What is a well built decision engine?

The well built decision engine is not simply a rules engine where it follows a prescribed set of rules and produces a repetitive customer experience day in and day out. A well designed decision engine produces results as soon as it starts to digest information from the applicant or a third party data provider.

A decision engine should be interactive with the applicant based on the information that it receives from the applicant as well as information gathered from third party data providers. It should make decisions to engage the client as quickly as possible.

What are the bank's impending issues today?

Most of the banks are facing a dire situation today. They have managed to silo themselves not only in the traditional organizational structure sense of the way, but also from a technical point of view. 

Out of necessity, the banks have to grow. However, their banking system, their banking core systems were traditionally built for single products and a single marketing channel, namely branch interactions.

These traditional systems don't have a client interface, it requires bank tellers to input information in their bank terminals and it makes it extremely difficult to scale.

On top of these limitations, the banks have amassed multiple “cores” to run their checking account, loans, credit cards, document store and an enterprise data warehouse that is strained to keep up with smashing multiple database designs from various “cores” and make sense of it all. Even the biggest banks have this problem. 

The end result prevents banks from innovating and growing. Their entire technology staff are sucked into keeping their current technology stack working or slaving away to keep data integrity intact from various data sources gathered from all of their core systems.

The advent of “side cores”

“Side cores” are essentially newer technologies with a higher level of API (application programming interface) to let potential clients send in applications online.

Unfortunately and to further propagate these constraints, banks are forced to innovate by bringing in even more side cores to patch their offering. This adds more strain to their technology office and their database.

Although side cores can temporarily increase sales for the bank, the technical debt is still piling up and the ultimate cost to keep data integrity together and whatever technology to let these cores talk to each other eventually becomes a burden to the banks.

The problem still persists: “One stop shop for all product offerings.”

The original issues facing the banks is that their staff managing and working with multiple cores fails to talk to each other. Even though these cores are “data aware”, meaning that at a household level, the banks and their cores might know all of the products belong to the household, but it’s after the fact. The bank needs to present these product offerings right at the beginning of the customer interaction.

In other words, the banks’ original sin is that they need a platform that can present multiple offers during their first interaction with their potential clients. And if there is new offerings coming online, the decision engine and product engine should be able to create these product offerings quickly so they can present these new products to their potential clients without having to launch another onboarding process.

What if the banks can use one onboarding platform with a decision engine that can offer multiple products? What would that look like?

The solution is a multi-product decision engine

To solve the banks multi-core strategy that’s been plaguing their growth, it must move to a digital onboarding process that offers multiple products. So what are some of the components required to offer multiple products?

• A single intake form: Most of the common product offerings require the same information from the applicants. In essence, the bank needs to collect enough information to pull credit as well as verify income. In some cases, if there is an asset involved such as a car, solar panel, the application might need to take into account the information on the assets themselves. However, to pre-approve the applicant, only a small amount of information is needed before moving onto collecting additional information.

• Multi-Product Underwriting: From a singular credit pull and income verification process, the decision engine must be able to run product specific underwriting paths and render a product specific decision. For instance, a bank’s DDA (demand deposit account) opening requirements is vastly different from offering a $5,000 line of credit on a credit card. However, these multi-product underwriting rules must exist in a decision that can run these pre-approval rules and return the results back to the client. If there are 3 product offerings, and the client only approved for 2, then 2 offers (DDA account and a Credit Card, sans Auto Loan) should be displayed to the client in real time.

• Multi-Product Offerings: A critical part of a decision engine is its ability to house multiple product offerings and multiple flavoring of each product offering. If the client is approved from a DDA account and a Credit Card, the decision engine’s product engine should show a DDA account with fee and without fee and Credit Card offerings with different reward offerings. 

• Multi-Product Workflow Orchestration: From an applicant’s perspective they should be able to agree and sign documents for multiple products. As such, the back office staff should also be advised on their tasks depending on the customer’s selection. If we use the previous example, if the client picked two products, a DDA account and a credit card. The workflow to prepare for a DDA account is different from getting additional information to the applicant for a credit card. The back off staff might have interacted with different vendors to fulfill these products.

• Multi-Core Integration: The onboarding system and its decision engine should have the ability to deliver the approved applicant and the product(s) they selected into the appropriate core and therefore solve the multi-core issue and the cross sell issue that’s hampering the growth of the banks

The solution is here:

Today, we possess the ability to run parallel decision trees and product offerings for banks to present to their clients multiple offerings in a single onboarding interaction.

The bank only has to pull credit bureaus once and verify the applicant’s income once and make multiple product offerings in one sitting and in real time.

About LendAPI:

LendAPI is a no-code, DIY (Do it yourself) digital originations platform with a fully customizable product designer with the ability to manage applications fields, upload documents and customize consents. LendAPI also has application workflow management, visual decision engine, third party integrations as well as various customer communication methods such as SMS and email all in one super affordable investment. A complete multi-tenant, point-of-sale capable,  white-labeled solution that can help you launch your finance products quickly.

To sign up, please visit app.lendapi.com/signup. We have a 30 day free trial with no obligation.