Real Estate SEO in Australia: Getting Named When Vendors Ask AI

When a vendor asks ChatGPT or Google AI Mode who the best real estate agent in their suburb is, the answer is usually a shortlist of two or three agencies, assembled from portal sales data, rating-site reviews, Google reviews and the agencies' own suburb pages. Real estate SEO in Australia now means making those sources agree about you, keeping your site readable by AI crawlers, and publishing suburb content specific enough to be quoted - without breaching underquoting or misleading-advertising rules.

This guide is written for principals and marketing managers at Australian agencies. It covers which sources AI answers draw on and what each one contributes, a structure template for suburb report pages, sold data and RealEstateAgent schema, the state compliance rules that change this month (NSW from 17 November 2026, Victoria from 1 October 2026), a 60-day plan and a way to measure the result. Regulatory details were checked against NSW Government and Consumer Affairs Victoria pages on 2 November 2026.

How do AI answers choose which real estate agent to recommend?

AI assistants do not keep a ranking of agents. When search is switched on, they run several web searches behind the scenes, read the pages that come back and summarise the agencies those pages agree on. Google describes this for AI Overviews and AI Mode as query fan-out, issuing multiple related searches across subtopics and data sources, and it states that a page must be indexed and eligible for a snippet to be used at all. So the question is which pages those searches return for "best agent [suburb]", "agents who sold the most in [suburb]" and "[agency] reviews", and what they say.

In practice the sources fall into six types. Each one contributes something different to the final answer.

The pattern across agency-industry write-ups is the same: the assistant compresses the field to a few names. Sandy Abram of Foundable put it bluntly in coverage of its August 2026 agency audit, describing AI as a recommendation engine rather than a search engine. That is why the six source types matter more together than any one of them alone: the agencies that appear are usually the ones several independent sources agree on.

Why vendor searches matter more than buyer searches

Buyers already have a strong default: the portals. realestate.com.au went further in February 2026 and launched an app inside ChatGPT that lets logged-in users search homes for sale in plain language and contact the listing agent. For buyer journeys, the portal sits between the assistant and you.

Vendor and landlord questions are different. "Who should sell my house in Brunswick?", "which agent gets the best auction results in Paddington?" and "how do I choose a property manager in Newcastle?" have no single portal answer, so the assistant assembles one. These are the listing-winning prompts, and they are where an agency's own content and reputation decide whether it is named. With 48.6% of Australian adults having used generative AI at least once, according to the ANU and Google "AI Adoption in Australia" survey of more than 3,500 people (July 2026), a meaningful share of vendors will at least try the question.

What did AI Mode return for a suburb vendor prompt?

We ran one dated check rather than a study, and we report it as an observation, not a measurement. On 2 November 2026 we asked Google AI Mode "best real estate agent to sell my house in Marrickville NSW" from a signed-in browser outside Australia with the Australian region parameter set. AI Mode ran what it labelled a deep search and returned a short paragraph naming three agencies with offices in or near the suburb, justified by sales volume and local market knowledge, followed by a note that the right choice depends on whether the vendor wants a boutique local team or a larger office. We are not naming the agencies: one answer is not evidence that any agent is better or worse, and the same prompt can return different names an hour later. We also did not capture the cited source list in that session, so we make no claim here about which sites it drew on.

Two things from that check are still useful. The answer was a shortlist of three, which matches what agency-industry commentators describe. And the justification was framed around evidence the portals and rating sites publish (who sells the most, who knows the suburb), not around website design or advertising. If you want to see this for your own suburbs, run the check properly:

1. Write 8 to 12 vendor prompts per suburb, covering discovery ("best agent to sell a house in [suburb]"), comparison ("[your agency] vs [competitor] [suburb]"), situation ("agent for a deceased estate in [suburb]", "selling an apartment off the plan in [suburb]") and property management.
2. Run each prompt in ChatGPT, Gemini, Perplexity, Claude and Google AI Mode, logged out or in a fresh session where possible, and include the suburb and state in the prompt text.
3. Record four things each time: whether you are named, your position in the list, the competitors named and the sources cited.
4. Repeat the same prompts at least weekly for a month before drawing conclusions. Answers vary between runs, so a single run tells you little.

AI Rank Lab's brand visibility feature runs a prompt set like this across ChatGPT, Claude, Gemini and Perplexity when you run it and stores every run for comparison. The LLM runs do not emulate a location, so put the suburb and state in the prompt itself; for Google, the AI Mode tracker takes a country and language per query, so you can set Australia. For the wider pattern of local prompts, see our guide to "near me" queries in AI answers.

Can AI crawlers actually read your agency website?

