The following is a guest post from Cyrus Vaghar
Just under a year ago, my business partner and I received an email: “Can you help me sell my home?”
Within twenty-four hours, we were touring the property. The seller was incredibly welcoming, but whenever we tried to pivot the conversation to our marketing strategy, our background, or our sales record, they simply weren’t interested. Finally, at the end of the tour, the seller turned to us and said, “Let’s get started.”
I paused. “Wait a minute. Do you want to learn more about us? Are you interviewing other agents?”
“No, I’ve made my decision.”
“Well, how did you find us?” I asked.
“Claude.”
My partner and I looked at each other. Who is Claude?
The Quiet Rise of the AI-Generated Lead
Claude is a Large Language Model (LLM). You’ve likely heard of its competitors, such as ChatGPT and Google’s Gemini. Made famous by the release of ChatGPT in late 2022, LLMs have grown so rapidly in popularity that Pew Research estimates roughly half of all Americans now use one.
A simple explanation of an LLM is that it answers questions, anything from how long to boil pasta to identifying the 10th President of the United States. It sounds gimmicky at first, right? But what happens when the questions evolve? What happens when a user asks, “How is the real estate market?” or, more importantly, “Who is the top real estate agent in Boston?”
That last question is the critical one. Anyone typing that prompt into an AI has two defining characteristics: high intent and capital to spend.
Because LLMs speak with authority, people trust them. When you are called into a home or a buyer consultation after being recommended by an LLM, you often bypass the competition completely. Think about the psychological progression here: the AI has already given this person reliable answers to important life questions, progressively scaling from things like simple cooking instructions to complex tax advice. By the time they ask it to recommend a real estate professional, they inherently trust its answer. When we walked through that seller’s home, we weren’t in a competitive interview; we were a shoo-in.
If You Think Consumers Aren’t Using AI, You’re Wrong
In my local market of Greater Boston, specifically Newton, Massachusetts, we have received a steady drip of AI-generated leads. It’s one of the primary reasons industry giants are pivoting; just look at Zillow’s recent launch of an AI-powered search tool.
The industry recognizes a stark reality: if a consumer uses Gemini or Claude to find a realtor, they are operating outside of the traditional real estate ecosystem. If they leave that ecosystem, traditional aggregators don’t get to assign the lead. And as every agent knows, if you don’t own the lead, you don’t have the business. HomeLight became a unicorn company simply by referring agents to consumers.
Today, consumers are using a mix of AI tools and specialized networks to bypass the noise and find experts directly. It is exactly why I built Find Agents Now. I wanted to create a streamlined, data-driven platform where buyers and sellers could get direct, vetted connections to top-tier agents nationwide without getting lost in the traditional lead-generation shuffle.
Real-World Applications for Buyers, Sellers, and Agents
The utility of LLMs extends far beyond simply finding and comparing real estate agents. When used correctly, AI is a powerful operational tool. Two examples:
Marketing and Copywriting: AI can generate compelling listing descriptions, brainstorm social media content, and draft targeted marketing emails in seconds.
Negotiation Strategy: As an agent, I’ve pasted long, complex email chains into an LLM to help me better understand the opposing party’s perspective. The AI often picks up on a subtle shift in tone or a single word I missed, revealing the true intent behind an email and allowing me to negotiate more effectively.
Consumers can apply this same concept. While you should always rely on your trusted local agent, AI allows buyers and sellers to instantly pull global, professional perspectives on hyper-local issues.
A Word of Caution: The Danger of Over-Reliance
LLMs are not without their limitations. Recently, a buyer fed a complex home inspection report into ChatGPT to estimate repair costs. The AI generated a massive, intimidating price range. The buyers spooked themselves so badly that their repair requests became totally unreasonable, and the deal died. Even when presented with grounded facts from a licensed, local professional tradesperson, the buyers couldn’t be convinced otherwise.
That is how deeply people believe in the magic of an LLM. It reinforces my original point: people implicitly trust AI recommendations, for better or worse.
Adapting to the New Landscape
Two years ago, LLMs were just neat parlor tricks used to fix typos in emails. Today, tens of millions of people trust them with the most important financial decisions of their lives. It is a fascinating, albeit slightly scary, phenomenon.
As real estate professionals, it is our job to be intimately familiar with this technology, if for no other reason than the fact that our clients as well as those on the other side of a deal, are actively relying on it. The good news? It isn’t difficult to learn. It is quite literally plug-and-play.
Test it. Prompt it. Have fun with it. While my team has had great success acquiring new business through AI, I’m no tech prodigy. There are massive opportunities sitting right in front of all of us. As this technology is adopted by the mainstream, the room for growth is only going to get bigger.
About the Author
Cyrus Vaghar is a licensed real estate agent (License #9557783) at Coldwell Banker in Newton, Massachusetts. He is a proud member of The Vaghar Team and the founder of Find Agents Now, a national referral network powered by Coldwell Banker Realty that connects clients with top-tier, data-vetted real estate professionals across the United States.

