The most quoted document in SEO this month is Google's new "Optimizing your website for generative AI features on Google Search." It is being passed around LinkedIn as gospel. 

It also contains, in plain sight, an admission that Google thinks the entire real estate industry has been producing the wrong kind of content for fifteen years.

The guide defines "commodity content" (the type AI search will not reward) with one specific example: "7 Tips for First-Time Homebuyers." It defines "non-commodity content" with another: "Why We Waived the Inspection & Saved Money: A Look Inside the Sewer Line." 

Both examples are real estate. Of every vertical Google could have picked to illustrate the difference between the content AI rewards and the content it ignores, they picked ours. 

Twice.

Sit with that for a second. 

The largest search company in the world has now told real estate, in writing, that the dominant content pattern of the last decade no longer works. Google's mythbusting section will get a lot of attention this month. The fact that they picked our industry as the cautionary tale won't.

I think the cautionary tale is more important than the mythbusting. And I think most of the mythbusting, taken at face value, will get real estate brands cited less in 2026, not more.

What Google Said About AI Search (GEO, AEO…)

The guide makes three moves. It tells you the rules haven't changed: optimising for AI search is still SEO. It gives you five things to stop doing:

• don't bother with llms.txt files
• don't chunk content into tiny pieces
• don't rewrite your prose specifically for AI
• don't chase inauthentic mentions
• and don't over-rely on structured data for AI specifically

Then it tells you the one thing that does matter, which is creating unique, expert-led, non-commodity content.

Two of those five myths are right. Three of them are right only if Google is the only AI model that matters to you. And the non-commodity-content principle is the most important thing in the guide by a wide margin, except that the way it's being read in real estate is going to produce more commodity content, not less.

I want to take the guide seriously, because the parts that are right are very right. I also want to be specific about where it stops being useful for realtors, team leads and brokerages, which is roughly the moment Google's interests stop overlapping with yours.

"It's Just SEO" Is a Budget Argument, Not a Technical One

Google's framing is that "AEO" and "GEO" are not new disciplines… They're SEO under different names. 

This is the line being amplified loudest, and it has a structural cost that most brokerages haven't priced.

Inside a brokerage, the line item is the argument. When the marketing director walks into the budget meeting and says "AI search is the same thing as SEO," the next thing that happens is the SEO retainer absorbs the work. 

The agency keeps the same headcount, runs the same monthly content production, and now also handles citation tracking across five different AI platforms, agent profile optimisation inside Zillow's AI mode, ChatGPT app surfacing for Realtor.com, Perplexity index hygiene, and the entirely separate measurement problem of figuring out which of those is driving inbound business. Same money. Three times the surface area.

The skill set has diverged whether the title has or not. The traditional real estate SEO toolkit is keyword research, MLS feed hygiene, on-page optimisation, schema, local citations, and link building. The work of getting cited in AI search adds passage-level content engineering, entity disambiguation across third-party platforms, brand presence work on sites you don't own (Reddit, Wikipedia, local press, third-party guides), and the ongoing surveillance of how four or five different retrieval systems are interpreting your brand.

There is overlap with classic SEO. There is also vast new surface area that has never appeared on a brokerage SEO scope of work.

This matters because of how brokerage marketing budgets move. A team leader who allocates $4,000 a month for SEO is not going to triple it because Google said the work was the same. They will hand the existing retainer the new work, watch the existing rankings slip because the agency is now stretched across five platforms, fire the agency, and conclude that AI search "doesn't work." Meanwhile the team across town who treated AI search as a separate line, hired a separate specialist, and ran a different playbook will be on the citation list for every buyer query in their market.

This is not theoretical. In FlyDragon's 2026 ‘State of AI Search’ benchmark of 12,400 AI responses across 192 metros, 91% of US agents are effectively invisible in the AI search engines their buyers now use. Buyer-side searches starting in an AI engine rather than a traditional one have hit 61.3%. The agents who started AI search work in early 2025 hold 5.7× the citation share of agents who started the same work twelve months later, despite the latter group spending more.

My read on those numbers: the gap between the cited agents and the invisible ones is not about who is doing better SEO. It is about who treated AI search as a separate problem early enough to build a separate operating budget around it. The "it's just SEO" framing is not a clarification. It is the move that keeps the work uncompensated, which is great for Google and lousy for the agents and brokerages funding it. 

I think by Q4 2026 there is a hard split between brokerages with a line-item AI search budget and brokerages without one, and the gap will be visible in lead cost.

Google Wrote a Manual for Google, Not for ChatGPT, Perplexity, or Grok

Here is what makes the "still SEO" framing especially misleading in real estate specifically: Google's guide describes an AI model that has the lowest AI Overview trigger rate of any major US industry.

According to the 2026 Luxury Real Estate AI Discovery Report from Stepps and 5WPR, real estate triggers Google AI Overviews 0.14% of the time. Health triggers them at 13%. Finance at 4.2%. Retail at 2.1%. Real estate, the largest single asset class in the American economy, sits at one-thirtieth of retail. 

Whatever Google's AI Overviews are doing for other industries, they are not yet doing for ours at a meaningful scale.

Which is the first reason this document should not be read as advice for ChatGPT, Perplexity, Grok, or Claude. It is conveniently aimed at Google. The other large language models — the ones your buyers are increasingly typing into before they ever open Google — retrieve from entirely different substrates, with entirely different incentives, and Google's guide does not speak for any of them.

Here is what the retrieval map looks for different LLMs:

ChatGPT uses Bing as its primary search partner, supplemented by direct content deals with publishers. When a buyer asks ChatGPT for an agent in your market, the system is sending rewritten sub-queries to Bing's index and pulling from OpenAI's contracted publisher set. 

Optimising for Google does very little for you here. Optimizing for Bing (the engine the SEO industry has ignored for fifteen years, including me) matters again, suddenly, for a different reason than it ever mattered before.

Perplexity uses a blend of Bing, third-party Google data, and a proprietary index it now exposes as a developer API. 

The substrate is partly Google-adjacent and partly its own crawler stack. Citations on Perplexity routinely surface content that does not rank in either Google or Bing's top 10, because Perplexity's reranking treats source diversity as a feature.

Grok retrieves from two distinct layers: a real-time web search with no public dependency on Google's index, and a live feed of public X posts. 

For real estate, the X layer is the part most agencies are not thinking about — agents publicly active on X with named transactions and identifiable thread history have a substrate advantage on Grok that no amount of website SEO will replicate.

Claude uses Brave Search as its primary live retrieval layer. Brave's index is independent of both Google and Bing, with its own crawler and its own ranking signals. Content that ranks well in Google can be invisible to Brave, and therefore invisible to Claude.

Five LLMs. Four different retrieval substrates. 

Optimising for Google's index is optimising for exactly one of them, in the vertical where Google's AI surface fires least often.

Stanford's 2026 research captured the directional shift directly: in July 2025, 76% of AI-cited URLs ranked in the organic top 10. By February 2026, only 38% did. The rest came from positions 11–100 and beyond. 

AI is finding authoritative content from across the web, not just the top of Google's index — and "the web" here means four different indexes operated by four different companies with four different definitions of authoritative.

A guide that describes one of those indexes, written by the operator of that index, telling you that index is the same thing as the other four, is not neutral guidance. It is product positioning. 

Read it as that.

Chunking Is Wrong As a Slogan. It's Right As a Format.

Google's most-quoted line in the mythbusting section is that you don't need to chunk content. The implication, repeated approvingly across SEO LinkedIn this month, is that you should write naturally for humans and trust the systems to figure out the rest.

The technical reality is different, and it matters more in real estate than in almost any other vertical. 

RAG (Retrieval Augmented Generation) systems do not retrieve pages. They retrieve passages. A passage is a chunk of text. A few sentences, a paragraph, a section that the retrieval system scores against the user's search.

The chunking happens regardless of whether you optimised for it. The question is whether your content survives the chunking process with meaning intact, or whether it shatters into incoherent fragments that lose retrieval comparisons to your competitor's tighter content.

Bing has been publishing the opposite of Google's position on this for months. 

In their May 2026 update on the evolving role of the index, Bing's team stated plainly that "chunking and transformations must preserve meaning and claims used in the answer." The unit of value, they said, is shifting from documents to groundable information — discrete, supportable facts with clear provenance.

The form factor where this matters most in real estate is the neighborhood guide. The neighborhood guide is the workhorse asset of every brokerage content marketing program in the country. 

It’s also where most real estate content fails passage-level retrieval, because the standard format is a single page that tries to cover ten neighborhoods with three sentences each, then loops the same generic structure for the next city. When a buyer asks ChatGPT "Is Mueller a good neighborhood for families with kids and a dog?" The retrieval system pulls passages, scores them on semantic matches, and selects the passage that most tightly answers the specific question. 

A page covering ten neighborhoods at the same density loses every time to a page covering Mueller specifically.

Google is right that you shouldn't break content into 50-word answer blocks. That advice was always dumb, and it produces content humans don't want to read. But the deeper principle — that passages are the unit of retrieval, and that one tight, self-contained passage beats three loose ones every time — is the most important format change in real estate content since the move to mobile. 

Calling it "chunking" makes it sound like a tactic. 

Calling it passage-level structure makes it sound like what it is: a basic constraint of how the systems work.

"Don't Rewrite for AI" Is the Lie That Will Cost Agents the Most

Google says you don't need to write differently for AI. AI systems, the guide claims, understand synonyms and general meaning. Just write naturally and trust the system.

I disagree with this for one specific reason in real estate, and I'd put real money on it: most agent content fails AI retrieval not because of word choice, but because of entity ambiguity.

Read a hundred agent blogs at random. Count how many times you see the phrase "great schools" without a named school district. "The luxury market" without a defined price floor. "This neighborhood" without a named neighborhood. "We've helped many buyers" without a named transaction, ZIP code, or year. 

The content is technically natural language. It is also, from a retrieval system's perspective, almost entirely ungrounded (and identical).

AI retrieval is entity-anchored. The query is parsed into entities. The candidate passages are scored on entity match. A blog post that says "I help families find homes in great school districts in the Austin area" has roughly three retrievable entities: families, Austin, school districts. 

A blog post that says "Last month I closed a $785,000 four-bedroom in Mueller for a family relocating from San Jose, where the buyer specifically needed walking distance to Maplewood Elementary" has fifteen. 

When the retrieval system is choosing which passage to surface for "Mueller homes near good elementary schools," the second post wins by a margin that isn't even close.

This is what Google's "don't rewrite for AI" framing obscures. 

You don't need to write in a special AI dialect. You absolutely do need to write with named entities, specific transactions, and dollar amounts woven into prose at a density that most real estate content has never approached. 

Three principles I would tell every internal writer or editor at any brokerage are:

• Name the entities every paragraph. Specific neighborhood, specific school, specific price band, specific year. Generic noun phrases are retrieval poison.
• Anchor in transactions. First-person specifics from real closings beat market commentary every time. "I closed three Mueller transactions in Q1 2026" outperforms "Mueller is a popular neighborhood for families."
• Write in paragraphs, not bullet lists. Bullets fragment the semantic context the retrieval system needs to score the passage well. The "chunking" advice produced a wave of bullet-hostage real estate content that retrieves worse than the prose it replaced.

Google's guide will not tell you any of this, because the guide's authors are not building real estate agent websites for a living. 

The retrieval math is the retrieval math regardless.

The One Thing Google Got Right, and Where Real Estate Is Misreading It

Non-commodity content is the most important sentence in the guide. AI systems are trained on the open web. Content that just restates what's on the open web does not survive synthesis. The systems are not stupid; they will pick the source that adds something the other ten sources can't.

Where the real estate industry is misreading this — including most of the AI SEO advice currently being sold to agents — is in thinking "non-commodity" means "long" or "comprehensive" or "with original photos." 