Often they cannot, and it is the cheapest problem to fix. Foundable tested 100 of Australia's largest agency websites against 11 live AI crawler identities, including those used by ChatGPT, Claude and Perplexity. According to the Australian Broker report (26 August 2026), 35% failed outright, 25% loaded normally in a browser but blocked AI crawlers moments later, and 7% told AI crawlers to stay away in robots.txt. Eleven of the blocking sites ran the same WordPress SEO plugin version, which points to a default setting rather than a decision.

Blocking matters because the assistants need permission to retrieve pages at answer time. OpenAI's crawler documentation says that sites opted out of OAI-SearchBot will not be shown in ChatGPT search answers. A five-minute check:

1. Open yourdomain.com.au/robots.txt and look for Disallow rules under OAI-SearchBot, ChatGPT-User, PerplexityBot, Claude-SearchBot or Google-Extended.
2. Ask your web developer whether the firewall, CDN or security plugin has a "block AI bots" toggle switched on.
3. Confirm your suburb and sold pages render their text in the HTML rather than only after JavaScript loads.

Our list of AI crawler user agents explains which bot does what, so you can allow the search and answer bots while still making a deliberate choice about training crawlers.

What should a suburb report page contain to get cited?

A suburb page gets cited when it answers a vendor's question with specific, dated, sourced detail that the portals do not already provide. Most agency suburb pages are a paragraph of lifestyle copy and a listings widget, which gives an assistant nothing to quote. The structure below is what we would build for each suburb an agency farms.

Keep each suburb page genuinely different. A templated page per suburb with only the name swapped is what Google treats as scaled content, and it gives assistants nothing distinctive to cite. Ten deep pages for the suburbs you actually sell in beat 200 thin ones.

Sold data pages and Victoria's new disclosure rules

Sold archives are the most persuasive content an agency owns, and in Victoria they are now partly public by law. Consumer Affairs Victoria says that for properties sold after 1 October 2026, agents must update the Property Price Statement with the sale price within 7 days of the sale becoming unconditional and keep it publicly available for at least 18 months, unless an exemption applies. That creates a steady stream of verifiable sold prices tied to your agency. Link each statement from the matching sold page, so your site, the statement and the portal record tell the same story.

Which schema should an agency use: RealEstateAgent, Person or Review?

Use RealEstateAgent for each office, Person for each agent, and do not mark up your own reviews to chase stars. On schema.org, RealEstateAgent is a subtype of LocalBusiness, so it inherits address, areaServed, sameAs, aggregateRating and review. Google's local business documentation requires name and address and says to use the most specific LocalBusiness subtype possible.

Reviews are where agencies most often go wrong. Google's review snippet guidance says pages where the business controls the reviews about itself are ineligible for the star review feature. Marking up RateMyAgent or Google testimonials on your own site will not earn stars and can look manipulative. Link to the review platforms with sameAs instead, and let those platforms carry the rating.

A minimal office block looks like this:

{
  "@context": "https://schema.org",
  "@type": "RealEstateAgent",
  "name": "Example Realty Inner West",
  "url": "https://www.example.com.au/",
  "telephone": "+61 2 0000 0000",
  "address": {
    "@type": "PostalAddress",
    "streetAddress": "1 Example St",
    "addressLocality": "Marrickville",
    "addressRegion": "NSW",
    "postalCode": "2204",
    "addressCountry": "AU"
  },
  "areaServed": ["Marrickville NSW", "Dulwich Hill NSW", "Petersham NSW"],
  "sameAs": [
    "https://www.realestate.com.au/agency/...",
    "https://www.domain.com.au/real-estate-agencies/...",
    "https://www.ratemyagent.com.au/real-estate-agency/..."
  ]
}

Two caveats. Google states that there are no additional requirements to appear in AI Overviews or AI Mode, so schema is not a ticket into AI answers; its value is making your entity unambiguous (which office, which suburbs, which profiles are yours). And the sameAs list only helps if those profiles carry the same name, address and phone number. AI Rank Lab's free schema generator produces this JSON-LD if you would rather not hand-write it.

Which portals and rating sites matter, and what should you fix on each?

Fix the profiles AI answers lean on before writing new content, because they carry the third-party proof your own site cannot. Here is what each contributes, as published on 2 November 2026.

• realestate.com.au. The largest portal and, since February 2026, present inside ChatGPT for buyer searches. Make sure every agent profile is complete, uses the same agency name as your website and lists the suburbs you actually sell in.
• Domain. Domain's agency profile pages carry sales statistics, and its suburb agent listings display review ratings supplied by RateMyAgent. Where an agent shows "sales statistics unavailable", that profile is giving AI answers nothing to work with.
• RateMyAgent. Its homepage claims 348,000 agent profiles and 2.7 million verified reviews, each tied to a property transaction, and it publishes state-level top agent lists. Claim every agent's profile and ask for a review after every settlement, not just the happy ones.
• OpenAgent. A comparison site that ranks agents using sales data and is paid by agents through referrals. Its suburb pages are written to answer exactly the "best agent" question, so they appear in AI retrieval; vendors may not realise the commercial model, which is worth knowing when you see it cited.
• Google Business Profile. Reviews, categories and service areas feed both Maps and AI Mode. Our Google Business Profile guide for Australian businesses covers setup and verification.

