TL;DR: Most US home buyers hire the first agent they speak to, and a growing number now ask ChatGPT who that agent should be. AI assistants don't show ten blue links. They name three or four agencies. If your agency isn't one of those names, you lose the lead before the buyer ever calls, and it never shows up in your CRM. Google SEO still matters, but AI visibility is now a second front door with its own rules.
In the United States, most buyers don't shop around for an agent. According to the National Association of REALTORS® 2025 Profile of Home Buyers and Sellers, 67% of first-time buyers and 76% of repeat buyers hire the first agent they speak to.
Read that again. For most buyers, the first agent they talk to is the one they hire.
So the real competition in real estate isn't at the listing presentation or the showing. It happens before the buyer ever picks up the phone, at the moment they discover you.
And that moment is changing.
How US Buyers Have Always Found Agents
For decades, the path was predictable. NAR's 2025 data shows:
| How buyers found their agent | Share |
|---|---|
| Referral from a friend, neighbor or relative | 43% |
| Went back to an agent they'd used before | 15% |
| Asked about a property they saw online | 7% |
| Found the agent through a website | 6% |
The biggest source by far is someone they trust telling them a name.
Here's What's Changing
Buyers now have a new "someone they trust," and it answers instantly, at 11 PM, with no awkward conversation.
- A 2026 Bank of America Homebuyer Insights report found that 20% of prospective buyers have already used AI tools or chatbots for homebuying research. That rises to 28% of Millennials and 32% of Gen Z, the core of today's first-time buyers.
- A 2026 Cotality survey found that 55% of prospective and recent buyers use generative AI tools at least monthly.
ChatGPT is quickly becoming the new referral. When a family relocating to Austin, Charlotte or Phoenix asks:
"Who are the most trusted real estate agents in [city] for first-time buyers?"
ChatGPT doesn't show ten blue links. It gives three or four names, with reasons.
That's a referral, and for most buyers the first name they hear is the one they call.
The Use Case: When You're Not Part of the Discovery Layer
Picture a well-established brokerage in a mid-sized US city. It has:
- a #1 Google ranking for "best real estate agency [city]"
- hundreds of five-star reviews
- a polished website
A buyer who's new to the city asks ChatGPT for a recommendation. Three agencies are named, and this brokerage isn't one of them.
It didn't lose to a better agent. It lost because the AI couldn't confidently verify it:
- →Inconsistent listings. Its name appeared inconsistently across directories and listings. (See our guide to AI local citation cleanup and NAP correction.)
- →No answer-ready content. Its website had no structured FAQ content answering the questions buyers actually ask.
- →Pages AI couldn't read. Key pages loaded through JavaScript that AI crawlers struggle to read.
- →No verified entity records. It had no verified entity records, such as Wikidata, that confirm it's a real, established business.
Google SEO still matters. It's still where a huge share of buyers search. But ChatGPT is a second front door, and it has its own rules.
The Math Nobody Is Running
At typical US buyer-side commission rates, a single missed buyer lead is worth $5,000–$15,000, depending on the market and price point.
It's not a bad review or a slow month. It's one lead.
The part that should worry you is that this loss is invisible. You won't see it in your CRM or your Google Analytics. The buyer never called, so there's nothing to track. They called the agent ChatGPT named, and in 7 out of 10 cases, they hired them.
The Window Is Open, but Not for Long
AI answers aren't locked in yet. The agencies that build their AI visibility now (consistent entity signals, content AI can extract, pages AI crawlers can read, verified third-party mentions) will become the names AI keeps repeating.
Being early is a real advantage here. Once an AI system trusts a set of names for a city, those names become the default answer.
Google ranks real estate agents. ChatGPT recommends them. The agencies winning now are visible in both.
For most buyers, the first agent they talk to is the one they hire. If AI is now making that first introduction, AI visibility isn't a marketing experiment. It's part of your lead pipeline.
Are You in the Same Situation?
If you're a US real estate agent, broker or agency owner and you're not sure whether ChatGPT is recommending you in your market, let's find out before your next lead goes elsewhere. Send us your city and agency name, and we'll tell you who ChatGPT recommends in your market today and whether you're on that list.
📩 Write to team@sheinnovatesai.com →Frequently Asked Questions
Sources
- How Home Buyers and Sellers Find Their Agents in 2025 (NAR 2025 Profile data) – BAM
- NAR 2025 Profile of Home Buyers, Sellers Reveals Market Extremes – National Association of REALTORS®
- AI Becomes Early Step in Homebuying Journey (Bank of America and Cotality data) – National Association of REALTORS®
- Key Takeaways from NAR's 2025 Profile of Home Buyers and Sellers – Virginia REALTORS®
The brokerage scenario and infographic are illustrative examples. Commission figures are estimates based on typical US buyer-side rates, which have varied since the 2024 NAR settlement.