It doesn't. 

A 4,000-word "Ultimate Guide to Buying a Home in Austin" with original drone photography is commodity content if everything it says could have been written by anyone with an MLS feed and a search engine. Length is not non-commodity. Imagery is not non-commodity. Even original data is only sometimes non-commodity, depending on whether the data is publicly derivable.

Non-commodity for an agent means content that could not have been produced by anyone who hasn't done the transaction. Three tests I use to score whether a piece of agent content earns the non-commodity label:

• The transaction test. Is this content grounded in a real closing? Not a hypothetical "if you're thinking of buying in Mueller" — a specific deal with a real address, real contingencies, real seller negotiation, real outcome.
• The named-entity density test. Does every paragraph name at least one specific entity — neighborhood, school, address, price, lender, inspector, builder, agent, year? If most paragraphs are abstract, the content fails this test even if the topic is specific.
• The replication test. Could a generic AI produce this from publicly available data alone? If the answer is yes, you have produced commodity content with extra steps.

Google's own contrast example — "Why We Waived the Inspection & Saved Money: A Look Inside the Sewer Line" — passes all three. It is grounded in a specific transaction. It is dense with named entities (sewer line, inspection, dollar saved). 

It cannot be replicated from public data because the decision logic only exists inside the head of the person who made it.

Most agent content marketing programs running today do not produce content that passes a single one of those tests, much less all three. The work to fix that is much harder than the work most brokerages are commissioning their content agencies to do, which is why the brokerages getting cited are the brokerages who pivoted their editorial brief in early 2025, not the ones still ordering "10 things to know about the Austin market in May."

So What Does a Real Estate Brand Do Now

I don't want to end on a checklist, because the work is not a checklist. The work is a reframing.

Treat AI search as five interfaces, not one. Build a separate operating budget for it, with a separate scope from your existing SEO retainer. 

Use Google's own "non-commodity content" definition as a hiring filter for whoever writes your blog, not a tagline for your homepage. Stop spending on content production into a substrate that doesn't retrieve, because the retrieval substrate — entities, structure, brand presence across third-party sites — is doing the work the content is taking credit for.

The hardest part of this is the part Google's guide cannot help you with at all. The retrieval systems are now plural. The optimization surface is now plural. The opinions about what works are now plural. The era of one document, one platform, one set of best practices is over for real estate, and the brokerages who keep operating as if it isn't are funding the gap their competitors will exploit.

Google's guide is one opinion. It is the opinion of the company with the most to lose from a multi-platform world. 

Read it. Take what is useful. Apply it where it applies.

The buyer asking ChatGPT about your market this week doesn't know Google wrote a guide. They want a name. Either yours is in the answer or someone else's is. The next twelve months decide which.

I run FlyDragon’s AI SEO strategy and AI infrastructure.

We work with 100+ real estate agents across the US and Canada, with clients on something like 20 different website platforms. Every sales call opens on the same question: is my website AI-friendly? 

For six months I've been answering it anecdotally but enough agents asked that I decided to run the audit properly: five live sites from each of the 14 most-deployed real estate platforms, 65 sites total, scored against the 2026 standards for what makes a page citable by AI search engines. 

This research ranks what I found, tells you what I think it means, and gives you the questions you can (and should) take to your vendor.

Why AI visibility is a different problem than SEO for websites

Traditional SEO was about ranking on page one of Google. Answer Engine Optimization (AEO) and Generative Engine Optimization (GEO) are about being the source an AI model cites when it writes, ranks and prioritizes an answer by itself.

When a buyer in Nashville opens ChatGPT and types who is the best real estate agent in East Nashville under $700k, the answer has nothing to do with yard signs or Zillow reviews. It depends on which agent's website an AI system can read, parse, and trust enough to recommend.

Being citable in AI means five things have to be true about the page (technical version):

1. Core content is present in plain HTML. OpenAI's OAI-SearchBot and Anthropic's Claude-SearchBot don't render JavaScript the way Googlebot does. If your bio, service area, and neighborhood guide isn't present when the AI search bots come to your page, nothing is seen.
2. Structured data is implemented in JSON-LD. RealEstateAgent, LocalBusiness, Person — these are the schema types that tell an AI system who you are rather than making it guess from unstructured prose.
3. AI search crawlers are allowed at the robots.txt layer and the CDN layer. These are two different checks — and 27% of B2B sites fail the CDN one without knowing (ziptie.dev via Mersel AI, February 2026). GPTBot and ClaudeBot are training crawlers and don't determine citation. OAI-SearchBot, Claude-SearchBot, and PerplexityBot do — plus the real-time user fetchers ChatGPT-User, Claude-User, and Perplexity-User. Most platform teams I've talked to don't know the distinction exists.
4. Core Web Vitals are in the green. LCP under 2.5s, INP under 200ms, CLS under 0.1. Slow pages get deprioritized by AI Overviews the same way they get deprioritized by classic ranking.
5. Entity signals are consistent. NAP data, mentions across the web, and disambiguation ("Sarah Chen, Douglas Elliman, Upper East Side") that let the AI resolve you to one specific professional.

And now, so you can understand what each thing means:

1. Your content needs to be visible on the page to search robots. If the language your website is built on isn't robot-friendly, it won't be seen (and thus, won't be ranked).
2. Structured data is robot language to help understand what the page is about in a structured way instead of having to read the headers and all the content (which burns tokens).
3. Robots.txt is a file most real estate website providers will have by default. However, the wrong instructions here can inadvertently block AI search engines from accessing your content.
4. Core web vitals means how quickly your website loads. It's a metric that Google introduced some years ago so we focused on more than just site speed.
5. Entity = your brand, where it appears and the information about your brand is consistent.

Most real estate platforms were built before any of this mattered. A few have caught up. Most are still selling IDX speed and CRM automation as if it's 2021.

How me and the team ran the website audits

I pulled five live agent sites from each of the 14 most-deployed platforms in the US market — 65 sites — and ran each through a headless crawler mimicking how an AI search system fetches a page. 

I checked the raw HTML that arrives pre-JavaScript, looked for JSON-LD schema in the document, tested robots.txt directives against named AI search user agents, and recorded the response code. 

Every site was scored on five criteria: 

• HTML readability
• Schema markup
• AI search crawler access
• JavaScript rendering risk
• Composite readiness

Three notes on methodology before the scorecard. A 403 response isn't automatically AI-hostile — some CDNs return 403 to any request they don't recognize, and named AI search crawlers with verified user agents may pass through the same middleware.

But a 403 against my crawler is at minimum unverified access for AI crawlers, and unverified is itself the problem. Second, schema judgments came from raw page source, not platform marketing claims. Third, the Bold Trail finding is a robots.txt-level block — categorical, unambiguous, not up for interpretation.

The best real estate agent websites ranked for AI search

Here's how the 14 platforms ranked, best to worst*.

Platform	HTML Readable	Schema	AI Bot Access	JS Risk	Overall
AgentFire	Yes	Active	Open	Low	Best
Sierra Interactive	403 blocked	Confirmed	Unverified	Medium	Mixed
Agent Image	Yes	WP-dependent	Open	Low	Mixed
Boomtown	Yes	Not visible	Open	Medium	Mixed
Brivity	Partial	Not visible	Open	Medium	Mixed
Luxury Presence	Partial	Not visible	Open	Med-High	Mixed
Curaytor	403 blocked	Claims AEO	Unverified	Unknown	Mixed
CINC	Yes	Not visible	Open	Low	Mixed
Ylopo (site varies)	Depends	Depends	Depends	Varies	Mixed
Lofty	403 blocked	Unknown	Unverified	Unknown	Poor
Market Leader	403 blocked	Unknown	Unverified	Unknown	Poor
Real Geeks	403 blocked	Unknown	Unverified	Unknown	Poor
Bold Trail	robots.txt block	Unknown	Blocked	N/A	Worst
Join Bulletproof	robots.txt block	Unknown	Blocked	N/A	Worst

Eight of thirteen platforms are somewhere in the middle — readable but under-built for AI citation. Four are unverified on access and haven't publicly addressed it. One is actively opted out. And one is genuinely ahead of the pack.

Which is the best real estate website for AI search?

AgentFire — the only platform shipping features for AI

AgentFire has a structural advantage: raw HTML is served to crawlers without waiting on a client-side render, plugin ecosystems cover schema cleanly, and robots.txt is agent-configurable rather than vendor-locked. 

What tipped AgentFire to the top of the scorecard wasn't the WordPress-esque base. In March 2026 they shipped a schema update adding Article schema for blog content, Product schema for listing pages, and VideoObject for video widgets — explicitly framed as Google and AI platform readiness. Of the 14 platforms I audited, they're the only one publicly treating AI citation as a product feature rather than a side effect.

AgentFire is where I'd send a new agent tomorrow for AI visibility.

Sierra Interactive

Sierra's documentation confirms schema markup is applied automatically to listing pages, agent profiles, and contact pages. 

Their reputation for traditional SEO is earned — the Sierra sites I've worked with rank genuinely well on Google. And Sierra has a great team of people who genuinely care about the experience they give agents.

But every Sierra site I crawled returned 403, and the collateral-damage problem applies: their aggressive bot protection may be catching AI crawlers in a net it was built to catch scrapers with. Sierra's support team is good. 

If you're on Sierra, send them the question list below and get the answer in writing.

Agent Image

Agent Image also builds on WordPress, giving it the same structural access as AgentFire. 

The sites I reviewed loaded cleanly, with strong hyperlocal content organization (floor plans grouped by community, neighborhood landing pages) and raw HTML without JavaScript dependencies. 

Schema depends on the individual build, because Agent Image is a custom platform rather than a templated one. Of the middle-tier platforms, Agent Image has the most workable foundation — an agent on Agent Image who layers in schema and entity work will beat an agent on AgentFire who does nothing.

Boomtown

Boomtown sites loaded clean HTML over Fastly's CDN. Property search is heavily JavaScript-dependent, putting the bulk of the lead-gen surface behind a client-side render — meaning to an AI crawler, the listings effectively don't exist. No schema visible. 

Boomtown is a lead-gen platform first; AI citation isn't a documented product feature and the architecture reflects that.

Brivity

Brivity sites loaded core page content in raw HTML (good) but showed JavaScript-dependent navigation, authentication flows, and IDX integration that relied on client-side rendering (same problem as Boomtown).

From first-hand experience, working with agents who use Brivity, I can tell you that performance is hindered by the platform. We’ve deployed a similar approach to AI search for most of our 100+ agents and the platform which gives us the greatest issues?

Brivity.

Luxury Presence

Luxury Presence prioritizes bespoke design and brand-forward aesthetics, and the agent sites on the platform are the prettiest in the sample. Sites loaded with partial content in HTML but heavy JavaScript dependency on Cloudinary-loaded assets. 

The bigger problem sits in property search, which runs on JavaScript frameworks AI search crawlers don't render. No schema on the listing pages I tested. 

The brand work is excellent. The AI citation readiness is not. Despite claiming they offer ‘AI SEO’ as a service, I highly suspect Luxury Presence is simply using it as an add-on, rather than a core offering.

Curaytor

Curaytor's platform page says their sites are "SEO and AEO ready at launch" — the most forward-facing AI visibility language of any platform in this audit. 

Yet every Curaytor site I crawled returned 403 (this means the site is blocked from AI search engines).

The bot-protection middleware that blocked me may or may not whitelist named AI search crawlers; I have no way to tell from the outside. 

I'd put the probability that Curaytor's middleware correctly allows OAI-SearchBot, Claude-SearchBot, and PerplexityBot at 50/50, and I'd want to see server logs before I moved an agent onto the platform on that claim alone. 

The intent is there. The proof isn't public. And, from what I’ve seen with traffic, most of their offerings do not support AI search.

