Dubai market · portals

What Bayut, Property Finder and dubizzle hold, and where a brokerage site can compete

I'm Aaron Zara. I built and run REN.PH, a Philippine property data platform holding 238,203 zonal value rows across 38,272 as of 21 August 2026, so portal-scale programmatic publishing is something I've operated rather than audited from outside. I also hold a PRC real estate broker licence (Philippines, 2015), and I run RealEstateSEO.ph, which does search and AI-answer visibility work for real estate businesses.

To be clear about standing: that licence is Philippine and carries no authority in the UAE, where RERA registration is what matters. I don't broker in Dubai and I'm not offering to. What transfers is the measurement work and the platform experience. RealEstateSEO.ph operates from the Philippines, which is also why the cost base sits below a Dubai or London agency retainer.

This page tests a claim I made on the Dubai market hub: the portals hold the transactional queries, and a brokerage should compete for the advisory layer instead. The evidence below sharpens that into something more specific and more usable.

01Method

Research snapshot

Market: Dubai
Checked: 4 September 2026
Listing counts: Read directly from the portals' own community pages
Search observations: External search API, run from outside the UAE, 4 September 2026
Transaction data: DLD H1 2026, via DLD releases and third-party analyst reports
Traffic estimates: Similarweb and Semrush public pages, dates noted per figure
Key limitation: Search observations are directional. They are not Dubai-localised Google rankings and should not be read as such.

Everything in the search-observation category below is phrased as "did not appear in this search run" rather than "does not rank," and the distinction is load-bearing. Re-running the same queries on the same instrument reproduces the same result sets, which confirms the instrument is stable and confirms nothing about what a buyer in Jumeirah sees. Anyone acting on this should re-run the queries logged out from a Dubai IP before committing budget.

The listing counts are a different class of evidence. Those came from the portals' own pages and are solid.

02The inventory

What each portal actually holds

Original observation, 4 September 2026

On the same query, same day, Property Finder listed 3,554 Dubai Marina apartments for sale, dubizzle 3,141 and Bayut 3,123, against 100 on Betterhomes, one of the emirate's largest brokerages.

Site	Apartments for sale, Dubai Marina	Source
Property Finder	3,554	propertyfinder.ae listing page count
dubizzle	3,141	dubai.dubizzle.com ad count
Bayut	3,123	bayut.com listing page count, confirmed in its own FAQ block
Betterhomes (brokerage)	100	bhomes.com pagination, "Showing 1 - 16 of 100 properties"

The gap is real, and the reason for it is worth stating precisely, because the obvious reading is wrong.

Betterhomes isn't a 100-listing firm. It has been operating since 1986 and publishes "300+ community specialists" on its own homepage. That 100 is one community and one property type. And those 100 units sit inside Property Finder's 3,554, because Betterhomes lists on the portals. Betterhomes appeared repeatedly as the listing agency on both the Property Finder and Bayut Dubai Marina pages I pulled.

The aggregation problem, stated plainly

Property Finder: 3,554 Dubai Marina apartments
    |-- includes Betterhomes: 100
    |-- includes every other listing agency in the community

This isn't 3,554 rivals against 100. It's an aggregator holding a pool that already contains the brokerage's own stock. A brokerage can't out-supply a page that includes its own supply, and that holds regardless of how large the agency gets.

The three portal counts land within 14 percent of each other, which is consistent with heavy cross-listing, though count similarity alone doesn't prove it. Three portals could each hold 3,000 distinct units and produce the same totals. Demonstrating overlap needs listing-level sampling, which is separate work and gets its own page in this cluster.

Bayut

Bayut is part of Dubizzle Group. Its content operation runs three separate guide systems visible in site navigation as of 4 September 2026: Area Guides, Building Guides and School Guides, plus the MyBayut editorial hub, TruEstimate valuations and a public Dubai Transactions tool fed by Dubai Land Department data.

Its Dubai Marina apartment page carries 18 FAQ answers with figures attached. On 4 September 2026 it reported an average asking price of AED 3,993,035 (about USD 1.09 million) from 6 months of its own listings, and separately an average sold price of AED 3,285,997 (about USD 895,000) across 2,248 apartment transactions recorded by DLD over the previous 12 months.

Bayut's AI position is the part most Dubai brokerage owners haven't priced in. It launched an app on ChatGPT on 1 June 2026, reported as the first UAE-based real estate platform to do so. On 27 August 2026 it added conversational AI Search on its own portal. BayutGPT, its on-site assistant, dates to 2023. Those are three separate products and they are worth keeping apart.

Worth being precise about what a ChatGPT app is and isn't. It doesn't fire ambiently on every property question. A user has to discover it or invoke it. What it does is put Bayut's live inventory inside the interface as a callable option, which is a different competitive surface from a page in a ranked list, and one a brokerage has no equivalent access to.

Property Finder

Property Finder raised USD 170 million led by Mubadala on 27 January 2026, alongside another UAE sovereign fund and BECO Capital, plus USD 250 million in debt from Ares Management and HSBC at the same time. Freshfields, which advised on the deal, published its own account on 29 January 2026. That followed a USD 525 million round led by Permira with Blackstone Growth on 9 September 2025, after which the company's valuation was reported to exceed USD 2 billion.

The relevant point for a brokerage is the publishing capacity that buys, not the deal history. Nearly USD 700 million of equity is going into the same result pages you're trying to appear on.

It held the largest Dubai Marina apartment inventory of the three on 4 September 2026 at 3,554 units. Its navigation carries Area Insights, Community Guides, Tower and Compound Guides, School Guides, a blog, market reports, a mortgage calculator, a rent-vs-buy calculator, sold-price data and resident reviews. The Dubai Marina page alone links to over 100 individual tower pages.

