1. Generalist, Consultancy-Style AI Providers
The lending industry and the application of AI within it is a tightly regulated and specialized environment that favors those with experience and knowledge of its inner workings. While some highly sophisticated generalists may be able to produce a functional AI model to score your applicants, they are highly unlikely to be able to pivot with industry disruptions and business shifts. As a result, these specially-built models tend to become one-trick ponies that are excellent at what they do in the short term, but will require an expensive replacement/extensive adjustment to remain function in the long term.
2. Over-Specialization in One Aspect of the Credit Ladder
Oftentimes, given that the conventional primes and super-primes are generally well-scored, companies will focus on subprime scoring. While specialization in subprime scoring is a necessity in order to provide meaningful a meaningful boost to your decisioning systems, companies that hyperfixate on subprime applicants tend to ignore the remainder of the market. By ignoring the rest of the credit quality ladder, you can miss out on a multi-billion dollar market and/or lose your customers as they build credit and graduate from subprime lending.
3. Inability to Answer Multiple Lending Questions
Sophisticated decision management platforms must be able to operationalize its insights in the context of your business. Credit scoring alone is generally insufficient. Look for platforms that can provide recommended loan amounts, maximum loan amounts, and recommended interest rates based on a superior scoring system: not single-purpose systems that are unable to expand to answer the questions you need answered.
4. Inability to Design Customized Workflows
By integrating different types of data at varying stages across the approval process in a waterfall workflow, sophisticated platforms can accommodate your unique workflow needs to mitigate costs, optimize efficiency, and maximize your profits. If your prospective solution is unable to customize to your needs, it’s time to make the switch to a true platform.
5. AI/ML Add-Ins
While AI alone is not sufficient to create an effective platform, it is a necessary component. AI/ML must be a fundamental piece of your platform and should ideally be an initial consideration in its construction. If a provider is attempting to sprinkle on AI at the end of the recipe, run away.