CINC

CINC sites loaded cleanly with neighborhood guide content in plain HTML. Then I checked the page titles. "Home Page." Multiple sites. Multiple states. Basic on-page SEO fundamentals that should have been solved in 2014 aren't. No visible schema markup. CINC is lead-gen-first, and SEO and AI visibility are clearly not where product engineering attention has been spent.

Ylopo

Worth clarifying because agents shop for Ylopo as if it were a website builder. It isn't. Ylopo is an advertising, IDX search, and CRM overlay that sits on top of a Squarespace site — every Ylopo client I've ever audited has been on Squarespace underneath. Ylopo's AI visibility ceiling is whatever Squarespace delivers (readable HTML, no native real estate schema, no llms.txt, limited structured data). Ylopo can't compensate for the underlying platform's gaps because it isn't touching that layer.

Lofty

Lofty (formerly Chime) returned 403 across every site I tested. Whether this affects AI crawler access specifically is unverified, and the unverified status is the problem. If you're on Lofty, you have no way of knowing whether your site is reachable by OAI-SearchBot or Claude-SearchBot. Product investment has been in CRM and workflow automation; organic visibility has never been the headline.

Market Leader

Market Leader (CoStar-owned) returned 403 across the sample. Built as a lead-generation machine — pay-per-lead pipes, turnkey conversion — AI visibility isn't a documented feature. I'd put probability of passing a competent AI-citation audit at under 30% without custom intervention.

Real Geeks

Real Geeks sites returned 403. Popular budget-friendly option with decent IDX integration and lead capture, but its SEO flexibility is well-documented as weaker than WordPress. AI visibility status unverified. If you're a new agent on a budget and Real Geeks is the cheapest IDX option, fine — but don't expect the platform to do AI visibility work for you.

Bold Trail

Bold Trail is the worst result in the audit and the gap is wide. Sites returned ROBOTS_DISALLOWED rather than 403 — a robots.txt directive instructing every crawler that reads robots.txt to stay away. OAI-SearchBot, Claude-SearchBot, PerplexityBot, Perplexity-User, ChatGPT-User, Claude-User — all of them honor robots.txt. 

An agent on Bold Trail has been opted out of AI visibility at the infrastructure level, by a policy they didn't write, which they probably don't know exists. I'd put this at 95%+ probability that no Bold Trail agent reading this has ever been told. 

Which makes it the biggest finding in the audit — a category-level exclusion rather than an execution gap.

Join BulletProof

Join BulletProof (joinbulletproof.com) is a Texas-based done-for-you marketing service, not a website platform. They sell "AI-ready schema, voice search optimization, and AEO" alongside a Google local SEO package. I'm including them here because several agents I've spoken to this year were weighing them against us specifically for AI visibility work.

The AI SEO layer is the part I can assess directly. Stripped of the terminology, BulletProof's package is Google Business Profile management, directory submissions, 15+ monthly templated posts, Local Service Ads guidance, and a CRM. Directory citations and GBP posts don't produce AI citations.

The local SEO work may well produce leads. GBP is still a real channel and LSAs still convert in some markets.

I wouldn't recommend them for AI visibility specifically, and the lack of any independent review record on a 12-month contract is a structural risk I'd flag to any agent considering the spend.

7 questions to ask your website provider

I've watched twelve agents in the last year have the AI visibility conversation with their platform's support team and I'll save you forty minutes. The seven questions that matter — copy them, send them, get the answers in writing:

• does our robots.txt explicitly allow all AI search bots to access our website?
• does any bot-protection middleware (Cloudflare, Imperva, AWS WAF) whitelist named AI search user agents at the CDN layer
• is RealEstateAgent schema applied to agent pages automatically, in JSON-LD, and does it validate in Google's Rich Results Test
• is core page content (bio, service areas, neighborhood guides) present in raw HTML pre-JavaScript
• is the property search on the main domain or a subdomain
• what are our current LCP, INP, and CLS numbers in Search Console

That's six.

The seventh is the one that matters most: who internally owns AI visibility on your platform, and can I talk to them?

If the answer is "that's not a role we have," you already have your answer about the platform. If the answer names a person and they respond within a week, that's a real signal about where the product is heading.

If you're unsure, book a call with us. It's no obligaiton. We'll tell you what's going on.

llms.txt is NOT important for AI visibility

One quick detour, because every AI-visibility piece aimed at real estate agents in 2026 is recommending llms.txt.

I don't.

Google's John Mueller and Gary Illyes have both publicly said no Google AI system consults llms.txt.

Multiple large-domain analyses have found zero statistically significant correlation between llms.txt presence and AI citation frequency. 784+ sites have implemented it, and server logs from the ones publishing their data show AI crawlers aren't fetching the file at meaningful rates. The proposal might become a standard in 2027 or 2028. It isn't one now. If your platform ships llms.txt support, fine — take it.

If you're paying a consultant to implement llms.txt, you're paying for cargo cult.

Your website platform is the floor

The platform is the foundation of AI visibility, nothing more.

You still have to build the building on top of it.

Even AgentFire has gaps. And a Nashville agent on CINC who does the right schema, the right entity-building, the right content cadence will outrank a Nashville agent on AgentFire who does none of that.

Reach is the first hurdle. Citation is the second, and the second is harder. AI citations favor sources that are authoritative (mentioned and linked across the web, not only on their own domain), entity-rich (clearly identified as a specific person in a specific market with specific expertise), fresh (updated with market-relevant content), and structured (schema and clean HTML so machines parse without guessing).

This is the work I do.

FlyDragon layers AI visibility infrastructure on top of whatever platform an agent is on — building the content, the entity signals, the citation trail that gets you named by ChatGPT and Perplexity when a buyer in your market asks. After 100+ agent partnerships across 20 platforms I can tell you the pattern clearly: the platform is one variable of maybe twelve.

The 403 blocks and the Bold Trail finding represent a systemic failure across an industry that hasn't caught up to how search works in 2026. The agents who win the next five years are the ones who figure out that being findable by humans and being citable by AI are two different problems with two different solutions — and who stop confusing the first for the second.

Most of the industry is still doing the first and calling it the second.

*This review is not personal or subjective to any of the real estate website providers mentioned. The information is true as of 20th April 2026 based on the live audits we conducted of each platform. These opinions are not swayed or influenced.

The pricing range for a real estate agent website in 2026 is wider than most realtors realize. 

Costs vary from $100/month up to $2,000/month leaves a lot to the imagination. And it’s never truly clear the benefit you’d get from investing in a website.I've consulted with agents at every tier over the past decade, and the pattern holds no matter how much they paid: the price of the site doesn't predict whether anyone finds it.

In my opinion real estate website costs are the wrong conversation to be having. 

The better question is: where the money goes inside that cost. Most agents invert the ratio that matters — they spend 80% on the build and 0–20% on the content that makes AI search models cite them. 

The top 10% of agent sites I audit have that ratio flipped. They cost less upfront and rank more.

Average prices of real estate website design

The market has roughly four tiers of pricing.

Template-based DIY tools (Placester, AgentFire, Squarespace with a Showcase IDX plugin, Wix) run $100 to $300 a month and get you a site that looks clean and does nothing else. 

All-in-one platforms (BoldTrail, BoomTown, CINC, Chime, Real Geeks) run $449 to $1,800 a month and bundle a site, a CRM, PPC tools, and lead capture into one monthly fee. 

Custom WordPress builds on Oxygen with plugins, Bricks, or Elementor run $8,000 to $10,000 up front with website hosting and maintenance on top (you own the site). Enterprise custom builds for top-producer teams run $40,000+ and include bespoke lead routing, iHomefinder or IDX Broker integration, and custom automation work.

Website development sounds fancy and impressive on paper.

But despite how great some of these websites look, 95% of them drive no traffic. I did the research in January 2026. Less than 5% of any real estate website was generating 100 or more visits.

No platform rep will tell you this. 

Three of those four tiers are designed for a market AI search models have already started to ignore. The IDX-heavy, listing-page-driven, forced-registration sites don't rank in ChatGPT, Perplexity, Claude, or Google AI Overviews. 

In fact, they barely rank in traditional search engines. But isn’t that what you’re paying for? Inbound leads?

Mistake 1: Paying monthly forever for a site you don't own

The all-in-one platform pitch is seductive. One monthly fee, everything included, nothing to manage. Inside a year you've spent $12,000 to $22,000 and you don’t have an asset, you have a liability.

As soon as you stop paying and the site disappears. Every URL, every page, every backlink you've built into it — gone. I’ve seen this at least 100 times in the last 5 years in the real estate industry. It’s a terrible way that most website providers trap agents.

The lead database stays with the platform in most cases if you move brokerages. You're renting, and the landlord holds the keys.

The deeper problem for AI search specifically: these platforms produce near-identical sites across thousands of agents. Same templates, same page structures, same thin neighborhood pages, same forced-registration gates on listing content. 

inboundREM's 2026 review of kvCORE said it directly — "SEO is non-existent. There is no organic means for Google to find your site." The quote comes from a review that broadly recommends the platform, which makes the admission sharper. Models like ChatGPT and Perplexity cite sources based on topical authority, entity recognition, and content uniqueness. 

A site that looks like 20,000 other sites has none of those.

My bet by the end of 2026: the all-in-one platform market loses 15–25% of its customer base to WordPress-based or headless CMS (content management system) alternatives, and the brokerages selling these platforms start pivoting to "hybrid" offerings where the site component is modular.

But no real estate website provider will ever be the best AI SEO agency. They will simply sell an ‘add-on’ which AI search isn’t so, keep that in mind.

Mistake 2: Spending 80% on design and development, 0% on content

This is the most common mistake I see, and it's the most expensive one in the long run. An agent pays a premium for a WordPress build with a designer, a logo, brand guidelines, and a styled IDX integration. 

The site launches looking beautiful and sits there ranking for nothing, because the agent never budgeted for content. No neighborhood guides. No market reports. No buyer-side or seller-side resources. No expertise pages. Nothing AI search can cite.

Design doesn't rank. Structure ranks. Content ranks. Entity signals rank. 

A professional website with 12 pages of thin content is worth less, in AI search terms, than an ugly site with 80 pages of specific, named, sourced, opinionated content that belongs to the agent who wrote it.

Imagine that for a second. A site that someone spent less than $1,000 on is converting more listings from AI search than yours which had a total cost of $10,000 to build.

Rule of thumb from the audits I run: if an agent has $10,000 to invest, I’d spend $2,000 on the website build and hosting. $5,000 on content. $3,000 on press releases and link building.

Why?

The agents who do this end up with sites AI search cites. This, in turn, gives you income. That is the point of a website in real estate: to convert.

Mistake 3: Paying the website cost and hoping for SEO wins

I've lost count of how many agents have forwarded me proposals from SEO agencies promising rankings in 90 days for $1,500 to $5,000 a month. 

The playbooks in those proposals are almost always identical. 

Keyword-optimized pages for 10 neighborhood targets. A monthly blog post. Generic link-building outreach. Google Business Profile optimization. 

It's the 2019 SEO playbook, and it hasn't ranked a real estate agent site in competitive markets for about three years.

AI search rewards different signals. 

• Brand mentions frequency across the web. 
• Entity recognition: is this agent known for this topic, in this market, with these named clients? 
• Source diversity: is the agent cited in places beyond their own site? 
• Content depth: does the site answer the specific questions buyers and sellers are typing into ChatGPT, Perplexity, and Google AI Overviews? 

The 2019 playbook optimizes for none of these. The agencies still selling it either haven't updated their product or are betting most of their clients won't notice. And most real estate website providers are still using this playbook.

How do I know this works? Well, take a look at Chris Speicher’s traffic. This is after working with us for the last 12 months.

What much should a real estate agent invest in a website?

My professional opinion is that you don’t need a gorgeous website if it doesn’t convert for you.

