By Zero per cent exact-entity agreement. That is what we found when we asked ChatGPT, Perplexity, and Google Gemini the same question: who is the best agent in Sydney? We tested 18 buyer prompts across all three AI engines, running 54 total queries in clean-room conditions with no login, no personalisation, and no conversation history. Not one agency appeared under the same name on all three engines. Through manual brand-level matching, we identified three brands present across all three — Sydney Sotheby's International Realty, Curtis Associates, and Ray White Inner West Group — but each engine named them differently. Of 254 agencies named in total, 92 per cent appeared on a single engine only.
Data collected: March 2026.
- 0% exact-entity agreement — 3 brands identifiable across all engines only through manual matching
- 92% single-engine — visible on one AI tool, invisible on the other two
- Suburb prompts reset the list — zero overlap between generic and location-specific results
- Boutique agencies dominated need-specific and suburb-specific queries over franchises
What Happens When You Ask AI for a Real Estate Agent in Sydney?
AI engines return specific agency names when Sydney buyers ask for real estate agent recommendations. Every one of our 54 queries produced named agencies. But the lists differ dramatically across ChatGPT, Perplexity, and Gemini. Of 254 unique agencies named in total, 234 appeared on only one engine. That is 92 per cent of agencies showing up in a single engine's results and nowhere else.
92 per cent of the agencies AI named for Sydney buyers appeared on only one engine. Showing up on ChatGPT tells you nothing about what Gemini or Perplexity will say.
How We Tested AI Real Estate Agent Recommendations
Eighteen prompts per engine. Six categories: generic recommendations, need-specific (first home buyers, investors, sellers, international buyers), comparative, suburb-specific, advisory, and cost/fees. Three city-specific prompts: off-plan apartments, luxury property, Inner West agencies. Fresh API call every time — no login, no history, no conversation carry-over.
| Engine | Prompts | Total Citations | Unique Domains |
|---|---|---|---|
| ChatGPT (OpenAI) | 18 | 89 | 55 |
| Perplexity | 18 | 132 | 92 |
| Gemini (Google) | 18 | 184 | 112 |
Gemini's 184 citations versus ChatGPT's 89. That gap is structural. These engines do not share a source list. Different domains, different signals, different answers to the same question.
Which Sydney Agencies Does ChatGPT Recommend?
ChatGPT recommended specific Sydney agencies for all 18 prompts, pulling 89 total citations from 55 unique domains. The distinguishing feature: 35 per cent of those citations were Google Maps links. That makes Google Business Profiles a primary data source for ChatGPT's real estate recommendations. When ChatGPT names an agency, it often includes the star rating, review count, address, and a direct Google Maps link.
For generic "best agent" prompts, ChatGPT surfaced a mix of franchises and independents. Levy Property Group, McGrath, Ray White, LJ Hooker, The Agency, CBRE, and Home7 all appeared. The response typically read like a curated list with review data attached: "5.0 stars, 8 reviews."
The buyer intent mattered. First home buyer prompts (B1) pushed ChatGPT toward buyer's agents: Brady Marcs and Propertybuyer replaced the franchise names entirely. Luxury prompts (S2) produced a different set again: Kincade Homes, Sydney Prestige, and Black Diamondz.
| Agency | Prompt Category | Context Provided |
|---|---|---|
| Levy Property Group | Generic, Need-specific, City-specific | Google Maps rating, review count |
| McGrath Estate Agents | Generic | Branch location, agent names |
| Brady Marcs | First home buyer | Buyer's agent, Google Maps link |
| Black Diamondz (Monika Tu) | Luxury, City-specific | Named principal, prestige market |
| Pillinger | Suburb-specific (Eastern Suburbs) | Google Maps link, reviews |
ChatGPT's top cited domains tell the same story. google.com led with 31 citations, followed by irec.com.au (3), whichrealestateagent.com.au (2), and sydneysothebysrealty.com (2). The source mix: 57 per cent agency websites, 35 per cent Google Maps, 3 per cent directory and review sites.
For suburb-specific prompts, ChatGPT named Pillinger for the Eastern Suburbs and Schwarz Real Estate and Domain Residential for the Northern Beaches. The off-plan prompt (S1) surfaced Leading Capital Group. These names appeared only on ChatGPT and nowhere else, reinforcing the single-engine visibility problem. An agency recommended by ChatGPT has no guarantee of appearing on Perplexity or Gemini.