Beyond the property sites, general Australian directories also get cited in local AI answers; our small study of which Australian sites AI answers cite shows which ones turn up. If an assistant states something wrong about your agency (an old office address, a departed agent, a merged brand), trace it to the source and correct it there, as set out in our guide to fixing wrong AI answers about your brand.

Compliance: underquoting, reviews and AI-written content

Anything your agency publishes about the price of a property it is selling is regulated advertising, whether a person or an AI tool wrote it. Two states changed their rules this spring, and both affect suburb and sold content. This is a summary, not legal advice; check the regulator in your state before publishing.

Queensland, South Australia, Western Australia, Tasmania, the ACT and the Northern Territory have their own rules on price advertising and auctions. Check your state or territory fair trading office before reusing NSW or Victorian templates.

Where AI-written content creates risk

• Price language in suburb copy. An AI tool asked to write "a suburb page that attracts vendors" will happily write "homes on this street are selling for over $2 million". If that sentence sits next to a current listing, it can read as a price representation for that property. Keep live-listing prices to the figure in the agency agreement.
• Unsupported superlatives. "Marrickville's number one agent" is a factual claim. Under the Australian Consumer Law it needs evidence (a stated period, a stated measure, a source), and assistants repeat such claims without the caveats.
• Reviews and testimonials. The ACCC says online reviews must be genuine and treats suppressing negative reviews as potentially as misleading as posting fake ones. Google's Maps policy separately bans incentivised reviews and says businesses must not discourage negative reviews or selectively solicit positive ones. Send the review request to every vendor after settlement.
• Personal information in sold archives. Publish sold results the way your agency agreement and privacy policy allow, and respect Victorian exemptions for family violence situations.

A 60-day plan for an agency

Sixty days is enough to fix access, profiles and a first set of suburb pages, and to take a second measurement. It is not enough to judge the final effect, because portals, reviews and AI answers update at different speeds.

1. Days 1-7: baseline. Pick your 5 to 10 core suburbs. Write the vendor prompt set for each and run it across ChatGPT, Gemini, Perplexity, Claude and AI Mode. Record mentions, positions, competitors and cited sources. AI Rank Lab's brand visibility and citation tracker can store these runs so the day-60 comparison is like for like.
2. Days 8-14: access and entity. Fix robots.txt, firewall and plugin blocks. Make the agency name, address and phone number identical across website, Google Business Profile, realestate.com.au, Domain and RateMyAgent. Add RealEstateAgent and Person schema with sameAs links.
3. Days 15-21: profiles. Complete every agent profile on the portals and RateMyAgent, with suburbs, bios and licence details. Remove departed agents or mark them clearly.
4. Days 22-45: suburb pages. Rebuild the core suburb pages with the template above, two or three a week, each with sourced market data, your sold results and vendor FAQs reviewed against your state's price rules.
5. Days 22-60: reviews. Turn on a post-settlement review request to every vendor and buyer, split across Google and RateMyAgent, and reply to every review, including the critical ones.
6. Days 30-60: third-party proof. Offer the local paper or property press a quarterly market comment with real numbers. Coverage that quotes your agents by name is one of the source types assistants trust.
7. Days 55-60: re-measure. Re-run the same prompt set on the same engines and compare. Note which sources started citing you and which prompts still return only competitors.

How do you measure whether AI search is sending you vendors?

Measure three layers: whether you are named, whether that produces visits, and whether visits become appraisal requests. No single tool covers all three.

For the visibility layer, the options are:

Disclosure: AI Rank Lab is our product. We list it first because this is our site, and we've kept the same criteria for every tool below.

• AI Rank Lab runs your prompt set across ChatGPT, Claude, Gemini and Perplexity, plus Google AI Overviews and AI Mode with an Australian location, records mentions, positions, competitors and cited URLs for each run, and connects GA4 for referral traffic. It is priced in USD, from $69 a month. Runs happen when you trigger them rather than on an automatic schedule.
• A spreadsheet and a calendar reminder. Free and perfectly adequate for one office and five suburbs, if someone actually runs the prompts every week and logs the answers consistently.
• Other AI visibility trackers such as Otterly.AI and Peec AI do the same core job and are worth comparing if you manage many offices or already use them for another brand.

Whatever you use, judge results over weeks, not single answers, and expect AI-driven appraisal requests to stay a small share of total leads for now. The point of measuring early is to see the trend before your competitors do.

If you want to see which agencies AI assistants name for your suburbs today, run your vendor prompts in AI Rank Lab and use the results as your day-one baseline.