If you’re considering between:

1. Investing $10,000 into a website that looks great and
2. Investing $10,000 into generating listings from a website that works

You go with option b.

You should invest no more than $2,000 into a real estate website. Unless you need a custom CMS, with huge hosting capacity, in a market that demands ‘luxury’ then spend your money on something that will give you a return on investment: AI search.

Your biggest website cost should be content. We’ve helped realtors secure upward of $30k in GCI from AI search in a matter of weeks thanks to the content we’ve published. 

It works.

How to choose between real estate website providers 

An agency is worth the premium if they've ranked real estate agent sites in competitive AI search queries. 

A freelancer is the better call for 80% of agents — lower overhead, direct access, same quality if you pick the right one. A DIY template is the right call for new agents with a small budget — use Placester, AgentFire, or Real Geeks and spend the real money on content and your sphere of influence.

Before you sign a $25,000 proposal, ask the agency to name the last three times one of their client sites appeared in a ChatGPT response, a Perplexity citation, or a Google AI Overview. 

One question agents don't ask enough: how long until the site ranks? 

And, to be honest, these website providers shouldn’t have an answer. Because they aren’t an AI SEO company. It’s like me asking Walmart ‘when will I lose weight?’ just because they’ve sold me healthy food.

The smoke and mirror element of real estate websites is this: they look great, they’re built for traffic but they’re not made to generate the traffic or build an online presence.

You could wait 3 years, spend thousands of dollars on the best looking website in your market, and you’d still get no listings.

Be smart in 2026.

Invest in a website if it makes sense but, without any hesitation, you should be investing in an asset that generates come list me phone calls.

Become the agent AI recommends as #1 and you’ll never have to worry about website design or builds ever again.

In January 2026, a Delta Media survey covered by Inman found that 97% of brokerage leaders report their agents actively using AI. 

A few weeks later, Ascendix put the daily-use figure at 87% across brokerages and agents. Set those numbers next to Gartner's forecast that 40% of enterprise applications will include task-specific AI agents by end of 2026 — up from under 5% at the start of 2025 — and you have the outline of a transformation that's already happened without most real estate agents realizing it.

My read is that the adoption curve is running ahead of the capability curve. 

The AI that agents have integrated into their weekly workflow is still mostly the 2024 generation — generative models that draft and summarize. 

The 2026 generation is different. AI is already causing mass disruption to every industry.

So we have to ask what makes real estate any different?

Where we are now with AI in the real estate industry

The tools agents are running day-to-day aren't impressive on their own and, if you’ve used any AI tools, you’ll know that’s true.

• Superhuman for email autodrafts.
  
• Otter.ai for meeting recaps.
  
• Canva for listing graphics.
  
• CloudCMA for pricing.
  
• RemyAI for lead follow-up.

Each one shaves time off a specific weekly task, but none of them change what an agent fundamentally does. 

Jason Ivens at KW Westfield in Orem, Utah runs 300 agents and told Keller Williams that his top 101 closed more deals and hit $183,000 median income after he pushed AI adoption two years ago. 

That's a good headline until you read it carefully.

300 agents, only 101 hitting the top bracket, which means the bottom two-thirds either aren't using the tools at all or aren't using them well, and they're losing ground inside their own office to the agents sitting two desks away.

The AI-curious agents — the ones who tell their broker they're "experimenting" with ChatGPT for listing descriptions — lose market share month over month to the AI-native agents in the same office, who've built the AI tools into their pipeline. 

The AI doesn't make the AI-native agents smarter or more skilled. It gives them back time. Ten extra hours a week compounded over 12 months is roughly 20% more client-facing capacity, and in a zero-sum market where every listing goes somewhere, that's where the transactions migrate. 

By the time the AI-curious agents realise they're being outcompeted by their own colleagues, the gap is too wide to close without a full workflow rebuild most of them aren't going to do.

The part that gets missed in the "97% of brokerages are using AI" headlines is a simple distinction: using AI and benefiting from AI are doing very different jobs in these stats. 

Most agents have the tools switched on and are barely touching them. A handful have built them into their workflow and are taking market share from the rest. McKinsey's estimate that AI could generate $110 to $180 billion in annual value for the US real estate sector is real, and I think it's probably low, but that value is heavily concentrated in the top 20% of agents who've figured out how to compound AI gains across a book of business. 

The middle 60% get marginal gains, mostly in admin and content. The bottom 20% are handing their leads to the top 20% through slower follow-ups and weaker positioning. Call it a redistribution event. The transactions are getting closed either way. 

Fewer agents are closing them.

The future of real estate and AI by the end of 2026

IDC expects AI copilots embedded in roughly 80% of workplace applications by the end of this year, and PwC's Emerging Trends in Real Estate 2026 draws a line between today's generative AI and what they call agentic AI.

KW Command relaunched in February 2026 with direct API integrations to Gemini, RemyAI, and Rejig.ai, which means a KW agent's CRM now talks to three different reasoning models depending on the task, and that capability is sitting in the platform whether the agent knows how to use it or not.

My bet on what end-of-2026 delivers in real estate specifically: a workflow where inbound leads get qualified, scored, nurtured, and scheduled into showings without an agent touching the top of the funnel. 

The agent gets involved at the showing and stays involved through the offer. 

Everything before, between, and after — including post-close nurture and sphere-of-influence drip — gets absorbed by the stack. I'd put that at 70% probability by December 2026 for teams at Compass, eXp, Real, and Keller Williams, because they have the platform infrastructure and the budget to deploy it. 

For independent agents without a tech-forward brokerage behind them, I'd put the probability closer to 40%, and dropping fast as the cost gap between a platform agent's stack and an independent's stack widens.

The tools that will matter most by December are those coming from the major models: Anthropic and OpenAI. Claude Cowork, as an example, can handle 80% of real estate admin tasks without much (or any) human interaction.

The model capability is already there. And it’s advancing incredibly quickly. Quicker than Sam Altman predicted. Quicker than any major AI player could’ve foreseen.

What the models won't be able to do by the end of 2026 — and probably not by the end of 2028 either — is sit across from a seller who's just been told their property is 15% overpriced and hold that conversation. 

Walk a first-time buyer through their second round of inspection findings. 

Read a room. 

Tell someone what they don't want to hear. 

The agreeable-by-default problem in current LLMs is baked in through alignment training. It's a feature, and the frontier labs have no incentive to change it (ask anyone who's tried to get ChatGPT to push back on a decision they've already made). 

Negotiation and judgment survive. Everything else doesn't.

Where we've been and what we got wrong

In August 2024, when the NAR settlement took effect, John Campbell at Stephens predicted a 50% agent attrition rate within two years, and the entire industry braced for it. 

It didn't happen. 

NAR's own mid-2025 numbers projected membership falling to 1.2 million by end of 2026 — down from a 1.6 million peak, a 25% drop — which is real, but slower than the doom forecasts.

Average buyer-side commissions, which were supposed to collapse under settlement pressure, actually held: Clever Real Estate's February 2026 data has them at 2.82%, higher than the 2.55% they averaged in early 2025, right after the settlement took effect. 

The lawsuit didn't deliver what the plaintiffs wanted, and most of the analysts covering the space have spent the last 18 months explaining why.

I was one of the people saying AI wasn't coming for agents in any serious way this year. I read the tools right and the deployment curve wrong. Even AI SEO has been adopted far more aggressively than we could've ever anticipated.

What I thought was a 2027 problem is already here for most real estate agents.

The reason the settlement failed to move commissions is clients kept asking for human representation and sellers kept paying for it, and regulation doesn't override what buyers and sellers are willing to do in a market. 

What moves markets is infrastructure. The commission settlement was a legal framework agents could resist by keeping their workflow unchanged, which is what happened. 

AI is an infrastructure change the agents themselves are driving by upgrading their own workflow, which is why it's going to move the number the settlement couldn't.

So will realtors get replaced by AI?

If your job is the tasks — listing copy, follow-up emails, MLS admin, CMA prep, scheduled calls, drip campaigns, sphere-of-influence nurture — you're already being replaced, and the only reason you still have a job is that your broker hasn't fully deployed the stack and your clients haven't noticed the difference. 

Both of those gaps close within the next 12 months. Max. 

The software is cheaper, more consistent, and by December 2026 it'll be orchestrating the whole task pipeline without a human pressing go between steps. You're not competing with other agents anymore. 

You're competing with a workflow that doesn't need to eat or sleep.

If your job is the decisions:

• should this seller take the $420,000 offer or wait three weeks
  
• should this buyer walk or renegotiate after inspection
  
• is this ZIP code about to get hot
  
• what does the price history tell you about a seller's motivation
  
• is this investor lying about their timeline
  

End-of-2026 AI won't touch you.

End-of-2028 AI probably won't either but that could change. The alignment problem means models will keep agreeing with whoever's typing into them, and the judgment tier requires the opposite: someone willing to tell a seller their expectations are wrong, or tell a buyer their dream house is overpriced by six figures.

The problem is most agents don't know which category they're in. 

They think they're decision-tier because they've been doing this for 12 years and closed $8M in 2024. But if they audit their week, 90% of the hours were task-layer work. 

The $8M closed on top of the tasks. Strip the task layer out and the closings stay. The software runs the pipeline. 

The agent sees the offer on the screen where their CRM used to send it.

The best agents at the task layer are the first ones to go.

By December 2026, I think NAR hits its 1.2 million projection, which is 400,000 fewer agents than the peak. Most of them will blame the market. Some will blame commission compression. 

A few will blame AI. 

Almost none of them will connect it back to the choice they made six years ago, when they defined their career as a series of tasks instead of a series of decisions.

Some real estate agents are sitting on a database worth six figures in commission income and doing nothing with it.

A case study from The Shift AI tracked a Tampa Bay brokerage that was losing an estimated $1 million per year in gross commission income because their average lead response time was 2.5 hours and 40% of leads were never contacted within 24 hours. 

When they deployed an AI agent, response time dropped to under one minute. Pipeline conversion improved 27% in 60 days. One in five AI-qualified leads booked showings.

And that’s a team with resources. Most solo agents have it worse.

NAR’s 2025 Home Buyers and Sellers Generational Trends Report found that 78% of buyers work with the first agent who responds. 

Not the best agent. Not the one with the best reviews. The first one. That stat has held steady for five years running.

So speed isn’t a competitive advantage anymore. It’s the baseline. And most agents are failing it.

I’m going to walk you through how to set up an AI follow-up agent for real estate leads using Claude Cowork, Anthropic’s desktop AI tool that launched in January 2026. 

You’ll need about an hour to set this up.

What Is Claude Cowork?

You’ve probably used Claude in a browser before. You type a question, it gives you an answer. That’s chat mode. Cowork is different.

Cowork runs inside the Claude desktop app (macOS or Windows, no mobile). When you open the app, you’ll see two tabs at the top: “Chat” and “Cowork.”

Click Cowork and you’re in a completely different interface. Instead of conversations, you’re creating tasks. 

You describe what you want done in plain English, and Claude breaks it into subtasks, executes them autonomously, and checks in with you at key decision points before doing anything irreversible.

Cowork can read, edit, and create files directly on your computer. It can open your browser and navigate to Gmail, Google Drive, or your CRM. It runs inside an isolated virtual machine (so it can’t accidentally wreck your system), but it has direct access to any folders you give it permission to touch.

Anthropic added MCP connectors in February 2026 for Gmail, Google Calendar, Google Drive, Slack, and others. 

These let Cowork search your inbox, read email threads, draft follow-ups, and update documents without you having to copy-paste anything between tabs. 

You authenticate once through OAuth (same process as connecting any app to your Google account), and Cowork can interact with those services on your behalf.

