NexJ : Intelligent Customer Management in Commercial and Corporate Banking

In my previous blog, we discussed how Intelligent Customer Management applies to Wealth Management and Private Banking. This blog will do the same but for Commercial and Corporate Banking. If you read my previous blog, you'll find the description of the mechanics familiar, even though the examples are tailored to Business Banking. You may wish to skip this blog and wait for my next blog: Intelligent Customer Management: Saving Time and Improving Loyalty.

Once again, to implement Intelligent Customer Management, we need to start with business goals. For Commercial and Corporate Banking, the business goals are often to increase share of wallet and improve customer loyalty. How can an application help bankers accomplish these goals?

A key process for achieving these goals is account planning. Today, the client team gets together on a regular basis to decide which products would be best suited for the client in the next year. Decisions are based on how well the team understands the client, the various geographies and trade corridors, and what worked in the past. A well-crafted, manually created account plan can deliver great results for a bank, but it is limited by the assumptions and past experience of the specific people involved. However, by letting the actual data speak for itself, we can get much better results with Guided Account Planning and Deal Management.

Guided Account Planning and Deal Management occurs in two stages. At the planning stage, we need to understand the client and recommend products. At the execution stage, we need to drive the deals to sell the products.

In the planning stage, we start by better understanding the clients and how they relate to other clients in given geographies and trade corridors. We use a form of machine learning called unsupervised learning to comb through data and find clusters. In this case, grouping clients into cohorts of similar companies based on who they are, where they are located, and who they do business with.

Now that we have the cohort, we need to determine the correct patterns of products these similar clients typically purchase. This is akin to the process that Amazon uses to recommend products based on previous purchases. Another machine learning method, supervised learning, is used to train a decision model to make predictions or deliver a score. The decision model is trained by using historical data to find patterns. The data must include both inputs and outcomes. In this case, the inputs are all of the client profiles and the outcome is the collection of products they purchased. The result is a decision model that predicts the best set of patterns for a given client.

In the execution stage, we also use supervised learning to create a decision model that predicts how to best sell a given product. The inputs are all of the interactions that occurred for historical deals and the outcomes are the subsequent decisions to purchase the product. The result is a decision model that predicts the best set of interactions for a given deal.

The prediction can be further tuned if the inputs include information locked in unstructured meeting notes and emails. Natural Language Processing is used to extract names, relationships, intentions, and tone from the unstructured text. All of these attributes are added to the input to create better decision models.

The benefit of Guided Account Planning and Deal Management is recommendations for both products to sell and actions to close the deals.

Is your firm considering the use of Artificial Intelligence (AI) to influence banker behavior? I would love to hear your experiences.

Tune into my next blog where I will discuss how machine intelligence can be used to save time and increase customer loyalty.