for these free STRATMOR
online publications
If you’ve been to a conference in the past few months, you probably noticed that AI is dominating nearly every conversation. It doesn’t seem to matter whether the topic is marketing, sales, customer service, operations, or recruiting — someone inevitably steers the discussion back to artificial intelligence and how it’s poised to transform the industry.
Listening to all of it, you’d be forgiven for thinking borrowers have already changed the way they shop for mortgages. But when we looked at the data, we found something surprising: while the mortgage industry is racing toward AI, borrowers trying to pick a mortgage lender largely aren’t – at least not yet.
Our question this month: If mortgage lender AI adoption is outpacing borrower adoption, what does that mean for lenders?
The funny thing about every major technology shift is that we tend to assume consumers are adapting faster than they actually are. I’ve seen this happen multiple times throughout my career. When websites became mainstream, we were told borrowers would no longer need loan officers. Then came social media, and we were told that personal relationships would become less important than digital presence. Then online reviews exploded, and many predicted that borrowers would begin choosing lenders the same way they chose restaurants or hotels.
Now, it’s AI. And to be fair, AI really is different. The technology is remarkable. I use it almost every day, and chances are you do too. I spoke at length in my last article about the importance of making sure you’re discoverable by AI engines. But one thing I’ve learned over the years is that technology adoption and behavior change aren’t always the same thing. Just because consumers have access to a new tool doesn’t necessarily mean they’ve changed how they make important decisions.
So, we decided to take a closer look at what borrowers are actually doing today when it comes time to choose a lender. What we found was fascinating.
Among refinance borrowers surveyed through STRATMOR’s MortgageCX program, more than 90% started by contacting a lender directly. Half reached out to their current servicer. Another 26.6% were contacted by a loan officer. By the time most borrowers decide they want to refinance, they’ve largely already decided who they want to work with, with nearly 87% choosing a lender or loan officer that came from a referral or one they already knew.
What really surprised us was how little independent research many borrowers were doing. More than 80% either didn’t read online reviews at all (76.4%) or couldn’t remember whether they did (4.1%). Less than 2% found their lender through independent research, including online searches, reviews, and AI recommendations combined. In fact, AI accounted for just 0.8% of lender discovery among refinance borrowers.
The purchase side told a similar story.
Nearly 90% of homebuyers surveyed revealed they selected their lender through a referral or existing relationship. Realtor referrals alone accounted for nearly half of all lender selections (49.4%), followed by friends and family (13.6%), builders (8.9%), and existing lender relationships (10.5%).
While purchase borrowers conducted slightly more research than refinance borrowers, only about 4% found their lender through online searches, reviews, or AI recommendations. Just 1.0% reported selecting their lender because of an AI recommendation.
The takeaway?
Despite all the attention AI is receiving across the mortgage industry, borrowers continue to choose lenders overwhelmingly through relationships, referrals, and prior experiences.
One of the challenges of leading through any period of rapid change is figuring out what deserves your attention today versus what might matter tomorrow.
The mortgage industry has never been short on trends. Every few years, a new technology emerges that has the potential to reshape how borrowers find lenders and make decisions. Some of those innovations end up changing the industry in meaningful ways. Others take much longer to influence consumer behavior than anyone expected.
That’s why I found this data so interesting.
If borrowers were suddenly turning to AI as their primary source of lender recommendations, the implications would be significant. It would suggest that lenders need to rethink how they market themselves, where they invest, and how they compete for attention. But that’s not what we’re seeing — at least not yet. To me, the lesson isn’t to ignore AI. It’s to keep it in perspective.
Most lenders think they have an AI problem they need to solve, but what they really have is a relationship problem, a retention problem, or an experience problem. The lenders that consistently earn referrals, stay connected to past borrowers, and create processes that people genuinely want to talk about still have a tremendous advantage.
The future may eventually belong to AI-assisted discovery. But today’s business is still being driven by trust.
As I dug into the data, I found myself thinking about the online review boom from a few years ago. Remember when many of us thought reviews were going to fundamentally change how borrowers chose lenders? The assumption was that borrowers would start shopping for mortgages the same way they shop for restaurants, hotels, or consumer products. More reviews would mean more visibility, and more visibility would mean more business.
What actually happened was something a little different. Reviews became a validation tool. Borrowers would hear about a lender from a Realtor, a friend, a family member, or a prior relationship, and then they might go online to make sure that lender looked credible. The relationship created the opportunity. The review simply reinforced the decision.
As I look at the AI data from YTD 2026, I can’t help but wonder if we’re seeing something similar unfold again.
AI is clearly becoming part of the consumer journey. Borrowers are using it to gather information, compare options, and educate themselves. But at least for now, it appears to be helping borrowers validate decisions more than it is helping them make decisions. And that’s an important distinction. Because if borrowers are still overwhelmingly choosing lenders through relationships and referrals, then the biggest growth opportunity isn’t necessarily becoming more discoverable. It’s becoming more recommendable.
The lenders winning today aren’t necessarily the ones with the most sophisticated AI strategy. They’re the ones creating experiences that Realtors want to refer, borrowers want to repeat, and customers want to talk about.
Technology may eventually change how people discover you, but great experiences are still what make people choose you.
Here are three actions lenders should prioritize today:
1. Protect Your Referral Ecosystem. The data is clear: referrals continue to dominate borrower acquisition. That means Realtors, builders, financial advisors, past borrowers, and servicing customers remain your most valuable growth channels. Invest in those relationships aggressively. They are still generating far more business than AI, search engines, or review sites.
2. Treat Every Borrower Interaction as Future Marketing. Most lenders think of marketing as something that happens before a borrower chooses them. In reality, the loan process itself is often the most powerful marketing activity you’ll ever undertake. Every expectation you set, every update you provide, every problem you solve, and every frustration you prevent contributes directly to future referrals and repeat business. The borrower experience is still your most effective lead-generation strategy.
3. Plant Seeds for AI Without Betting the Farm. This isn’t a recommendation to ignore AI. Public-facing reviews still matter. Reputation management still matters. Creating content that can be found and cited by AI systems is a worthwhile long-term investment. But treat AI discoverability like planting an orchard, not harvesting a crop. The benefits may be significant in the future, but they are unlikely to outperform relationship-driven growth in the near term.
Final Thought: AI may eventually reshape how borrowers discover lenders. But today, borrowers are still overwhelmingly choosing lenders the old-fashioned way: through trust. And trust is still earned through relationships, referrals, and consistently great customer experiences. The technology may be changing rapidly. Human nature isn’t.
To find more Customer Experience Tips, click here.
MortgageCX is now integrated with Encompass®! Michael Seminari
Start receiving monthly Customer Experience Tips
Feel free to reach out with any questions or needs we might be able to help with.
