client is a Beijing-based financial technology and cloud service provider dedicated to providing specialized technology, operations and knowledge services to all types of financial institutions through fintech and cloud computing.

The company has strong channel resources and geographical advantages to help financial institutions to provide more efficient corporate and personal financial services at a lower cost.

Client Pain Points

China’s banks have focused on corporate banking for years. Personal banking has grown gradually in the past few years. It is challenging for banks to take advantage of AI-powered data mining, data analysis, and predictive analytics as there is relatively less data in the field. Many small and medium-sized banks have only collected quality data for up to two years, and it is impossible to form a good data training set for machine learning.


Reducing the loss of important customers, growing deposits, and increasing customer loyalty and stickiness have always been the immediate needs of various financial institutions.


Mans International helped our client find a Canadian company with more than a decade of data mining and analytics experience in the personal finance sector. The Canadian company leveraged artificial intelligence technologies to optimize their solutions, and reduce the cycle of project deployment.

At present, our client has completed the POC phase for the provincial and municipal banks in East China by using the advanced technology and experience of the Canadian company, and has begun to deploy the solutions to various branches.

Challenges Facing the Financial Industry

Competition in the financial industry is increasingly fierce, and the cost of acquiring quality new customers can be 5-10 times more than retaining existing customers. How to effectively retain existing customers has always been one of the major challenges facing the financial industry.

The loss of commercial bank customers is very serious: the monthly average net loss of banks with $50 billion in monthly transaction volume is about $7-10 billion; the monthly average net loss of banks with $100 billion in monthly transaction volume is from several billion dollars to nearly $10 billion.

The growth of new customers has concealed the loss, making the bank’s growth slower or halting, or even retreating; banks have invested a lot of manpower and resources to attract new customers and drive deposits, but due to customer losses, especially the loss of high-end customers, The bank’s overall income could not meet the expected growth target, and some banks also experienced negative growth. Some banks have already restricted the loan manager’s loan quota, focusing on growing bank deposits. Banks has been operated with less and less net profits.

The Financial Industry has Faced the Following Problems in the Process of Retaining Customers:

  1. The sheer volume of customer data makes it difficult to find the most effective way to recover the most customers and assets cost effectively.
  2. It is challenging to effectively distinguish between general customer churn and key customer churn.
  3. Without solid data scientists with in-depth financial domain expertise to mine the variables and contributions related to loss, the current forecast model on the market has a high error rate and banks cannot implement the possession.
  4. Without a large amount of data training set and tempered AI models, the effect is not guaranteed.
  5. Bank-related data and tables are quite complex. There are few mature data cleaning and preprocessing AI tools available, which are time-consuming and labor-intensive and cannot be effectively implemented.
  6. Banking departments and sub-branches are over-reliant on traditional customer retention methods, without systematic retention strategies and marketing techniques.

The AI-powered Churn Prediction Solutions

We offer three kinds of solutions to banks and other financial institutions:

  1. Highly effective AI-powered churn prediction
  2. Personalized marketing recommendations
  3. Professional sales and customer retention


  • The accuracy of loss prediction is more than 95% on average, and the error of loss is less than 10% (more than 3 times the accuracy of other similar products).
  • Retained nearly 40% of high-end customers for banks.
  • Save about $20 million marketing costs.
  • Through high-end customer retention, the bank recovered 26% of the lost revenue in three months, equivalent to $450 million.

(Note: This is the result of a bank with $10 billion in monthly transaction volume, the bigger banks will save more than ten times of their revenue.)