You hired the agency, you paid the retainer and four months in, you were reading LinkedIn drafts that sounded like every other loan officer in the country. Dang, that sucks.
So, you fired them and haven’t touched marketing since.
That’s a pattern I see often lately. One bad experience, sometimes two, and then a year or two of silence while the new guy in your space with a fraction of your experience starts getting opportunities you should be getting.
Then somebody mentions AI, and you feel the same skepticism rise up before the sentence is even finished.
You’re right to feel it. Most AI content is worse than the agency draft you fired someone over.
Here’s why, and here’s what you should consider instead.
Why the Agency Failed Before AI Was Even Part of the Conversation
The agency failed because they never asked who you actually are.
They most likely asked what you do and maybe a little about your team. But they probably never asked about the referral partner who sent you a deal because you called her back at 9pm on a Sunday. And they may have missed asking what you believe about this job that’s different from the guy down the street who closes the same loans you do.
So they wrote content from a template, and the content sounded like a template. Nineteen years of reputation, reduced to a posting schedule.
AI tools have the exact same failure mode, just faster. Ask ChatGPT to write you a LinkedIn post about mortgage rates and it will hand you back something that could belong to any of the four million loan officers currently licensed in this country because you gave it nothing to work with.
I’ve watched this exact scenario play out with clients who came to me after trying AI on their own. They pasted in a few thoughts, got back something polished and empty, and closed the tab. The tool did exactly what it was asked to do. It just wasn’t asked to do anything specific.
What Real AI Implementation Looks Like for a Top Producer
What does good AI implementation look like in a mortgage practice?
Good AI implementation starts with a documented foundation: your positioning, voice, and story, captured before any content gets generated. The AI then works from that foundation instead of a blank prompt, producing output rooted in specifics: your referral partner dynamics, your closing philosophy, the exact way you talk about your work.
Without that foundation, you’re not doing AI implementation. You’re doing prompt roulette. Every output is a slightly different guess at who you might be, and none of them land, because the model was never told.
This is the order of operations most people get backward. Foundation first, then the AI system, then the content. Skip the first step and every downstream piece inherits the same emptiness the agency draft had.
I spent six years ghostwriting for top producers before I ever touched an AI tool. I know the difference between content that sounds like a person and content that sounds like a category, because I built both, by hand, long before AI could do either one for me. That’s not a credential I’m reaching for. It’s the reason I trust myself to tell you when a tool is being used right and when it’s being used lazy.
The AI Gold Rush Is Producing More Noise Than Results
Right now there’s a course, an app, or a coach promising to solve your content problem with a prompt library. Most of them will be gone in eighteen months. Some already are.
The people running that gold rush aren’t wrong that AI is powerful. They’re wrong about what it’s powerful for. AI didn’t lower the bar for good content. It raised it for the people already doing the work, and it exposed everyone who wasn’t.
What’s the difference between AI-assisted and AI-generated content?
AI-generated content starts and ends inside the tool, with no human foundation underneath it, which is why it reads generic no matter how good the model is. AI-assisted content starts with a documented brand foundation, uses AI to produce faster from that foundation, and gets edited by a human who knows what “sounds like me” actually means.
Your name is attached to whatever goes out. That’s the part the noise ignores. A junior agency writer publishing something generic costs them a client relationship. You publishing something generic costs you nineteen years of a reputation you built one closing at a time.
Why Compliance Makes This More Complicated, Not Less
Most outside marketers, human or AI, get nervous around mortgage compliance and either go generic to stay safe or go sloppy because they don’t understand the regulatory texture of what you do. Neither one protects you.
A brand strategist who understands the difference between retail and wholesale, who knows what TRID means without you having to explain it, builds an AI system that produces content inside your guardrails instead of testing them. In a regulated industry, it’s essential
This is the same reason you don’t let a junior LO talk to a referral partner without training. The stakes are the stakes whether a human or a tool is generating the words.
What This Requires From You
You need a documented foundation, built once, that any AI tool you use afterward can work from.
That’s what a brand foundation does that a marketing agency retainer never did: it becomes a permanent asset. You own it. It doesn’t walk out the door if you switch tools, switch VAs, or switch strategists five years from now. It’s the thing that makes every future AI system you touch actually sound like you instead of like a template with your name on it.
The Next Decade Looks Different Whether You Build This or Not
You’ve hit the ceiling of personal production volume and the next chapter, whatever it looks like for you, requires a digital presence that travels beyond the relationships you’ve already built.
The newer agents in your market are checking LinkedIn before they call a top producer for a referral. Your real-world reputation hasn’t stopped mattering. It’s stopped being visible to the people who’ve never met you yet.
AI isn’t the reason to finally deal with this. It’s just the tool that makes it possible to deal with it without adding another job to your calendar. The foundation is still the foundation. It was true before AI existed and it’s still true now.
Frequently Asked Questions
Is AI-generated mortgage content getting flagged by referral partners?
Referral partners and clients increasingly notice when content feels templated or off-voice, even if they can’t articulate why. The fix isn’t avoiding AI. It’s giving the AI a real foundation to work from so the output sounds like the loan officer’s actual voice instead of a generic model default.
What are top-producing loan officers actually using AI for?
Top producers who use AI well use it for speed on the execution layer: drafting from an established voice, repurposing one piece of content across platforms, and cutting the blank-page time down. They don’t use it to invent a positioning they haven’t done the work to define.
Should I disclose when I use AI to write my content?
There’s no universal rule here, and it varies by brokerage policy and personal preference. What matters more than disclosure is whether the output is actually true to how you think and talk. If it sounds like you, edited and reviewed by you, most readers and partners care less about the tool than the result.
Comments +