Mortgage Originators’ Guide to Success in the Age of AI

The future is coming into focus more and more each day. Pretty soon, our cars will all drive (maybe even fly) themselves, we’ll be able to play a round of golf at Pebble Beach with nothing more than a set of clubs and an Apple Vision Pro headset, and we’ll be adding skills like “Master GPT Prompter” to our resumés. Imagining the future can be fun, but it can also be unnerving. Will robots take over the world? Or scarier yet, will they simply take all our jobs? That gives “Terminator” a whole new meaning!

We may not be in danger of those dire scenarios just yet, but here’s a prediction that is guaranteed to come true: AI will continue to improve and expand at a rapid pace. Consider this article to be your “Robot Defense Manual” for showing you how to remain indispensable as a mortgage professional in the age of AI. Our question this month: How can mortgage professionals benefit from AI while also taking advantage of its limiting factors?

Benefit from AI by Taking Advantage of Its Limitations

If you haven’t yet played around with ChatGPT, you should. It’s more than a search engine on steroids or a way for students to enlist robots to write their term papers. It will soon completely transform the way we search for and acquire information. Jobs won’t disappear as much as they’ll just morph into something different. Take a mortgage originator, for example. Here is a list of the main tasks that a loan officer (LO) completes during a loan process. See if you can spot any that could not be performed by AI.

1. Educate the borrower on product options
2. Explain what the loan process will look like
3. Gather basic information about the borrower
4. Calculate loan amount (house price) they can afford
5. Assess borrower’s risk tolerance
6. Provide a rate quote and payment options
7. Assist with loan application (and potentially collect documents)
8. Convey periodic progress updates
9. Prepare borrower for closing by reviewing closing disclosure
10. Attend closing virtually or in person

The last one is the only one I see, and even that may be possible with AI sooner than we think. But this doesn’t necessarily mean LOs are suddenly on the endangered species list. They will need to adapt and reinvent themselves by harnessing, not competing with, AI.

The mortgage industry has been one of the slowest industries to adopt new technology, mainly because the burden has been on the lenders to invest in and implement new software tools, and most are chronically too busy or too cash poor to make those investments. The ones that are forward-thinking enough to pull the trigger often lack the perseverance to see the projects through and get internal adoption. What makes the AI revolution different is that customer adoption, not lender adoption, is driving the change. Customers are not waiting around for lenders to buy top-of-the-line Point of Sale (POS) tools or chatbot software. They’re just going on ChatGPT and prompting a virtual assistant to do their planning and research for them. This will be Gen Z’s M.O. Are you unsure what getting a loan involves? Are you unsure how much house you can afford? Do you know your financing options?

Consider this prompt:
I am in the market for a mortgage loan to purchase a new home. I am going to give you some details about myself, and I would like you to ask me questions to gather any other information you’ll need to recommend several loan programs for my situation (singling out your top recommended option), as well as a few lenders with excellent customer experience reputations. I would also like to know how much house I can afford and what I can expect to pay per month, including taxes and insurance with each option. I am __ years old, married, and make $___K annually. I have a current home worth $___ and my current mortgage balance is $___K. I live in ___. What else do you need to know to complete this task?

In ChatGPT’s current iteration, the results for this query are impressively good, but not comprehensive. I’ll bet you dimes to dollars, however, that they will be in the coming years or maybe even months. What this means is that an originator’s main value will not be in their ability to gather information, quote a rate or complete any of the tasks listed above. Their value will, more than ever before, be centered around soft skills like creating rapport and building trust. This implies recruitment of a whole new breed of LO, not to mention retraining for your current team.

Diagnosis

So, where are the limits where AI cannot and will not venture? Answer that question and you will be well on your way to AI-proofing your career. Keep in mind though, that our relationship with AI need not be like that of Terminator’s John Connor and Skynet. We’re not mortal enemies…at least not yet.

Here are some ways that I foresee AI helping LOs and the limiting factors where humans will never be replaced. Successful originators of the future will harness the efficiencies of technology but focus their personal growth on the softer skills that AI cannot replicate.

Here is a shortlist of what AI will and will not be able to do for borrowers:

• Will educate borrowers on guidelines, financing options, industry changes
• Will not build borrower trust and become their trusted advisor. Trust requires emotional history with someone, not just accumulated knowledge. It also requires believing that someone has your best interests in mind. Both are very big leaps from where AI technology sits today.

• Will ask deep, personal questions, find out the borrower’s “why,” mimic getting to know you.
• Will not actually get to know you, care about you, empathize, be emotionally supportive, or connect on a personal level. AI will likely never be programmed to ask, “How are the kids?” or “How are you feeling about everything?” because the consumer knows it doesn’t actually care.

• Will provide basic steps for shopping, gathering documents, what to expect in a straightforward loan process, as well as generate endless scenarios and options based on risk tolerance, budget, and other factors.
• Will not opine on the best fit for you or think outside the box (creative problem-solving). Sometimes having a lot of options is more overwhelming than it is helpful. And there’s nothing more frustrating than a robot who isn’t tracking with your unique, nuanced situation. Robots are notoriously bad at handling requests for “exceptions” too, which are often reliant on relationships and situational judgement calls.

• Will request and keep track of uploaded documents, give progress updates, find and correct errors.
• Will not apologize when something goes awry. According to STRATMOR data, at least one thing goes wrong on 55 percent of loans. An apology from a robot won’t cut it. This must involve a real human.

• Will learn to pick up on social cues, sarcasm and humor, but will remain far from perfect.
• Will not reliably interpret subtle cues and cultural nuances, which will lead to misinterpretations or inadequate responses. In other words, it can’t be fully human, and if soft skills like emotional intelligence are still being refined by 30-year mortgage veterans, it’s not going to be easy for AI to catch up.

Prescription

Here are three things you can begin doing today to robot-proof your career:

• Build Trust. As mentioned earlier, trust takes time, so the sooner you start the better. Trust is built when someone sees you as an expert, when your advice is tailored to their specific set of circumstances (meaning you listen well), and when they believe you have their best interests at heart. A robot can compete on the first two, but not the third.
• Get Emotional. One of the best questions you can ask your borrowers during the loan process, ideally multiple times, is “How are you feeling about everything?” It’s sincere. It’s open-ended. It gives someone who may be tied up in knots with anxiety a release. And it can’t be mimicked by a robot. Best of all, it gives you a chance to help right the ship if anything is amiss.
• Get Personal. If your knowledge of your borrowers is limited to their financial picture, you’re doing it wrong. Don’t just get to know them (their kid’s names, their dog, their dream vacation, their favorite foods). Document it and ask them about these things periodically, like you would in any friendship. The thing about getting someone to feel like you care is…to actually care. Show me a robot that can do that.

How can you learn more about creating a better customer experience and about how the customer experience impacts your company?

Find out more about STRATMOR Group’s MortgageCX program and how it is revolutionizing the CX space. Contact Mike Seminari at mike.seminari@stratmorgroup.com.

To find more monthly Customer Experience Tips, click here.

How Can We Help?

Feel free to reach out with any questions or needs we might be able to help with.

Follow us on social media:

Site Map