From our data
31 of ChatGPT's 89 citations pointed to Google Maps. No other engine used Google Maps as a source at all. If your Google Business Profile is not complete, ChatGPT likely cannot see you.
Which Sydney Agencies Does Perplexity Recommend?
Perplexity generated 132 citations across 92 unique domains for Sydney real estate agent queries. It pulls heavily from directory and review platforms. WhichRealEstateAgent appeared 6 times. LocalAgentFinder, 5. RateMyAgent, RealEstateBusiness, and several agency sites each appeared 4 times. Where ChatGPT leaned on Google Maps, Perplexity leaned on third-party platforms that rank and review agents.
That 20 per cent directory/review share is the highest of any engine. Perplexity also cited award data explicitly — REB Top 100 honourees, "Agent of the Year" winners named by title. One response mentioned a "$50.1 billion portfolio" as context. Reputation data matters here. Structured, third-party, verifiable reputation data.
For generic prompts, Perplexity recommended Ray White, Morton-Sydney City, Ayre Real Estate, DiJones, and Sydney Sotheby's International Realty. It also included agencies that the other engines ignored: CST Properties, Palladium Property, and Dexus appeared in mid-tier results.
| Prompt Category | Agencies Named | Notable Pattern |
|---|---|---|
| First home buyer (B1) | Propertybuyer, Cohen Handler, PK Property | Buyer's agents dominate. Selling agents disappear. |
| Investment (B2) | inSynergy, Meridian Australia, The Investors Agency, Rethink Investing | Specialist advisory firms. None appear for generic. |
| Luxury (S2) | Sydney Sotheby's, The Rubinstein Group | Award citations. Historical context ("since 1902"). |
| Inner West (S3) | Hudson McHugh, Ray White Inner West, Adrian William | Suburb-specific content drove these results. |
Perplexity generated 88 related questions from our 18 prompts — questions it suggests to users after delivering an answer. Examples: "How do the fees of different buyer's agents compare for first-home buyers?" and "Which real estate agents in the Eastern Suburbs specialize in property management?" Eighty-eight of them. That is 88 prompts your next client might already be typing.
Which Sydney Agencies Does Gemini Recommend?
Gemini produced the broadest coverage: 184 citations from 112 unique domains across our 18 prompts. That is more than double ChatGPT's 55 domains. Gemini also exhibited a behaviour the other engines did not: it named individual agents within agencies, complete with their suburb. Alexander Phillips (Phillips Pantzer Donnelley, Woollahra). John Paranchi (McGrath, Hunters Hill). Adrian Bo (McGrath, Coogee). Dib Chidiac. Allen Yan (Ray White). The response format felt like a referral from someone who knows the market block by block.
Where did Gemini find these names? The top cited domains: salebyhomeowner.com.au (8), top10realestateagent.com.au (7), top3realestateagents.com.au (7), realestate.com.au (7), ratemyagent.com.au (6), oneflare.com.au (5). Its source mix: 65 per cent agency sites, 17 per cent directory/review platforms, 8 per cent consumer guides, 6 per cent portals, 2 per cent Reddit, 2 per cent government sources.
Reddit. Four citations from Reddit appeared in Gemini's responses. Neither ChatGPT nor Perplexity cited Reddit once. For advisory prompts, Gemini also pulled from consumer.nsw.gov.au — the NSW government's consumer affairs site. Wider net. More diverse sources. More ways in.
A pattern we noticed
Gemini was the only engine to name individual agents rather than just agencies. If you are an agent building a personal brand, Gemini is the engine most likely to surface your name specifically.
Gemini decomposed our prompts into 57 fan-out queries, the search queries it ran internally to build each response. Examples: "best real estate agents Eastern Suburbs Sydney" and "Sydney real estate agents specializing in overseas clients." For agencies, these fan-out queries are a roadmap for content gaps.
Do the Three AI Engines Agree on Who's Best in Sydney?
