Despite being the stuff of science fiction for more than 50 years, AI burst into the business world with amazing speed. Even in the mortgage industry, where advancements can lag up to a decade behind other industries, these new tools found champions.
One of the most interesting of these individuals is Scott Schang.
Schang has spent 25 years at the intersection of technology and human behavior, primarily serving American homeowners and real estate professionals. Known for his ability to demystify complex topics, Scott has a particular focus on AI’s role in business and marketing.
In fact, he was instrumental in the writing of our recent white paper.
Since 2022, he has immersed himself in the world of artificial intelligence, reading dozens of AI news sources daily and dedicating hundreds of hours to educating others.
Currently, Schang serves as Chief Data Scientist for Redefining Business Intelligence (RBI), a firm trying to help lenders and other industry executives get a return out of their Business Intelligence (BI) investments.
We sat down with Schang to find out more about AI in the mortgage business and specifically why it seems to be so hard to get it right in the mortgage industry.
How is AI Transforming Consumer Direct Marketing?
Schang sees AI as a game-changer for consumer direct marketing, especially in the real estate and mortgage industries.
"AI helps businesses meet customers where they are," he explains. By leveraging AI tools like chatbots and predictive analytics, companies can offer more personalized experiences. This means delivering targeted messaging based on a customer's unique needs, habits, and preferences, significantly enhancing engagement.
For mortgage and real estate professionals, this level of personalization is invaluable.
"Think about AI as a bridge," Schang says. "It connects companies to clients more directly, eliminating some of the guesswork and making interactions more meaningful."
AI helps professionals refine their messaging, making complex products and services more accessible and understandable to consumers.
What Are the Challenges in Building Custom AI Tools?
As someone who has worked with virtually all of the large language models (LLMs) and is very familiar with AI tools, Schang highlights two critical aspects of creating successful AI products: functionality and user experience.
"The first is the functionality—does it do something interesting, unique, or offer a profound perspective on something?" Scott explains. "The second, and I think the most important, is the user experience."
He emphasizes the need to consider the perspective of non-technical users.
"When you build these tools, it’s important to put yourself in a user’s shoes, especially someone who doesn’t have the experience or insight as the GPT builder, presumably a subject matter expert," Scott says.
The challenge lies in designing an experience that is intuitive and accessible. For him, this means continuous testing with various input types, even intentionally vague ones, to ensure that the tool provides useful and clear outputs for a wide range of users.
Is Prompting as Important as People Think?
In the early days of AI, prompting was a major focus, but Schang believes it’s often overemphasized.
"I think early on, we all got caught up in prompting, but I don’t think it’s nearly as important as most people think," he says. The problem with relying on pre-written prompts is that they lack the context and personalization needed for effective AI interaction.
"The trick to prompting is fully thinking out the goal of your GPT, and being able to concisely communicate what the functionality and user experience should be," he explains.
Schang’s approach involves using a custom-built prompt generator, which transforms his brainstorming sessions into tailored prompts.
"In short, I communicate to the AI what I want it to do, and ask it to write the prompt. This works for me most of the time. There are often small tweaks required, but for the most part, it’s very effective," he notes.
His method demonstrates that the real power lies not in copying prompts but in understanding how to articulate specific goals and needs to the AI.
What Does It Take to Make AI Interact Like a Peer?
Schang believes that getting AI to engage with you on a more personal level can be both simple and effective.
"Role-playing is one of the easiest things for ChatGPT to do," he says.
By telling the AI to "act as" a specific persona or context, like a conversation at a bar after a conference, users can create more dynamic interactions. He also recommends experimenting with the voice feature in the ChatGPT app for a more natural back-and-forth.
Schang’s insights highlight a central theme: AI can drive significant improvements in business, but only if it is implemented thoughtfully.
For those in real estate, mortgage, and beyond, embracing AI is less about jumping on the latest trends and more about finding ways to truly connect with clients.
By keeping a clear focus on the customer experience, businesses can use AI to enhance their offerings and create lasting value.
As Schang puts it, "AI is not the future of business—it's already here. The real question is, how will you make it work for you?"
Download our new white paper to learn more. To visit with Schang, or anyone at RBI, about making AI work in your business, visit us online at https://www.rbiplaybook.com or email us at info@rbiplaybook.com.
Comments