South Florida · Mexico City The OGs of digital marketing · now fluent in AI ● AI systems in production
Journal · Real Estate Websites

How Home Buyers Find a Real Estate Agent in 2026: What NAR’s Survey Shows

Most home buyers still find their real estate agent through someone they know. In the National Association of REALTORS® 2026 survey, four in 10 buyers were referred by a friend, neighbor or relative, and 1% found their agent by searching with AI. A record 90% of buyers used an agent.

Here is what the survey says about how agents get chosen, what the 1% does and does not mean, and where that leaves an agent’s website and follow-up.

What did NAR publish about buyers and agents?

Its yearly survey of people who just bought or sold a home. The National Association of REALTORS® has run the Profile of Home Buyers and Sellers since 1981. The 2026 edition covers people who completed a transaction between July 2025 and June 2026, and NAR published its own summary of it on 5 November 2026.

The headline for anyone who markets an agent is that agent use went up, not down. NAR says 90% of buyers used a real estate agent or broker, the highest in four decades, and 92% of sellers did, an all-time high. Homes sold by the owner without an agent stayed at 5% of sellers, which NAR calls a historic low. Of those owners, 62% already knew the buyer.

The same summary shows who these clients are. First-time buyers were 22% of buyers, up from 21%. Their median age stayed at 40. The median repeat buyer is 62, and sellers had owned their home for a median of 12 years before selling. This is an older client, moving less often, than most agent marketing is written for.

How do buyers actually find their agent?

Through people they know. The detail on agent choice was reported by the trade publication Real Estate News on 11 November 2026, from the full report. Four in 10 buyers said they were referred to their agent by a friend, neighbor or relative. Among first-time buyers, 52% found their agent through a referral. On the selling side, 63% either used an agent they had worked with before or one who was referred to them.

What NAR’s 2026 survey foundShare
Buyers who used an agent or broker90%
Sellers who used an agent or broker92%
Buyers referred by a friend, neighbor or relative4 in 10
First-time buyers who found their agent by referral52%
Sellers who used a past agent or a referred one63%
Buyers who found their agent by searching with AI1%
Repeat buyers who interviewed only one agent77%
First-time buyers who interviewed only one agent67%

The first two rows are from NAR’s own summary of 5 November. The rest are from Real Estate News’ report on the full survey. HousingWire, writing about the same report, also gives the 1% figure for buyers who found their agent through artificial intelligence.

Why 1% from AI search is not the whole story

Because the question was where the buyer found the agent, and nothing more. A buyer who got a name from a neighbor and then asked ChatGPT or Google about that agent still counts as a referral. The survey figure says AI search is rarely the first place a name comes from. It does not say how many buyers used AI to research a neighborhood, a building or the agent they were about to call.

It is also a look backward. These buyers closed between July 2025 and June 2026, and most of them started looking months before that.

So we read the 1% two ways. An agent who is told that AI search will fill the pipeline this year should ask for evidence, because NAR’s buyers did not report that. And an agent who concludes that AI search can be ignored is reading more into the number than it holds. We have no data of our own on how many referred buyers check a name in an AI assistant first, and we have not seen NAR publish one.

Most buyers interview only one agent

This is the number we would put on the wall. Real Estate News reports that 77% of repeat buyers interviewed only one agent, and 67% of first-time buyers did the same. There is no shortlist for most clients. There is a name, a quick look, and a call.

That changes what a website and a Google Business Profile are for. They are rarely how a stranger discovers an agent. They are where a referred buyer goes to confirm that the name they were given is real, active and works in the right area. If that look goes well, the agent gets the only interview. If the site is out of date, or a search for the name shows nothing useful, the buyer asks the friend for a second name.

For sellers the pattern is similar. An agent’s reputation was the top consideration when choosing who to work with, cited by 31% of sellers. Reputation, for someone who has not met you, is what they can read about you in two minutes.

Where does the search for a home start?

Online, and usually before any agent is involved. NAR’s own description of the 2026 report says 46% of buyers took looking online for properties as their first step, while 19% contacted a real estate agent first.

Put that beside the referral numbers and the order becomes clear. The buyer starts with listings on a screen. At some point they need a person, and they ask someone they trust. Then they check the name. An agent’s marketing has to be present at the third step even if it had nothing to do with the first two.

It also explains why the agent still matters once hired. Real Estate News reports that agents were the most frequently cited source buyers used in the home search, at 86%. Help finding the right home was the most common thing buyers wanted from their agent, cited by 51%.

The gap between happy clients and repeat clients

Here the survey is uncomfortable. According to Real Estate News, 88% of buyers said they would definitely or probably use the same agent again or recommend them to others. Yet just 24% of repeat buyers said they used an agent they had worked with before.

NAR’s summary offers part of the reason. Sellers now stay in a home for a median of 12 years. Twelve years is long enough to lose a phone number, change email addresses and forget a surname. The client was happy. Nobody kept in touch.

For a business that runs on referrals and repeat clients, this is the most direct revenue point in the report. It is a follow-up problem, and it is solved with a contact list that is kept current and a reason to write a few times a year. That is the kind of work we set up in GoHighLevel systems for agents. We wrote about the limits of automated outreach in does AI lead generation work.

What sellers want from an agent’s marketing

Marketing and price, in that order of mention. NAR’s summary says sellers want an agent to help market the home, price it competitively and sell within a specific timeframe. Real Estate News gives the shares from the full report: marketing the home and pricing were each cited by 20% of sellers, and timing the sale by 17%.

