Analytics & Big Data
Dec 20, 2018

Why predictive selling is set to become a game changer

In the commercial lending space, one of the current challenges facing mid-market players is the heavy investment they must make in their sales organizations and their relationship managers (RMs). Fortunately, to help them get more complete insight into where the sales opportunities are, companies can now apply predictive analytics in selling.

A lot has been happening with analytics in the small-market space but very little in the mid-market. This makes the possibilities for analytically driven predictive selling all the more important to mid-market players, their sales organizations, and, particularly, their RMs. Thanks to the latest developments in artificial intelligence (AI), predictive analytics can be used to help commercial lenders identify opportunities to sell loans. This type of predictive selling takes three basic forms.

Selling to a prospect who's not an existing customer

A good example of this is how lenders can estimate the revenue of a restaurant franchise based on secondary signals like its location and the number of Yelp ratings. They can also use internal data from their current portfolio of restaurant loans and assess whether or not the prospect is within the target market. RMs can then prospect the business with all these analytics in hand.

Cross-selling to existing customers

We've seen that a large number of our clients, in particular commercial banks, are good candidates for using predictive cross-selling. For example, banks can increase the number of checking-account holders who also have a commercial line of credit simply by creating a 360-degree view of the customer across silos. Key predictive data from these silos could come from the RM's call report in the customer-relationship-management system, from a checking account transaction history, or even from customer-service transcripts. Imagine the predictive power of all these data streams coming together. Now imagine that collective insight being aggregated with astonishing speed with the help of AI. For instance, checking-account data may show a customer made a large down payment. This intelligence, combined with the call report and other data, could strongly indicate the customer is in the process of expanding in some areas and may need a loan.

Targeting potential repeat borrowers for sales

Our internal research shows that existing borrowers are the source of 60%-80% of new loan originations. Pursuing repeat business is the most effective way of keeping origination costs low, so preventing customer attrition is vital. But in going after such business, it's important to remember that the risk of attrition is much higher when the borrower doesn't have a sticky product with the lender, such as a checking account or working capital loan. In this scenario, data mining can be used to create an early alert system for potential trouble, such as when a customer calls for a quote for an early loan payoff. This type of simple query could be a huge red flag that the customer is looking to move elsewhere. With predictive data, smart lenders can act before it's too late.

About the author

Anu Sachdeva

Anu Sachdeva

Global service line leader, Commercial Banking

Follow Anu Sachdeva on LinkedIn