You need a paid plan. The Claude Max 5x plan is $100/month and gives you around 225 messages per 5-hour window. Max 20x is $200/month with around 900 messages. For a solo agent running follow-up workflows a few times a week, $100/month is more than enough. If you’re on a team processing 50+ leads per week, go with the $200 tier.

One thing to know upfront: Cowork uses significantly more tokens than regular chat because of how much computation it does behind the scenes. So you’ll burn through your message allowance faster than you would just asking Claude questions in a browser. 

Plan your usage around your lead processing schedule.

How to set up a lead follow-up workflow in Cowork, step by step

Download the Claude desktop app if you don’t have it. Sign up for the Max plan. Open the app and click “Cowork” at the top of the screen.

Before you create your first task, do two things:

First, set up your global instructions. 

Go to Settings, then Cowork, then Global Instructions. 

Type something like: “I’m a real estate agent in [your city]. When drafting emails, use a casual, first-name tone. Never use corporate language. Keep emails under 100 words. Always reference the lead’s original inquiry date and what they were looking for.” 

These instructions apply to every Cowork task you create, so you don’t have to repeat yourself each time.

Second, create a folder structure on your computer. Something like:

• Lead Follow-Up (main folder)
  • Incoming Leads (where you drop your CRM exports)
  • Drafts (where Cowork saves its email drafts)
  • Templates (your follow-up email templates)
  • Sent (where you move drafts after sending)

In the Templates folder, create a text file with your base follow-up template. 

Here’s one that works:

“Hey [First Name], this is [Your Name] with [Brokerage]. You reached out about [buying/selling] back in [Month]. I realized I dropped the ball on following up with you, and I didn’t want you to think I forgot. Are you still thinking about [buying/selling], or has anything changed? Either way, no pressure. Just wanted to reach out.”

That last part matters. “I dropped the ball” takes ownership of the gap instead of putting it on the lead. 

People respond to honesty about the lapse more than they respond to a polished sales email pretending the gap didn’t happen.

Now create your first task. Click the “+” icon in Cowork to give it access to your Lead Follow-Up folder. Then type your instructions. 

Be specific. The more specific your prompt, the better the output. Here’s what I’d use:

"I have a CSV of leads in my Incoming Leads folder. For each lead, I want you to:

1. Read their first name, email, original inquiry date, and what they were interested in (buying or selling, and the area if listed).
2. Draft a personalized follow-up email using the template in my Templates folder. Reference the specific month they reached out and what they were looking for. If their inquiry was about buying, ask if their timeline has changed. If it was about selling, ask if they’ve had a recent home valuation.
3. Keep each email under 80 words.
4. Save each draft as a separate text file in my Drafts folder, named [FirstName]-[LastName]-followup.txt."

Cowork will read the CSV, identify the columns, match each lead to your template, personalize every message, and output individual draft files. You review them (takes 10 minutes for 50 leads), tweak anything that sounds off, and send them through Gmail.

If you’ve connected the Gmail MCP connector, you can go further and tell Cowork to create the drafts directly in your Gmail drafts folder instead of saving text files. 

But I’d recommend starting with text files first so you can review the output quality before letting it touch your email.

Scheduling recurring lead follow-up with Claude Cowork

The manual workflow above is useful for an initial blast through your database. But the real power is scheduled tasks, which is Cowork running the same workflow automatically on a cadence you set.

There are two ways to create a scheduled task. 

The easiest: type /schedule inside any Cowork task. A setup wizard launches and walks you through a few questions, usually with multiple-choice options, so you’re not guessing what to type. 

You’ll set the task name, describe what it does, pick your cadence (hourly, daily, weekly, weekdays, or manual), choose which folder it has access to, and confirm.

The second way: click “Scheduled” in the left sidebar of Cowork, then click “+ New task” in the upper right. This opens a modal where you fill in the same details.

For a real estate follow-up agent, here’s the schedule I’d set up:

1. Export your CRM leads into a CSV every Monday morning. Most CRMs (Follow Up Boss, KvCORE, Sierra Interactive, BoldTrail) have an export feature. If yours doesn’t, copy-paste from your lead list into a Google Sheet and download as CSV. Drop the file into your Incoming Leads folder.
2. Set a weekly scheduled task to run every Monday at 8 AM. The prompt:

“Check my Incoming Leads folder for any new CSV files added this week. For each new lead, draft a personalized first-touch follow-up email using the template in my Templates folder. Reference their inquiry date and interest. Save drafts to my Drafts folder. After processing, move the CSV to a ‘Processed’ subfolder so it doesn’t get re-read next week.”

3. Set a second weekly task for Wednesday at 9 AM. This one handles the second touch for leads who haven’t responded:

“Check my Drafts folder for any files older than 3 days that haven’t been moved to the Sent folder. For each one, draft a second-touch email with a different angle: include a recent comparable sale in their area or a brief market update for their zip code. Save the second-touch draft as [Name]-followup-2.txt in the Drafts folder.”

One limitation you need to know: your computer must be on and the Claude desktop app must be open for scheduled tasks to run. 

If your laptop is closed at 8 AM on Monday, the task gets skipped and runs automatically when you open the app. So either keep a desktop machine running or time your schedules for when you know you’ll be at your computer.

It’s not ideal, but it’s the tradeoff for a $100/month tool vs. a $500/month purpose-built platform.

Reactivating your old real estate leads with Claude Cowork

This is where the real money is for most agents reading this, and the data backs it up.

The same Shift AI case study tracked a multi-state real estate team in Texas and Colorado that had accumulated over 18,000 inactive leads (dormant 120+ days). 

They ran an AI reactivation campaign. In 90 days, they re-engaged over 900 leads, moved 62 back into active pipeline, and closed 15 listings from contacts the team had completely written off. 

They doubled their contact touchpoints without adding a single person to the team.

An independent broker in Bergen County, New Jersey, from the same study, reduced admin workload by 40%, saved 3-4 hours per day, and increased listing acquisitions by 30% in the following quarter, all by having AI handle 100% of first-touch communications.

Research from BoldTrail shows that reactivating dormant contacts costs 5-10x less than acquiring new leads and converts at 3-4x higher rates. Dormant prospects typically convert at 8-12%.

Here’s how to do it with Cowork. 

Export everyone from your CRM who came in over the past 24 months and didn’t convert. Drop the CSV into your Incoming Leads folder. Give Cowork this prompt:

"Read through this lead list. For each person, draft a re-engagement email with the following rules:

• Reference the specific month and year they first reached out
• Reference what they were interested in (buying area, selling address, price range, whatever the CSV contains)
• Ask one simple yes/no question: ‘Are you still thinking about this, or have things changed?’
• If they were a buyer lead, add one sentence about how the market in their target area has shifted since they last looked
• Keep each email under 80 words
• Tone should be casual and direct, like a text from someone they’ve met before, not a marketing email
• Save each draft with the lead’s name and email as the filename"

Review the drafts. Send the ones that feel right. For the ones who reply with interest, go back to Cowork and have it draft a second-touch email with a different angle: a recent comparable sale in their area, a note about interest rate changes, or a question about their timeline.

Say you have 500 old leads. BoldTrail’s benchmarks suggest dormant prospects convert at 8-12%. 

Even at the low end, 8% of 500 is 40 re-engaged conversations. The Shift AI data shows roughly 1 in 5 of those will book a showing. That’s 8 showings from contacts you’d written off. Close a third into listings and you’ve got 2-3 new listings from a dead database and a $100/month tool.

Scale that up. The Shift AI study found that a brokerage with 50 agents and 35,000-60,000 dormant contacts could generate 1,400-2,400 additional transactions annually through segmented reactivation campaigns, translating to $2-6 million in commission revenue.

‘Do I need Claude Code or just Cowork?’

If you’re not a developer, you want Cowork. Full stop.

Claude Code is a separate tool that runs in your terminal (the black screen with the blinking cursor). It’s built for software engineers who want to build custom systems. It can do things Cowork can’t, like build a full conversational SMS agent that texts leads via Twilio, holds back-and-forth qualifying conversations, scores leads automatically, and books appointments into your calendar without you touching anything.

But it requires you to know (or hire someone who knows) how to write code, manage APIs, and deploy to a server.

That’s a different article and a different budget.

Cowork was built for people who want AI to do real work without writing a single line of code. For the highest-impact use case most agents need (drafting and scheduling personalized follow-up at scale), it handles it without any technical overhead.

Why this matters more in 2026 than it did last year

The Inman Real Estate Technology Survey from 2025 found that the average agent response time to a new lead is 917 minutes. That’s over 15 hours. And InsideSales research on 55 million sales activities found that 57.1% of first call attempts happen after more than a week.

After more than a week.

Meanwhile, NAR data shows 62% of real estate inquiries come in after business hours, between 6 and 9 PM and on weekends, exactly when you’re not at your desk.

The agents who set up a system like this now, even a basic one, end up in the same position as the agents who claimed their Google Business Profile before everyone else did. Or the ones who started building an email list in 2014 when it felt pointless.

$100/month and an hour of setup. That’s the gap between you and the agent in your market whose old leads are booking calls while they sleep.

I’ve watched this pattern play out in SEO for over a decade. The people who move early on shifts like this build advantages that compound.

The people who wait until it’s obvious end up paying 10x more for the same result (hello every agent who’s now paying $500/month for Zillow leads and calling them back the next day).

Most real estate agents are using the wrong AI search engine to get found.

I went through the data. The market share, referral traffic, citation behaviour, conversion rates across ChatGPT, Google's AI Mode, Gemini, Perplexity, Grok, and Copilot. I wanted to know which ones are genuinely sending leads to agents right now, and which ones are future opportunities.

Here's how I'd rank each AI search engine for real estate leads — and why.

1. Google AI Mode

This should be your number one priority. And almost no realtor is treating it that way.

Google AI Mode has 100 million monthly active users in the US alone. It sits inside the search engine that still controls 92% of the market.

When a homeowner types 'best estate agent in [your city]' into Google, there's an increasing chance they're getting sent straight into AI Mode without knowing about it.

AI Mode has a 93% zero-click rate.

That sounds terrifying until you realise there's still opportunity here.

But if you're not being cited in that AI response, you don't exist. There's no page two to fall back on. There's no 'well, at least I'm ranking somewhere.' You're either in the answer or you're invisible.

iPullRank's referral data from 2025 showed that high-authority, community-driven sites dominate AI Mode citations. Reddit, YouTube, Wikipedia, Zillow — these are the platforms Google's AI trusts.

What does that show you?

Google's AI is looking for agents, teams, brokerages, and brands with authority signals across the web. Your Google Business Profile, your reviews, your local content, your backlink profile. The stuff you should've been building for the last decade.

If you've been doing AI SEO properly, AI Mode is your biggest advantage. If you haven't, it's about to become your biggest problem.

2. ChatGPT

ChatGPT holds somewhere between 64% and 81% of the AI chatbot market depending on which data set you're looking at (SimilarWeb puts it at 64.5% as of December 2025, Conductor's data from November 2025 says 87.4% of all AI referral traffic comes from ChatGPT).

The numbers are massive. 800 million weekly users. The third most visited website on the planet.

And for real estate agents, the referral traffic story is incredibly high intent. The agents we work with see the highest intent leads from ChatGPT compared with any other LLM search engine.

Exposure Ninja reported that AI search traffic converts at 14.2% compared to Google's 2.8%. Microsoft Clarity's study across 1,200 sites found LLM traffic converting at 3x the rate of traditional channels.

Based on the successes of our clients (take Katelyn Warren for example) we can confidently say this is the case.

When someone asks ChatGPT to recommend an agent or compare neighbourhoods, it pulls from your existing web presence. Your site, your reviews, your mentions in local press, your YouTube transcripts.

The agents who focused on content distribution (i.e., different channels) are the folks winning in ChatGPT answers.

The real estate agents who get ahead of this now, while the space is still wide open, are the ones who'll own those citations as the rate climbs.