A listing presentation that shows how a home will be marketed, with real examples, answers the first thing sellers say they want. HousingWire notes that 25% of sellers in the survey were selling for the first time, so a plain explanation of the steps is not wasted either.

Jessica Lautz, NAR’s deputy chief economist, said in comments reported by Real Estate News that record agent use “might seem counterintuitive” given how much information consumers can reach. That information, she said, “can be misleading without proper context.” Context is something an agent’s own pages can supply before the first call.

What this means for an agent’s website

Build it for the person who already has your name. That visitor wants to see who you are, where you work, what you have sold or leased lately and how to reach you. Most of that is plain text and a working phone link.

When we ran our own checks on the homepages of 63 Miami real estate websites on 25 September 2026, 38% had no business identity in their structured data and 21% had no tap-to-call or WhatsApp link. Only 13% had two or more headings written as a question. Those are our own readiness checks, not Google requirements, and they do not prove lost clients. They do show that the two-minute look often goes worse than it should.

The same basics serve AI answers. Google’s documentation on AI features says there are no additional requirements to appear in AI Overviews or AI Mode, and that the same foundational SEO practices apply. A site that states the agent’s name, license, areas and recent work clearly is doing the work for search, for AI answers and for the referred buyer at once. Our guide to AI for real estate agents covers the tools side.

What we would do with these numbers

Start with the clients you already have, then fix the first impression, then worry about new channels.

  1. Clean the past-client list. Every closed client, with a current phone and email, in one place.
  2. Write to them on a schedule. A market note for their building or neighborhood a few times a year is enough to stay the name they give.
  3. Search your own name in Google and in an AI assistant, as a referred buyer would, and read what comes back.
  4. Fix the profile and the homepage. Current photo, areas served, license details, recent work, a phone link at the top.
  5. Answer the questions buyers ask about your buildings and neighborhoods on your own pages, in plain language.

This is the order we follow in our real estate marketing work, and for agents here in South Florida in our Miami real estate marketing work. The survey is national. It does not break out Miami, international buyers or the luxury market, so treat it as the baseline and check it against your own closed deals. To see how your homepage reads to search and AI crawlers today, run the free AI Visibility Checker.

Want your business found on Google and in ChatGPT? Ciao Bella engagements start at $5,000 a month, no setup fee. See how we work.

Verified 17 November 2026 against NAR’s 5 November 2026 summary and report pages, Real Estate News, HousingWire, Google Search Central and our own September 2026 scan of 63 Miami real estate websites

Building on the web since 1999 · South Florida and Mexico City

Frequently asked questions

What share of home buyers used a real estate agent in 2026?

NAR’s 2026 Profile of Home Buyers and Sellers says 90% of buyers used a real estate agent or broker, the highest share in four decades. Among sellers it was 92%.

How many buyers found their real estate agent through AI search?

About 1%, according to trade press reports on NAR’s 2026 survey. The figure counts where buyers found their agent. It does not count buyers who used AI to research homes, neighborhoods or an agent they were referred to.

Do most home buyers interview more than one agent?

No. Real Estate News reports from the NAR survey that 77% of repeat buyers and 67% of first-time buyers interviewed only one agent.

Is the full NAR Profile of Home Buyers and Sellers free to read?

NAR publishes highlights and a summary of the top findings on its website. Its report page offers the full report as a download that is free for members.

Does an agent’s website need special optimization for Google’s AI answers?

Google says no. Its documentation on AI features states there are no additional requirements to appear in AI Overviews or AI Mode, and that the same foundational SEO practices apply.

Sources

  1. National Association of REALTORS® — Top 10 Takeaways from NAR’s 2026 Profile of Home Buyers and Sellers (5 November 2026) Agent use by buyers (90%) and sellers (92%), for-sale-by-owner share, first-time buyer share and age, repeat buyer age, 12-year tenure, what sellers want. Read 17 November 2026.
  2. National Association of REALTORS® — Highlights From the Profile of Home Buyers and Sellers Survey period (transactions between July 2025 and June 2026), published since 1981, full report free for members. Read 17 November 2026.
  3. National Association of REALTORS® — 2026 Profile of Home Buyers and Sellers (report listing) First step in the search: 46% looked online for properties, 19% contacted an agent. Read 17 November 2026.
  4. Real Estate News — Share of buyers, sellers using agents hits record high (11 November 2026) Trade press report on the full survey: referral shares, 1% by AI search, agents interviewed, would use again against did use again, seller priorities, Jessica Lautz’s comments.
  5. HousingWire — Buyers are aging and sellers are holding longer, what NAR 2026 found Trade press report on the same survey: 1% found their agent through artificial intelligence; 25% of sellers were selling for the first time.
  6. Google Search Central — AI features and your website Last updated 10 December 2025. No additional requirements to appear in AI Overviews or AI Mode. Read 17 November 2026.

Related reading: our scan of 63 Miami real estate websites, our guide to AI for real estate agents and our post on AI lead generation. For the service side, see real estate marketing and GoHighLevel systems.

Want to know what a referred buyer sees when they look you up?

Tell us your site — we’ll run our checks and send what we find.

Want to see what it costs first? Engagements start at $5,000 a month. See how we work.

We use your details only to reply and talk about working together. We never sell them. Privacy policy.

Free newsletter

Get our AI-search notes by email.

One short take a day: what changed in Google and ChatGPT, and what to do about it. Unsubscribe anytime.

No spam. We've been online since 1999. We know what spam looks like.

Hey, want to chat?
Scroll to Top