Image: HousingWire (illustration)
By Ben Anderson | HousingWire
Commentary
I spent more than 20 years originating mortgages, and at one point ranked among the top 10 producers in the country. I know the ins and outs of loans and lending, which is why I’m always suspicious when lenders talk about AI. When they do, I listen out for one thing, and one thing only.
Has the person building the new AI tool ever dealt with three closings, a lock about to expire and a borrower who has gone AWOL in one afternoon? Chances are, they haven’t. And this is what concerns me the most.
AI is coming to mortgage lending whether originators like it or not. Yet, the real question is who shapes it. My view is simple: the people who know this business inside out should have the biggest say in how these tools work, because a tool that doesn’t understand how loans actually get done will create as many problems as it solves.
Where the business really leaks
Ask any producing loan officer at what stage they lose the most deals, and you’ll get a similar answer. It’s rarely the hot file. The hot file gets everyone’s attention.
It’s the lead who got one call and then nothing. It’s the past client who said, “Call me in six months,” and then they never heard back. It’s the pre-approved buyer who went quiet after losing a thrilling bidding war, and nobody checked in when the next listing came up.
None of this comes from pure laziness. Most originators quite frankly care a great deal about their clients. But a CRM full of thousands of names doesn’t pick up the phone, and someone chasing conditions, appraisals and closing dates only has so many hours in the day.
The gap between what a good originator knows they should do and what they actually have time to do is where the money goes missing. Although I’m sure this applies to many industries.
That gap is where AI can earn its place in the mortgage market. Not by replacing the loan officer, but by handling the steady, repetitive follow-ups that fall apart the moment a busy workweek hits. AI can be relentless and logical, while we often need a lot of caffeine to keep us going when the pressure kicks in.
Why generic AI falls short
Not all AI tools are made equal. As I’m sure you’re aware. A general-purpose chatbot can quite easily reveal your office hours to a curious customer. But mortgage conversations are different and deeper.
When a first-time buyer asks, “Should I wait for rates to come down?”, they’re rarely asking about rates. In fact, this is where emotional intelligence comes into play. More often than not, they’re actually nervous about the monthly payment, their spouse isn’t sold yet or perhaps a previous lender has sadly let them down. An experienced originator hears that simple question and considers the reasons behind it. A generic tool answers the literal question and moves on, and so the borrower drifts away.
Then there’s compliance. Borrower communication in this industry sits under disclosure rules, fair lending obligations, state licensing requirements and consent rules for calls and texts.
An AI tool that quotes terms it shouldn’t or answers a question that belongs to a licensed loan officer isn’t saving anyone precious time. It’s creating exposure. It’s potentially putting your business’ reputation at risk. Now, software built by people who have lived with those rules tends to know where the lines in the sand have been drawn. Software built by those unfamiliar with these rules can overstep them and cause significant issues.
Great production can be taught
We’re social creatures. Even introverts enjoy human interaction, even if it’s few and far between. In fact, one of the easiest marketing strategies these days is personal branding, because we buy into people, not products. Being a coach taught me that habits, timing, consistent follow-ups and knowing just what to say when a borrower hesitates make a whole world of difference to your sales potential.
Years ago, a notary named Daniel came into my office to sign documents. I thought he’d be good at this business and told him to get licensed. He did. Years later, he tracked me down to tell me he’d become the top originator in his market. He didn’t start with some rare gift. He learned the habits and stuck with them.
If those habits can be taught, they can be built into tools. That’s the real opportunity I see in the market. AI built on the knowledge of experienced producers can give a second-year loan officer some of the discipline it took the rest of us a decade to develop.
I also learned something from building software for loan officers. Tools that depend on someone logging in and remembering to act don’t change much. The originator still carries all the work. The next generation of tools has to take some of that weight off, not add another dashboard to check.
That frustration is actually what led me to build Cindie, an AI assistant for mortgage teams that keeps up follow-up with leads and past clients, and hands the conversation back to the loan officer when a licensed professional needs to step in.
Five questions to ask before you adopt an AI tool
If you’re evaluating AI for your team, these are the questions I’d want answered first.
- Who built it? Has anyone on the team originated loans, or managed people who did? You’ll feel the difference within a week.
- Does it know when to stop? Good tools hand the conversation to a licensed person at the right moment. Ask the vendor to show you exactly where that handoff happens.
- How does it handle compliance? Ask about disclosures, licensing, fair lending and consent for calls and texts. Vague answers here are a red flag.
- Does it sound like you? Borrowers can tell when a message is canned. The tool should reflect your voice and the way you work with clients.
- Does it take work off your plate? If it adds another screen you have to remember to check, it probably won’t get used.
Originators need a seat at the table
The mortgage industry gets to decide who shapes its AI. If originators sit this one out, the tools will be designed by people who have never explained a rate lock to a nervous first-time buyer, and we’ll spend years working around their blind spots.
So I urge you to get involved. Test the tools on the market. Push vendors with hard questions. Tell them what actually happens in your day, and share what’s working with your peers.
AI shouldn’t erase the value of experience in this business. Built the right way, it can be a superpower.
Ben Anderson is the founder of Cindie, an AI workforce built for mortgage lending. He spent more than 20 years originating mortgages and has coached hundreds of loan officers through Ben Anderson 365.
This column does not necessarily reflect the opinion of HousingWire’s editorial department and its owners.
Originally published by HousingWire.
HousingWire
© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.