The three AI engines agree on almost nothing. Of 254 agencies named across all responses, exact-entity agreement was 0 per cent — no agency appeared under the same name on all three engines. Through manual brand-level matching, we identified 3 brands that appeared on all three engines under slightly different names or office contexts. Another 17 brands appeared on two engines. The remaining 234 (92 per cent) were recommended by a single engine only.
| Agency | Total Mentions | Prompt Categories |
|---|---|---|
| Sydney Sotheby's International Realty | 6 | Generic, Need-specific, City-specific |
| Curtis Associates | 3 | Need-specific (international buyers) |
| Ray White Inner West Group | 3 | City-specific (Inner West) |
Sydney Sotheby's International Realty had the broadest reach: 6 total mentions across generic, need-specific, and city-specific categories. Curtis Associates appeared only for international buyer prompts, but it appeared on all three engines for that niche. Ray White Inner West Group appeared for Inner West-specific queries across all three.
The 17 agencies on two engines included Propertybuyer (Gemini + Perplexity), Cohen Handler (Gemini + Perplexity), DiJones (OpenAI + Perplexity), McGrath Estate Agents (Gemini + OpenAI), and Black Diamondz Property Concierge (Gemini + OpenAI). The full list:
| Agency | Engines | Categories |
|---|---|---|
| Propertybuyer | Gemini, Perplexity | Need-specific, City-specific |
| Levy Property Group | ChatGPT, Perplexity | Generic, Need-specific, City-specific |
| Morton-Sydney City | Gemini, Perplexity | Generic, Need-specific, City-specific |
| Ayre Real Estate | Gemini, Perplexity | Generic, Need-specific, City-specific |
| Cohen Handler | Gemini, Perplexity | Need-specific, City-specific |
| DiJones | ChatGPT, Perplexity | City-specific, Suburb-specific |
| The Investors Agency | Gemini, Perplexity | Need-specific |
| PK Property | Gemini, Perplexity | Need-specific |
| BFP Property Group | Gemini, Perplexity | Need-specific |
| Coposit | Gemini, Perplexity | City-specific |
| The Rubinstein Group | ChatGPT, Perplexity | City-specific |
| McGrath Estate Agents | Gemini, ChatGPT | Generic |
| Ray White | Gemini, ChatGPT | Generic |
| Black Diamondz Property Concierge | Gemini, ChatGPT | Generic, City-specific |
| Rose & Jones | Gemini, ChatGPT | City-specific |
| SydneySlice | Gemini, Perplexity | Need-specific, City-specific |
| BuyEast | Gemini, Perplexity | Suburb-specific |
The sources disagreed too. We tracked 198 unique domains across all engines. Of those, 147 (74 per cent) appeared in only one engine. Ten domains were cited by all three:
cohenhandler.com.au, curtisassociates.com.au, irec.com.au, levypropertygroup.com.au, propertybuyer.com.au, ratemyagent.com.au, sydneysothebysrealty.com, sydneyvendoradvocates.au, titlespace.com.au, whichrealestateagent.com.au.
Ten out of 198. Visible on one engine? That tells you almost nothing about the other two.
Interactive: Search for your agency
Data from our 54-query experiment. See all free tools.
Does the Question Change the Answer? How Buyer Intent Shifts AI Recommendations
The question changes the answer completely. A buyer asking "best agent in Sydney" sees franchise names. A buyer asking "best buyer's agent for a first home" sees a different set of agencies with zero overlap. We tested six prompt categories and three city-specific variations. The agencies recommended shifted with every change in buyer intent.
How Do AI Recommendations Change for First Home Buyers in Sydney?
We changed one word in the prompt — from "best agent" to "first home buyer" — and the entire recommendation list reset. Propertybuyer. Cohen Handler. BFP Property Group. PK Property. All buyer's agents. Not one selling agent survived the switch. None of these four appeared in generic or investment results either. The match is precise: say "first home buyer" on your website, and AI connects you to the buyer typing those words.
Change one word in the prompt — "best agent" to "first home buyer" — and the entire recommendation list resets. Not one selling agent survived the switch.
Which Real Estate Agents Does AI Recommend for Sydney Property Investment?