BrightEdge found that ChatGPT makes it easy to get mentioned — but only 2 in 10 mentions include a clickable link. So your brand shows up, but the user doesn't always have a direct path to your site. That's a limitation worth knowing about, not a reason to ignore the platform.

ChatGPT is the volume leader. The conversion data is strong. The real estate-specific citation rate is still low, which means there's a window right now for agents who move first.

3. Perplexity

Perplexity is small.

Around 22 million monthly users. Roughly 6.2% of the US market. It processed 780 million queries in May 2025 — big growth, but a fraction of ChatGPT.

So why is it third?

Because Perplexity users are researchers. The platform averages over 5 citations per answer (BrightEdge data). That's more than any other AI engine.

Every response comes with sources, links, and attribution. When someone uses Perplexity to ask 'who are the top-rated estate agents in Austin for luxury homes,' they get a sourced, linked answer — and they click through.

The conversion data backs this up. Seer Interactive's analysis found Perplexity converting at the second-highest rate of any AI platform. Microsoft Clarity's study showed Perplexity referrals converting sign-ups at 7x the rate of direct traffic.

Perplexity's user base skews professional. Students, analysts, decision-makers. The kind of people who read the sources, compare options, and make informed choices. That's exactly who you want finding your content when they're researching a move.

Only 1 in 5 Perplexity answers mention a brand at all (BrightEdge, May 2025). So while the traffic it does send is high quality, the total volume is small.

You're not going to build a pipeline off Perplexity alone. But as a supplement to your Google and ChatGPT visibility, it's punching above its weight.

4. Gemini

Gemini is the story of 2026. Google's AI assistant nearly quadrupled its market share — from 5.7% to 21.5% in twelve months (SimilarWeb, January 2026).

It has 400 million monthly active users globally. And its referral traffic to external websites grew 388% year-over-year from September to November 2025, dwarfing ChatGPT's 52% growth in the same period.

Those numbers look incredible on paper.

But for real estate leads, Gemini has a problem.

It's tightly integrated into Google's ecosystem — Android, Workspace, YouTube — which means most people encounter it inside products they're already using, not as a standalone search tool. The US market share is only about 3.4% (SimilarWeb), way below the global figure, because American users are still defaulting to ChatGPT for direct AI search.

We don't have strong data on Gemini-specific conversion rates for real estate. Seer Interactive's study showed Gemini converting at about 4x the rate of direct traffic for general sign-ups — decent, but behind Perplexity and Copilot. The real estate-specific data simply isn't there yet.

I'd watch Gemini closely. The trajectory is aggressive.

But right now, it's a visibility play, not a lead generation one. Make sure your content is structured for AI citation (clear headings, direct answers, schema markup) and Gemini will likely become more important over the next 12 months.

Just don't bet your pipeline on it today.

5. Grok

Grok went from zero to 3.4% global market share in a year (SimilarWeb, January 2026). That's impressive for a platform built by xAI and integrated into X (formerly Twitter). Around 29.6 million visits in July 2025 alone.

Grok's audience skews toward high-net-worth, entrepreneurial users on X.

And we've seen early signals of conversions coming from affluent buyers and sellers who use the platform as part of their research process. That's a small but valuable segment — the kind of clients most agents would love to attract.

The platform rewards real-time data, trending topics, and social-media-native content. If you're an agent who's already active on X and building a personal brand there, Grok's integration could surface your content to exactly the right audience.

The volume isn't there yet for most agents to prioritise it. But if you work in luxury or high-value markets, Grok is worth paying attention to. The affluent buyer profile and the platform's growth trajectory make it one to watch closely heading into the second half of 2026.

6. Copilot

Microsoft poured $13 billion into OpenAI. Copilot is baked into Windows, Office, Teams, Outlook, and Edge. It should be dominating.

It's not.

Copilot sits at roughly 1.1% global market share — and it's declining (SimilarWeb, January 2026). In the US it performs slightly better at around 14% in some data sets, but that's heavily skewed by enterprise usage within Microsoft 365, not consumer search behaviour.

The irony is that Copilot's conversion data is technically the best of any platform. Microsoft Clarity's study found Copilot referrals converting subscriptions at 17x the rate of direct traffic. But the volume is so low that the stat is almost meaningless for real estate. You might get one incredibly qualified lead per quarter from Copilot. Maybe.

The brand confusion doesn't help either. Microsoft has Copilot in Windows, Copilot in Edge, Copilot in Bing, Copilot in Teams — users don't know which one does what, and most of them aren't using any of them for property searches.

I'd put Copilot last. Not because the technology is bad, but because the user behaviour isn't there for real estate.

So which sends the best real estate leads?

Stop treating 'AI visibility' as a single strategy. Each platform has different citation behaviour, different audiences, and different conversion patterns.

If you had to pick one thing to focus on right now, it's making sure Google's AI Mode can find you and trust you. That means your Google Business Profile needs to be flawless. Your local content needs to answer specific questions about specific neighbourhoods. Your reviews need to be recent and consistent. Your site needs schema markup that tells AI exactly what you do and where you do it.

For ChatGPT visibility, the play is broader authority — being cited on industry publications, having YouTube content with solid transcripts, getting mentioned on Reddit and in local press. ChatGPT pulls from Bing's index, so your Bing Places profile matters more than you think.

Perplexity rewards depth. Long-form, well-sourced content with clear attribution. If you're publishing data-backed neighbourhood guides and market reports, Perplexity will find them.

And for the rest — Gemini, Grok, Copilot — keep an eye on them, but don't restructure your business around platforms that aren't sending consistent real estate traffic yet.

I've been in SEO long enough to know that the agents who move first on distribution shifts like this are the ones who build advantages that compound for years. The ones who wait for the 'definitive guide' to AI search in 2027 will be fighting over whatever scraps are left (I'm speaking from firsthand experience).

The data's here. The rankings are clear. The question is whether you'll act on them before your competitor in the next postcode does.

Watch on YouTube

'I don't have time to make videos.' Yes, you do.

'I don't know what to talk about.' Yes, you do.

'I'm not a video person.' Nobody is until they start.

The excuse factory runs 24/7 in real estate. And I get it. You're busy showing homes, managing clients, and trying not to drown in admin. Video feels like one more thing on an already overflowing plate.

But you're probably sitting on 10, 20, maybe 50 blog posts right now that could be turned into video scripts with AI in under 10 minutes (not an exaggeration).

From content you've already written but haven't distributed yet.

And the reason this matters more in 2026 than it ever has before is because YouTube is the #1 cited source in Google's AI Mode and ChatGPT.

This means, it's a traffic and lead driver.

P.S. If you need an AI SEO company to do this for you... well, book a call.

YouTube Is Now The #1 Cited Source In AI Search

BrightEdge tracked AI citations across ChatGPT, Perplexity, and Google's AI Overviews from May 2024 to September 2025. YouTube averages a 20% citation share across all AI platforms.

That makes it the single most cited video source. 200 times more than any competitor.

And this isn't just Google playing favourites with its own platform. ChatGPT and Perplexity have zero corporate reason to prioritise YouTube.

They do it anyway.

Because YouTube has the depth, the transcripts, and the structured content that AI models need to pull answers from for real estate searches.

Surfer's AI Citation Report backed this up — across 36 million AI Overviews and 46 million citations, YouTube sat at approximately 23% of all citations. Ahead of Wikipedia. Ahead of Reddit. Ahead of every government site, every news outlet, every niche blog.

Let that sink in for a second.

When a potential buyer types 'best neighbourhoods in [your city] for families' into ChatGPT, search engines aren't only pulling from website content, it's pulling from YouTube transcripts, too.

And if you don't have the right video content, you won't exist in ChatGPT's answers by the end of 2026.

It's that simple.

How Do You Turn Real Estate Blogs Into YouTube Videos With AI?

I'm not asking you to become a YouTuber. I'm not asking you to buy a ring light, learn Final Cut Pro, or start doing jump cuts.

(You're probably exhausted from dancing on Instagram already to sell a house... am I right?)

I'm asking you to take the blog posts you've already written — the neighbourhood guides, the market updates, the first-time buyer tips — and turn them into video scripts using AI.

Here's the workflow and prompt:

1. Take a piece of content you've already written (or somebody else's highly ranking blog).
2. Use the prompt we're giving you for free.
3. The prompt will create the video title, description, key talking points, the length of time you need to record... everything.
4. Use Claude Opus 4.6 or 4.5 (we don't recommend ChatGPT... it sucks lately).
5. Your entire video transcript will be done in 3 minutes.
6. Go record it.
7. Post it and distribute.

That's it.

You've just created a piece of content that AI models can cite, Google can index, and potential clients can watch at 2 AM when they're lying in bed thinking about moving.

The blog post was the hard part.

You already did the research. You already organised the thoughts. The video is just you saying what you already wrote — but now it lives on the platform that AI trusts more than any other.

Do You Need To Go Viral For AI Search To Cite Your Video?

'Oh but I'll only get 200 views...'

This is probably the most common pushback I hear.

And it tells me that most agents are measuring video success the same way they measure Instagram success.

Views don't matter. Not in the way you think.

A neighbourhood guide with 200 views on YouTube isn't competing for virality. Your job is to attract high-intent buyers and sellers in your market.

That's it.

Nate Clark started his YouTube channel 30 days ago and has already secured a listing. His YouTube videos, on average, generate 15–45 views. He generated 300 views in 30 days.

What does that show you?

These are the warmest leads you'll ever get. By the time they contact you, they already feel like they know you.

Agents with small YouTube channels — under 500 subscribers, sometimes under 300 — report consistent inbound leads from their content. Not thousands of views. Just the right views.

Compare that to the leads you're buying from Zillow or whatever platform you're currently renting your pipeline from. Those people don't know you. They don't trust you. They gave their number to a form and now six agents are fighting over the same callback.

Will AI Clones Matter For Real Estate Video?

I knew this was coming.

The idea is seductive. You record a few minutes of yourself talking, feed it to some AI avatar tool, and suddenly 'you' are pumping out videos while the real you is at a showing.

Sounds efficient. Sounds smart.

It's also the fastest way to destroy the one thing that makes video work for you in the first place — trust.

The entire value of video for agents is that it's you. 

Your face, your voice, your knowledge of that specific street, that specific neighbourhood, that specific market. When someone watches a 6-minute video of you walking through your local area and explaining why families love it there, they're not just absorbing information. They're deciding whether they like you. Whether they'd trust you with the biggest financial decision of their life.

An AI clone can't do that.

It looks like you. It sounds close to you. But something's off. 

And people feel it… You’ve seen it, I know you have. The uncanny valley isn't just a tech problem.

It's a trust problem. And in an industry where trust is literally your product, that's not a risk worth taking.

Don't get me wrong, AI clones will improve. They already have. But the moment your audience finds out (and they will), you've lost something you can't get back.

There's another issue nobody talks about. If every agent starts using AI avatars to mass-produce video content, what happens? 

Saturation. Hundreds of identical-feeling videos flooding YouTube.

That's the opposite of a moat. That's a race to the bottom.

The agents who win on YouTube over the next few years won't be the ones who produced the most content. They'll be the ones who produced the most real content. The stuff that's imperfect, a little rough around the edges, but unmistakably human.

AI Gives You No Excuse To Not Make Video in Real Estate

Let's go back to where we started.

'I don't have time.' You don't have to write anything new. You're repurposing what already exists.

'I don't know what to talk about.' Your blog posts are a content library waiting to be spoken out loud.

'I'm not good on camera.' Nobody watching a local neighbourhood guide expects you to be a TV presenter. They expect you to know the area. That's it.

AI has made this process absurdly simple. Feed your blog into a model. Get a script back. Record it on your phone. Upload it to YouTube.

