By
Katie Shive
•
August 18, 2026
For most of the internet's life, you had one audience to win people over. You wrote your bio, your posts, emails, and website copy for a human being who might become a client or a referral partner. Everything you did to get discovered was aimed at another person.
That's not true anymore. Sometime in 2024, automated traffic quietly passed human traffic for the first time in the internet's history. The 2025 Imperva Bad Bot Report clocked bots at 51% of all web activity, and by the end of 2025 Cloudflare, which handles requests for roughly one in five websites on Earth, put the split closer to 53% machine to 47% human. Cloudflare's CEO had predicted this crossover would happen in 2027. It arrived 18 months early.
Read that again, because it should get your attention. More of the internet is now machines than people, and a growing share of those machines are the large language models (LLMs) your clients are asking for a mortgage recommendation.
So you have two audiences now. Humans, still. And the LLMs that increasingly stand between you and those humans.
Here's the part most loan officers haven't caught up to. When someone opens ChatGPT and types "Who is a good mortgage lender in Boise that's great with first-time buyers" the model is scanning everything it can find about the people who might fit that request, weighing the signals, and deciding who to put its name behind.
That last part matters more than anything else I'm going to tell you.
Think about how you refer people in real life such as your hair stylist, your kid's dentist, and the buyer's agent that you send half your business When you refer someone, your own reputation rides along with it. If the person you vouched for delivers, you look good; if they drop the ball, that's on you too. So you refer the people you trust to make you look correct.
An LLM works the same way, minus the feelings. It has one goal: match the best answer to the question in front of it. It doesn't care how many followers you have. It isn't impressed by a viral post. It has no loyalty and no grudge. But when your signals are clear, when you're findable in the places it looks, and when you consistently deliver the thing people are searching for, the model starts to trust that recommending you is a safe bet. It refers you for the same reason that you refer your dentist because you're the answer it can stand behind.
This is the whole premise of AI visibility. It's a credibility check, run by a machine, at scale, all day long. And loan officers are being surfaced and skipped by that check right now, whether or not they know it's happening.
Thankfully, none of this requires you to become a tech expert. It does, however, require you to stop leaving evidence on the table.
Be findable, and be the same person everywhere. This is the one people fumble the most. You're John on your license, Jonathan on LinkedIn, and Johnny on your Facebook profile. To you, that's just how your name works. To an AI trying to confirm you're a real, consistent professional, it reads like three loosely related strangers. Pick one version and use it on your NMLS record, website, review profiles, and socials. Clear beats clever. A model can't vouch for a person it can't pin down.
Collect the third-party reviews. Anyone can write a blog post announcing their greatness. The machine knows that, which is exactly why your say-so carries almost no weight. A verified Google review from a real borrower who closed with you is worth ten paragraphs of self-description, because it's evidence someone outside your own head can confirm. Third-party proof is the strongest social currency you have. So, ask for the review every time. I promise, it's not vanity; it's the paper trail that tells an AI you're the best thing since cotton candy.
Be consistent, not occasionally brilliant. One newsletter that went semi-viral three years ago is a nice memory. The machine doesn't care about it. What builds trust is a steady body of you, spread across the internet, saying useful and consistent things over time. A blog here, a podcast appearance there, answers to the questions your borrowers are asking, all pointing back to the same expertise. Consistency is what separates a professional from a person who posted once. It's slower than a viral moment, and it's worth far more.
The instinct, when you hear all this, is to start writing for the robot. Cram in keywords, flatten your personality, and turn your website into a spec sheet a crawler will love.
Don't do this; that's the trap. And even though it might be tempting, your audience will read right through it. The content that wins the AI still has to win the person first. When a borrower finally clicks through from that ChatGPT recommendation to your page, they need to feel like they're reading something a real human wrote for another real human. If they land on machine-optimized mush, you've won the referral and lost the client in the same breath.
So the work is to write content that sounds unmistakably like you, rooted in your real expertise and your real way of explaining a rate lock or walking a nervous first-timer through underwriting, and then make sure that content is structured so the machines can find it, read it, and file you under 'trustworthy'. Human on the surface. Legible to the LLM underneath. Both, or it doesn't work.
Here's the good news: AI can't be charmed, and it can't be gamed for long. This means the loan officers who've been excellent at their jobs for years finally have the thing that matters most: substance. The only task left is making it findable.
You already have the expertise. Now there's a second audience that needs to be able to see it.
Get your name straight. Grow your review base. Show up consistently and sound like yourself when you do. That's how you’ll get referred in a world where half the traffic on the internet is a machine deciding whether it can trust you.
Katie Shive is a brand strategist who helps mortgage professionals and service-based business owners build the foundation their content and AI visibility depend on. After six years ghostwriting and building personal brands for top-producing loan officers, she launched KS Marketing to help the people behind the work finally get seen for it. Her philosophy is simple: brand before AI. The tools only work when there's a real, documented voice underneath them. She hosts the AI After Hours podcast and works with clients through her Brand Brief and done-with-you Brand Builder. Find her at katieshive.com.