Different story for investment. inSynergy, Meridian Australia, The Investors Agency, Rethink Investing. A completely different set. None appeared for generic or first-home-buyer prompts. The sources shifted too: Perplexity pulled from industry publications, Gemini from portals and review platforms. If you focus on investors but AI cannot see that, you are invisible to the exact buyers you serve.
| Prompt Type | Agency Types Recommended | Example Agencies | Key Observation |
|---|---|---|---|
| Generic ("best agent") | Franchises, large independents | Ray White, McGrath, LJ Hooker, Belle Property | Brand recognition drives generic results |
| First home buyer | Buyer's agents | Propertybuyer, Cohen Handler, PK Property | Selling agents disappear entirely |
| Investment | Advisory firms, specialists | inSynergy, Meridian, Rethink Investing | Zero overlap with generic results |
| Luxury | Prestige specialists | Sydney Sotheby's, Black Diamondz, The Rubinstein Group | Named principals (e.g., Monika Tu) |
| Suburb-specific (Eastern Suburbs) | Local specialists | Phillips Pantzer Donnelley, BuyEast, Pillinger | Completely different lists per engine |
| Suburb-specific (Northern Beaches) | Local specialists | Cunninghams, Clarke & Humel, Schwarz Real Estate | Different agencies per engine again |
Suburb-specific prompts produced the widest divergence. Eastern Suburbs: Gemini named Phillips Pantzer Donnelley (through agent Alexander Phillips). Perplexity picked BuyEast. ChatGPT listed Pillinger. Three engines, three completely different agencies. Northern Beaches was the same: Cunninghams and Clarke & Humel on Perplexity, Schwarz Real Estate on ChatGPT, Ben Spackman (Raine & Horne) on Gemini. Zero overlap.
Cost and fees prompts? All three engines agreed: 1.8 to 2.5 per cent average commission, 2.1 per cent as the Sydney-specific average. Perplexity gave the most granular breakdown — tiered structures, fixed versus percentage, marketing cost separation. Sources overlapped more here than in any other category: whichrealestateagent.com.au, irec.com.au, finder.com.au across multiple engines.
Comparative prompts ("buyer's agent vs regular agent") were mostly educational. ChatGPT and Gemini gave pros-and-cons without naming anyone. Perplexity named names: Propertybuyer, buyersagencyaustralia.com.au. Want to be cited in a comparison? Perplexity is where it happens.
Why Are Most Sydney Real Estate Agencies Invisible to AI?
Of the 254 agencies our experiment surfaced, 234 appeared on only one engine. That means 92 per cent of named agencies have, at best, partial AI visibility. And those are the ones that showed up at all. The vast majority of Sydney agencies were not named by any engine for any prompt.
We did not set out to study what makes agencies visible. But after reviewing 54 responses and 405 total citations, patterns emerged in the agencies that appeared on two or more engines.
What Do AI-Recommended Agencies Have That Others Don't?
Five characteristics repeated across the 20 agencies on multiple engines.
| Factor | How It Showed Up | Which Engines? |
|---|---|---|
| Review platform presence | RateMyAgent, WhichRealEstateAgent, LocalAgentFinder, OpenAgent among top cited domains | All 3 |
| Award recognition | REB Top 100, RateMyAgent Awards, "Agent of the Year" cited by name | Perplexity, Gemini (not ChatGPT) |
| Specialisation language | Curtis Associates states "international buyers" on site — only agency on all 3 for that prompt | All 3 |
| Suburb-level content | No suburb pages = invisible for D1/D2 prompts. Direct relationship. | All 3 |
| Crawlable website | 60–76% of all citations across engines pointed to agency websites directly | All 3 |
ChatGPT
Perplexity
Gemini
Review platforms were the single most consistent factor. Award data helped on Perplexity and Gemini but ChatGPT ignored it entirely. Specialisation language had the sharpest effect: state what you do, and AI matches you to the buyer asking for it. Say nothing specific? You get generic prompts. If any.
Franchise vs. Boutique
In the top 50 most-mentioned agencies, franchise branches accounted for 22 per cent of mentions. Boutique and independent agencies accounted for 78 per cent. Franchises dominated generic prompts. For need-specific, suburb-specific, and city-specific queries, independents and specialists took over.
What Does This Mean for Sydney Real Estate Agents?
Buyers are already using AI to find agents. BrightLocal's 2026 Local Consumer Review Survey found that 45 per cent of consumers now use AI for service recommendations. That includes real estate.
So the question is not whether this matters. It does. The question is whether your agency shows up when a Sydney buyer asks.