The agent who does this 20 times over the next six months will have 20 indexed, citable, searchable video assets working for them around the clock.

The agent who doesn't will keep wondering why their competitor shows up in AI answers and they don't.

This isn't about being a content creator. Trust me, I would never wish that on you. Video is going to be the only surviving moat left when AI models become so good, people won't know who to trust.

Even if you don't have the content, there are millions of blogs online you can repurpose. As long as you're adding your unique view and opinion, use whatever form of inspiration you want.

The gap between agents getting cited in AI using video, and agents who aren't is embarrassingly small.

The only question is whether you'll close it.

I've watched this pattern play out in SEO for over a decade. The people who act early on a shift like this build an advantage that compounds.

The people who wait until it's obvious end up fighting for scraps (hello Zillow owning Google for 2 decades or hello agents who didn't use Google My Business when it first launched).

Yes, there's a difference between AI SEO and traditional SEO for real estate agents. 

The noise is in the nuance.

And it's the nuance that catches out most real estate teams. They end up paying for SEO and AI SEO separately. Which is both unsustainable and unnecessary.

I have 12 years in traditional SEO and 5 years in AI SEO (I'm likely the only person who has in real estate) and so, I'm uniquely qualified to tell you the differences.

These differences will save you money and bring you more listings so, it's worthwhile you read this if you're an agent who feels like you maybe are being mislead by agencies online.

The Key Differences Between AI SEO and SEO in Real Estate

TLDR: the biggest difference between AI SEO and traditional SEO in real estate is how you deliver content and links to each LLM.

Most LLMs (ChatGPT, Copilot, AI Mode) steal from traditional search engines. 

• ChatGPT uses Google's index (it's library of content) to return answers.
• Copilot uses Bing's index to return answers.

AI SEO cannot exist without normal webpages. However, AI search engines do not use pages that have no traffic, no structure, no intent and, no value.

How do we define no value in real estate? 

Simple. Look at any of your last 50 blogs that your CMS provider has given you and ask yourself these questions:

• Do these blogs answer questions a buyer/seller might ask?
• Are these blogs long enough to be valuable?
• Are people actually searching for these topics?
• What intent does your blog capture?

Ranking in AI models works in the same fashion as you assessing content manually. Traditional SEO could be easily manipulated whereas AI SEO is more heavily in favor of how your content speaks to a person specifically.

Optimizing for AI Search vs. Google

Here's where things should get technical. Search engine optimization (SEO) is anything but simple. Websites are complicated, algorithms are messy and CMS providers have done their absolute best to offer the least SEO-friendly platforms available (that's for another time, though).

But, despite that, I'll outline the comparative differences between the two disciplines. You'll notice how much overlap there is. And then, hopefully, you'll see where the nuances are.

Hyper-specific content about your market

Traditional search results were ranked on traffic and links. The algorithm changed the quality threshold based on the industry but, for real estate content, if you could get people to visit your site, and other websites to give you backlinks, you did pretty well.

With AI SEO, it's now about the specificity of your answers.

Your content should no longer be: how do I sell my home in Florida?

And instead become: how can I sell my 4-bedroom home, with a pool, in Florida for higher than the market average in the next 3-6 months?

This is natural conversation. This is how normal homebuyers and sellers query ChatGPT. 

How do you apply this to your content?

Like this.

Let's say you've written a blog called 'Moving To Florida: Everything You Need To Know'. It's a pretty common topic and something AI tools like to reference in real estate.

To make your content match up to the conversational intent of somebody's question, you'd need to break it down into chunks:

• Safety
• Different neighborhoods
• Home prices
• The current market
• Jobs
• Transportation

Etc, etc.

It's less about opinion, and more about depth. When someone asks AI that question, your content needs to be "chunked" (i.e., segmented) down in this fashion so it's easy to scan and easy to extract answers from.

Building brand citations across the web

A brand citation is where your brokerage, team name or personal name are mentioned on different websites. Citations are the key to AI SEO.

Your visibility in AI search engines will be controlled by:

• How consistent your information is online
• Where your information is being listed
• How often you keep your AI citations fresh

A big problem with real estate professionals is they change teams or brokerages... a lot. 

And so, ChatGPT, for example, could find 3 different brokerages attached to your name. If at any point the AI-driven search is confused, it simply will not return your business.

You will miss out on leads, rankings and organic traffic (inbound, at that).

Traditional SEO relied on citations, too. But not to the same extent. AI SEO is 95% reliant, whereas traditional SEO was 50-60% reliant.

Creating key real estate profiles

Do you have a Zillow profile? How about a Fast Expert? Or Rate My Agent?

The chances are you either:

a) have the profiles but don't keep them updated

b) you don't have the profiles apart from Zillow

This is typical in real estate (unfortunately). Real estate SEO relies on trusted authority platforms. This means most LLM engines and Google go to the same sites over and over for information, because they trust them.

If ChatGPT expects to see you on 10 different real estate profiles, but you're on 1, what are the chances of you being the agent AI recommends?

I'll save you the trouble of thinking: the answer is zero.

AI results work in the same way as traditional SEO when it comes to what we call 'seed sites'. The play a huge role in real estate, whether it's for lead generation, reviews or brand building.

The nuance here is that ChatGPT, Perplexity, Grok etc., rely on the same 40 profiles. Whereas Google used to rely on hundreds.

Reviews and trust

Reviews are where the two disciplines balance out. Local SEO is heavily skewed on the reviews you get, how often you get them and where you get reviews.

Google's AI Mode uses Maps to find the agents people are looking for when they search for 'who's the best?'

Best is subjective but, most platforms treat 'best' based on:

• Reviews
• Sales history
• Local expertise

And it extracts that from your reviews. You can't claim you're the best, you need your clients to do it for you. Which is why reviews make such a big difference in you appearing in AI-generated answers (or not).

Here's where you should get reviews in priority order:

1. Google Business Profile
2. Zillow
3. Realtor
4. Fast Expert
5. Yelp
6. RateMyAgent
7. Angi
8. Homelight
9. Expertise
10. Bestrated

And, as we said about citations, make sure every profile you make is consistent.

Traffic from different channels

This is the biggest difference in real estate SEO in 2026. It's where you get your traffic. 

Traditional SEO has gone from Google only to now being searched everywhere. The age old SEO strategy was based on driving traffic to your website from Google. And it worked well.

But, now, you need to consider:

• Instagram
• X
• YouTube
• Reddit
• Quora
• Facebook Groups

Traffic needs to come from everywhere. AI scans every platform for mentions of who you are. Buyers and sellers are in subreddits asking for recommendations for agents to work with. Mortgage brokers are setting up Facebook groups to build local lead generation inquiries. 

Getting traffic from every platform makes you authoritative. It's essentially what digital marketing should have been but now, LLMs have made it a necessity rather than an afterthought for real estate agents.

What Are The Cost Differences?

AI SEO will cost you much less than a traditional SEO campaign.

Based on my experience of well over 100 SEO campaigns in the last 12 years, I can tell you the average retainer cost was north of $4,200 per month. This was to cover content writing, links, consultancy, technical SEO: the whole 9 yards.

That was for an SMB. Enterprise SEO campaigns used to range in the region of $10,000 - $20,000 per month.

For real estate agents specifically, you could find a local SEO campaign for $1,500 a month. But it would likely underperform.

In the age of AI-powered SEO, you can get results for $800 - $1,000 a month.

It depends on:

• How much work you've done to your site in the past
• How well-known you are already in the AI answers
• If you appear in Google's AI overviews
• How many backlinks your website has
• How good your CMS provider is

We have helped agents with no website history get to well over $2m in pipeline in 6 months. And others, we've helped achieved that in 90 days or less.

Which Will Give Me Quicker Results (More Listing Appointments)

AI SEO will give you much quicker results and inbound listing appointments.

A typical SEO campaign timeline could be anywhere from 6 - 12 months. There were/are ways you could speed it up, but it involved risk.

AI search can start working immediately. As in, within 24 hours in real estate. We've helped agents become the #1 results in ChatGPT and Google's AI Mode within a week of starting with us (and they're still there today).

Which Will Send Me Better Qualified Leads?

AI SEO will send you much better leads.

Why?

Because people interact with AI engines as if they were an assistant. It's personal. It's a conversation, rather than a search action.

What we find is that realtors who rank better in ChatGPT, typically get:

• Shorter sales cycles
• More affluent buyers
• More frequent inbound calls

Traditional SEO can absolutely do this, too. But it's the length of time vs. the ROI (return on investment) that makes all the difference.

They don't compare.

That's coming from someone who has helped generate north of $100m from traditional SEO.

Which One Is Best for Realtors: AI SEO vs Traditional SEO

AI SEO is a much better choice for real estate agents in comparison with traditional SEO.

But, a lot of the deliverables and strategies (as I've outlined above) are largely the same. It's the frequency, the quality, the intent and the distribution that makes all the difference now.

It should no longer be a choice between a traditional SEO agency or an AI SEO agency: because ALL agencies should be catering for AI now.

If they're not, they're already years behind. And, they're likely not going to help you, if they can't help themselves.

FAQs

What is GEO (Generative Engine Optimization)?

GEO is the same as AEO (Answer Engine Optimization) and AISEO. It's a different name for the same practice. The most commonly used term is AI SEO.

Will traditional SEO still work for real estate agents?

Yes traditional SEO will still work for agents who have more budget and are willing to play the long game. This carries risk. Google is moving everything to AI Mode and so, traditional search, will change thanks to artificial intelligence.

Should I pay for real estate SEO and AI SEO?

No, you shouldn't pay for real estate SEO and AI SEO separately. You should choose an AI-first SEO agency to do your work because the deliverables are 90%. the same. It's not worth the additional expense to pay for both.

Watch this on YouTube

OpenAI announced they will be testing ads within ChatGPT on January 16th, 2026.

This comes as no surprise to most users.

After Sam Altman did a complete 180 on his stance on not using advertising as a business model, we can see it as a bad thing or, capitalize on the opportunity of a lifetime.

Realtors will have access to millions of homebuyers and sellers for a fraction of the cost they’d pay on Meta or Google PPC (Pay Per Click).

AI search already proved that traditional marketing channels could (and should) be different.

ChatGPT ads will be hyper-personalized, they’ll have higher intent signals than any other advertising platform and, the best part is, people already trust ChatGPT’s responses.

Let’s take a look at how ChatGPT ads could work for real estate agents in the very near future*.

How will ChatGPT ads work?

Based on OpenAI’s release article, ads will be shown natively at the bottom of a ChatGPT window.

(Source: OpenAI)

Initially, ChatGPT ads will be shown to free users and to Go tiers (a newly introduced subscription at $8 per month). 

95% of ChatGPT’s users are on the free tier. 

That’s 750 million users.

That’s most homeowners in your market. That’s most of the motivated sellers in your area, too.

Can real estate agents use ChatGPT ads for leads?

Yes, real estate agents can use ChatGPT ads to generate leads.

The advantage of early movers’ advantage cannot be stressed enough. Most realtors will ignore this announcement. 

Leaving the market open for agents who want:

1. Cheaper paid traffic
2. Higher-intent leads than any other platform
3. The ability to lower CPL (cost per lead) by an estimated 30-40%

How do we know this will happen for real estate?

Well, answer this…

When you have 18 months of a ChatGPT user’s memories, chat windows and preferences, what do you think they’re going to do with this data?

This (as much as OpenAI says they won’t use it) will be the initial draw for advertisers.

For people looking to buy, sell or rent a home, ChatGPT ads will likely prove more useful than other platforms for that reason. If the ads are integrated correctly, most consumers won’t even notice they’re clicking ads.

Google has used ads for over a decade and people still don’t know when they’re clicking a sponsored result versus an organic. 

How will ChatGPT ads drive leads to an agents’ site?

Traffic will be sent directly from ChatGPT to the agent’s website. 