Our 54 queries produced three clear findings.
| Finding | Data | What It Means |
|---|---|---|
| Each engine builds its own list | 74% of source domains appear on only 1 engine. 0% exact-entity agreement. | Visibility on ChatGPT tells you nothing about Perplexity or Gemini. |
| Intent resets the results | Zero overlap between generic, first-home-buyer, and investment results. | Specialist positioning wins intent-specific prompts. Generic positioning wins generic prompts only. |
| The sources are identifiable | 10 domains cited by all 3 engines. Review platforms most consistent factor. | Not a black box. We mapped every source each engine relies on. |
| If you... | AI sees you as... | What to do |
|---|---|---|
| Have no suburb-specific content | Invisible for location queries | Create one page per suburb you operate in |
| Use generic "full-service" language | A generic result (if any) | Add specialisation language matching buyer intent |
| Are on portals but not review platforms | Visible to Gemini, harder for Perplexity to find | Claim your RateMyAgent and WhichRealEstateAgent profiles |
| Have a strong Google Business Profile | ChatGPT-visible, but that alone does not carry to other engines | Google Maps citations ≠ Perplexity or Gemini citations |
Calculator: What is AI invisibility costing you?
Estimate based on market data and AI adoption rates. See all free tools.
We ran this same experiment in Melbourne and Perth. The cross-city patterns are striking: 12 brands appeared on all three engines across all three cities, and not one was a franchise. Read the full cross-city analysis.
The Bottom Line
Each AI engine builds its own recommendation list from its own sources. Most Sydney real estate agencies are not on any of them. Of 254 agencies named across 54 queries, 92 per cent appeared on only one engine. Three appeared on all three. The sources that AI relies on are identifiable, the signals are measurable, and the gaps are fixable.
Want to see where your agency stands? Take our 2-minute AI readiness assessment, or request your free AI visibility report.
Frequently Asked Questions
Which real estate agency in Sydney has the best AI visibility?
Sydney Sotheby's International Realty appeared on all three AI engines with the highest mention count (6 total) across generic, need-specific, and city-specific prompts. Curtis Associates and Ray White Inner West Group also appeared on all three, but for narrower query types.
Does ChatGPT recommend real estate agents?
Yes, with names, star ratings, and Google Maps links.
Are there any boutique real estate agencies in Sydney worth considering?
Boutique and independent agencies accounted for 78 per cent of mentions in our top 50 most-mentioned list. For need-specific and suburb-specific prompts, boutique agencies dominated over franchises. Curtis Associates, a boutique firm specialising in international buyers, appeared on all three engines. Cohen Handler, Propertybuyer, and PK Property (all independents) appeared on two. Franchises like Ray White, McGrath, and LJ Hooker led the generic "best agent" results, but the moment a buyer added a specific need or suburb, the smaller firms took over. Specialist positioning gives boutique agencies a measurable advantage in AI recommendations when buyers have a clear intent.
How do I find the best real estate agent for my specific suburb in Sydney?
Ask ChatGPT, Perplexity, and Gemini separately. Our Eastern Suburbs and Northern Beaches tests produced completely different lists on each engine, with zero overlap.
How do commission rates in Sydney compare across AI recommendations?
All three engines agreed: 1.8 to 2.5 per cent is the typical range, with 2.1 per cent as the Sydney average. Perplexity provided the most detailed breakdown, including tiered commissions and marketing cost separation.
Can a real estate agent also act as a buyer's agent?
Legally, yes. In practice, AI engines treat them as separate categories. When we asked about buyer's agents, selling agents disappeared from results entirely. The engines clearly distinguish between the two roles based on how agencies describe themselves online.
Sources
- Cited Research, Sydney AI recommendation experiment, March 2026. 54 queries across ChatGPT, Perplexity, and Gemini. Raw data collected via API.
- BrightLocal, 2026 Local Consumer Review Survey. Referenced for 45 per cent AI usage statistic.
- RateMyAgent, WhichRealEstateAgent, LocalAgentFinder, OpenAgent — Australian real estate agent review and comparison platforms referenced in AI engine citations.
- NSW Fair Trading — government regulator cited by Gemini for advisory prompts.
- REB (Real Estate Business) — industry awards data cited by Perplexity.