It’s that simple.

There won’t be any middle interface between a user’s conversation and your website.

This means you will need:

1. A dedicated landing page for ChatGPT ad traffic
2. The ability to track conversions on these pages
3. The ability to receive form enquiries on your website

Those agents with difficult CMS platforms should consider how this works for them now. If your CMS platform doesn’t allow you to add forms freely, or create paid landing pages, you’re going to be restricted.

Which will be costing you listing appointments… 

The alternative is that ChatGPT ads will send people directly to your Map (using Google’s index) or to call you from within the chat window.

The less friction, the better. If OpenAI takes any inspiration from Google or Meta, they benefit from keeping users in their ecosystem (think Google Local Service Ads or Meta’s Lead Form).

This could mean less traffic overall to your website but… better qualified leads are great, no matter how you get them.

Right?

When will ChatGPT ads be available for real estate?

OpenAI is rolling out testing in the United States as of January 16th, 2026.

And with existing partnerships with portals like Zillow, there’s a high likelihood that we will start seeing ads imminently. With a full rollout likely to be within Q1 2026.

As soon as ChatGPT starts serving ads based on geography, home interests, demographics and personalised history, we will be entering a new ecosystem of advertisement.

If OpenAI is willing to do this, you can be sure Google will introduce ads into AI Mode. As will Copilot, Perplexity and Grok.

Where does this leave the real estate industry?

Real estate is the industry I see benefitting from this change in the ad ecosystem.

Individual agents, smaller teams and even larger brokerages find it hard to generate paid leads, consistently, through current channels.

With rising CPM (cost per mille) costs on Meta, and CPL (cost per lead) higher than ever before, this announcement from OpenAI is a much needed change.

Will this replace Zillow leads?

Will this mean Realtor.com will need to revisit its business model?

No, it won’t. 

This, as with most marketing channels, will be another route to customer acquisition, rather than a replacement.

Our advice would be to pay attention to how this pans out and jump at the opportunity as soon as it presents itself to you.

We’ll be watching and updating throughout the coming weeks and months.

If you want to stay updated, join the waitlist, and be the first to win in this new market.

*This article is FlyDragon’s opinion, not fact. ChatGPT ads have yet to roll out as of the 19th January 2026. We’re using official information from OpenAI to provide you this information.

For the last decade, the real estate industry has operated on a "pay-to-play" model. 

You paid the portals for access, you called the leads within 5 minutes, and you prayed for a conversion. 

It was a simple, albeit expensive, transaction.

Unfortunately, paying Zillow for seller leads leaves you with small margins, a co-dependent business model and, if you’re honest, it’s not why you got into real estate.

Sellers are using AI search to disqualify agents they don’t see as the right fit. Social media is becoming ruined by slop and every agent is using the exact same lead generation strategies as they were 5 years ago.

If you’re tired of being a realtor who spends their time chasing seller leads, rather than sellers coming to them, I’ll show you how to change that in 2026.

What percentage of home sellers are using AI search?

Recent data shows that 82% of Americans are now using AI tools (like ChatGPT, Gemini, and Perplexity) to gain housing market insights.

Let that sink in.

Four out of five sellers in your market are consulting an AI algorithm before they ever consult a human.

Why? Because the AI doesn't try to sell them something. And that’s what makes AI SEO so powerful.

It gives them unbiased data on customer acquisition costs, market trends, and neighborhood safety. It answers their specific, high-intent questions without demanding a phone number in exchange.

67% are using ChatGPT and 54% are using Gemini. That means sellers are having natural conversations with an AI assistant because they’re trusted.

AI is quickly becoming a companion for many people and so, sellers trust their information more than they trust their friends or family (that’s not a joke, I promise).

If you’re not capitalizing on AI search to generate seller leads, you’re losing GCI and you don’t even know it.

Are inbound seller leads motivated?

The average conversion rate for a "cold" outbound lead (cold calling, door knocking) or a "forced registration" lead (Zillow, Realtor.com) sits at a miserable 0.2% to 1%.

That means you have to outreach 100 people to find one person who might sell.

In contrast, inbound AI leads convert at over 15%.

Let’s say you get 500 visits to your website per month. That’s 75 inbound seller leads, every month.*

When a seller finds you through an AI Overview or any other LLM, they have already performed a zero-click search. They have consumed your data, verified your brand authority, and decided you are the expert.

If you use the right content strategy, by the time a seller is reaching out to you, you’ve already answered their questions (more on that later in the article).

*This doesn’t account for market downturns or seasonal swings. And most agents don’t invest in SEO to get 500 visits per month.

How real estate agents can get more seller leads in 2026

I’m tired of seeing the same lead generation tactics online in every real estate blog. The only brand we know who share battle-tested ways to get more seller leads is ListingLeads.com. 

So, we’re going to share with you how we’re helping agents generate more seller leads every month instead.

Make AI models recommend you

Real estate agents were never able to outrank property portals. And it’s why so many of you never invested in SEO.

That isn’t the case with AI search. Most AI search engines recommend agents, teams and brokerages because they understand intent better. A conversation is not a search. And so, property portals rarely make it as a recommendation.

To make an AI model recommend you as the ‘best agent to sell a home in…’, here’s what you need to know:

1. You need to teach AI models who you are, where you work and what you sell.
2. You need to specialize in something (price brackets, home types).
3. You need to consistently reference that everywhere online.

It sounds like a lot of work (it is) but it’s the most powerful way to generate seller leads today.

Here are two tactics you can use today.

Firstly, you have NO authority so, go and piggyback on someone who does.

Think about:

• Social media profiles.
• Local press.
• Business news and media outlets.
• Other businesses in your local area you could sponsor.

You can use these websites to promote your brokerage, or your team, and AI assistants eat it up. Talk about popular listings you’ve sold, how many homes you’ve sold in the last 12 months, any awards you’ve won for being a great salesperson.

(If you want to see a full breakdown, download the full AI Masterclass PDF).

Sales-focused AI chatbots

Most real estate chatbots are useless. They are just annoying digital gatekeepers trying to force an email address out of a seller.

That doesn’t work anymore. 

Consumers are tired of clicking "Chat" only to be met with a glorified contact form saying, "Hi! Someone will be with you shortly."

To make a chatbot actually convert a seller, you need to stop thinking of it as a greeter and start training AI to provide value and qualify your sellers.

Here is what you need to do:

• Train it on your brain: Upload your past blogs, local market reports, and tax guides.
• Stop the gatekeeping: Do not ask for an email address before you provide value.
• Answer the hard questions: It needs to handle complex queries, not just "schedule a viewing."

The standard for "value" has gone up.

A seller is on your site at 11 PM. They aren't looking for a generic "Free Home Valuation." They are stressed about specific problems.

They are asking: 

• "What are the capital gains tax implications if I sell my rental in Austin this year?"
• "How will the new high school rezoning affect my property value?"
• “How can I sell my home fast in this slow market?”

If your bot replies with "Please enter your email to chat," they bounce. You lost them.

If your bot replies with "Based on current Austin tax codes and your zone, here is the breakdown...", you win.

You establish trust by giving the answer away. You get the lead because you were the only one awake to help them.

Your sales chat should come to a natural conclusion (i.e., speak to the agent) or, you should follow up as soon as convenient to see if you can offer any more value.

Seller-intent landing pages

The "What is my home worth?" landing page is a vanity metric.

You cannot compete with Zillow on this. They have billions of data points, better engineers, and they own the consumer's attention on price. When an agent runs Facebook ads to a home valuation tool, they are usually just paying $15 a lead to generate a list of curious neighbors who aren't selling for three years.

So, stop burning cash on generic valuation traps.

To generate actual seller leads, you need to target trigger events in someone’s life.

People don't sell homes because they woke up and checked a Zestimate. They sell because of life transitions: Death, Divorce, Debt, Downsizing, or Relocation.

You need to build high-intent seller pages that capture them during the research phase of that transition.

Here are the three high-value assets you should build immediately:

Thousands of homeowners in your city are sitting on equity, terrified to sell because they have a 3% mortgage rate. They are debating renting it out vs. cashing out.

Build a breakdown of local rental yields vs. capital gains tax exposure.

"If you rent it for $2,500, you net $200/mo. If you sell, you net $150k tax-free. Here is the math."

The result? 

You capture the seller who is financially stuck.

Next, think about probate and inheritance.

When a parent dies, the children inherit a property they often don't want, filled with furniture they can't move, in a tax bracket they don't understand.

"Selling an Inherited Home in [City]: The Step-by-Step Guide." is the page you should create on your site. Deeply explain the "Step-up in Basis" tax rule (which Zillow won't tell them about) and they’ll start trusting you immediately.

These pages target long-tail searches. Which means there will be fewer of them happening. But, you’re competing for intent, not volume.

Use AI for predictive market growth

Most agents pick a farm area because they "like the houses" or it’s close to where they live.

If you are sending postcards to a neighborhood with a 2% turnover rate, you are setting money on fire. 

You are marketing to people who are statistically guaranteed not to move.

The top 1% of listing agents don't guess how to find sellers. They use AI to perform predictive prospecting. 

They know who is selling 6 months before the "For Sale" sign goes up by analyzing market signals that otherwise would’ve taken days of manual searching.

Take the "pre-listing" renovation spike, for example.

Homeowners rarely replace a roof or upgrade an HVAC system just for the fun of it. 

They do it to pass an upcoming inspection. By using AI tools to track local permit data, you can spot sudden clusters of exterior permits in specific subdivisions. It could signal a wave of deferred maintenance being fixed right before a wave of listings hits the market.

You can also use data to identify absentee owner fatigue.

Rental properties have a psychological lifespan. An out-of-state owner who bought in 2018 has seen massive appreciation, but now faces rising insurance costs and maintenance headaches. By filtering for out-of-state owners with high equity who have held the property for more than seven years, you aren't finding investors. 

You are finding tired landlords ready to cash out.

These are just two of the ways you can use AI to identify market trends. Tapping into a tool like Manus to do this research will dramatically speed up the process.

Once you identify these pockets, do not insult these homeowners with a recipe card.

Position yourself as the economist of that micro-market. When the homeowner finally decides to sell, they won't call a generalist; they will call the person who clearly knows more about their street than they do.

Should I buy seller leads?

The cost of buying seller leads is skyrocketing. In competitive markets, you are paying upwards of $60 to $200 per lead for contact info that has been sold to three other agents.

This is the "Zillow Tax” you all know and love.

You are renting an audience. The moment you stop paying, the pipeline dries up.

Investing in AI SEO is buying the building. It costs more upfront in time and effort. But once you rank in the AI Overviews, that traffic is free. It compounds.

Does outbound still work to generate seller leads?

Yes, outbound still works to find motivated sellers but if your business model relies on interrupting strangers to beg for attention, you are missing deals.

Privacy laws are tightening. Spam filters are aggressive. People under 40 do not answer unrecognized numbers.

Combine your outbound efforts with inbound marketing. This will, by default, make you easier to trust. Who is a seller more likely to trust:

• An agent who called them out of the blue with no reputation
• An agent who cold called and has 400 hours of YouTube content on their market

Inbound, by proxy, improves the conversion rates of seller leads.

Agent takeaway

The window is closing.

Right now, most agents are ignoring AI SEO because it sounds hard. They are hoping things go back to "normal."

They won't.

You have a choice. You can keep fighting for scraps in the "paid lead" shark tank, watching your margins erode.

Or, you can build a discovery engine. You can create the assets—the data-rich articles, the local guides, the brand mentions—that train the AI to see you as the only logical choice.

Months 1-3 will suck. You will write content that no one reads. You will feel stupid.

Months 4-6 will be quiet.

But by Month 12? 

You will have a pipeline of sellers who call you, ask for you, and trust you before you even say hello.

Build the engine.