Its price insights on that page, same date, gave an average of AED 2,908,161 (about USD 792,000), an average of AED 2,190 per square foot (about USD 596), and a rental yield of 5.97 percent. The page states these are drawn from Property Finder's last 12 months of listing data.

dubizzle

dubizzle is the classifieds side of the same group as Bayut, and it carried 3,141 Dubai Marina apartment ads on 4 September 2026. Its property section shares TruEstimate with Bayut and runs an agent and agency directory, and its property blog carries categorised advisory content including area insight pieces and a laws and regulations category. What it doesn't have, in navigation, is a dedicated area-guide or building-guide hub of the kind Bayut and Property Finder both run.

Its property section also sits inside a marketplace running motors, jobs and classifieds, which produces a different authority profile than a pure property domain.

On "which portal is biggest," a caution

The tools disagree, and they disagree with themselves month to month. Similarweb's UAE Real Estate ranking puts bayut.com first, but it files dubizzle.com under Classifieds and dubai.dubizzle.com under Jobs and Employment, so one of the three portals under discussion isn't in that comparison at all.

Both category pages overwrite themselves every month and neither is in the Wayback Machine, so the three estimate rows below cannot be re-opened by a reader. They are recorded here as read on the date given. Only the listing-count row can be checked today, and that is the point of the section.
Source, period	Order	Re-openable
Similarweb, UAE Real Estate, Dec 2025	bayut.com, propertyfinder.ae, danubehome.com. dubizzle absent from category	No
Semrush, UAE Real Estate, Apr 2026	dubizzle.com 2.33M, bayut.com 1.68M, propertyfinder.ae 1.45M	No
Semrush, UAE Real Estate, Jun 2026	bayut.com 3.61M, dubizzle.com 3.04M, propertyfinder.ae 2.55M	No
Listing count, Dubai Marina apartments, 4 Sep 2026	Property Finder 3,554, dubizzle 3,141, Bayut 3,123	Yes

Anyone deciding where to put portal budget on the strength of a single third-party traffic estimate is deciding on noise. Count the inventory in the communities you actually sell in, and ask each portal for lead data on your own listings.

03The two contests

Two different competitions, and the real reason for the gap

Three query shapes, run through the search API described above on 4 September 2026.

Transactional

"apartments for sale in Dubai Marina"

Returned Bayut, Property Finder and dubizzle, alongside JamesEdition, Zoopla, Luxhabitat, Betterhomes and Dacha.

Advisory, cost

"Dubai service charges per square foot 2026"

Every result in the returned set was a brokerage or agency site, including Luxhabitat, Engel and Volkers, Driven Properties, Excel Properties, Pearlshire, EGSH and Luxury Property. None of the three portals appeared in this search run.

Advisory, eligibility

"can foreigners buy property in Dubai freehold areas rules"

Returned the UAE government platform at u.ae first, then Betterhomes, Engel and Volkers, McCone Properties, Mavrix, LYM Real Estate, Janus Developers and Kotook. Again, none of the three portals appeared in this search run.

The obvious reading of that absence is that the portals never wrote the material. That reading is wrong, and it's checkable in about ninety seconds.

Bayut has a full service charges guide on MyBayut covering what service charges are, the DLD index, how to look a building up three different ways, and the AED 3 to AED 30 per square foot range. Property Finder has a service charges guide on its blog covering the same ground. The content exists.

What's also true: the MyBayut guide is indexed with a 2024 date and its own table of contents still reads "How to Use Service Charge Index Dubai 2024." The portals wrote this material and moved on.

Meanwhile Luxhabitat's 2026-updated guide gives Burj Khalifa at approximately AED 67.88 per square foot and Downtown Dubai averaging around AED 21, tower by tower. That AED 21 is an average inside a stated range of AED 17 to over AED 40, with ultra-prime towers reaching AED 65 to 68, so it is the midpoint of a wide spread rather than a representative figure.

So the gap is maintenance and specificity, not subject matter. A 2024-dated general explainer from a large domain did not appear in this search run, while 2026-dated tower-level breakdowns from smaller domains did. That reading explains what the searches actually returned, it survives someone pasting a MyBayut URL into the comments, and it tells you what to do differently.

The strategic version, if you want one sentence: a brokerage doesn't beat a property portal by becoming a smaller property portal. It competes where the portal's aggregation model has no reason to keep the answer current.

One consequence worth accepting: the firms appearing on these queries are already doing this. Luxhabitat, Engel and Volkers, Driven and EGSH are all publishing tower-level and index-level cost content in 2026. The space is contested, not open.

04What to publish

What a Dubai brokerage should publish, in priority order

Tier 1: current, specific, and attached to an expensive decision

This is where the maintenance gap is widest and the buyer's money is most at risk.

Approved 2026 service charge rates for the specific towers you sell in. Dated, sourced to the DLD index, updated when the index updates. Not "service charges in Dubai run AED 10 to 30 per square foot," which Bayut, Driven and EGSH all publish already.
All-in acquisition cost on a named unit price. The 4% DLD transfer fee is on every site in the emirate. The total on a specific AED 2.5 million unit including agency fee, trustee fee, NOC and mortgage registration is not.
Current DLD transaction data for your communities, refreshed on a schedule you actually keep.
Off-plan versus ready framed by buyer type, using the correct dataset. See below.

Tier 2: interpretation of conflicting published data

Two separate disagreements sit in the Dubai Marina data above, and they're different in kind.

Comparison	Metric type	What explains it
Bayut AED 3,993,035 vs Property Finder AED 2,908,161	asking vs asking	6-month vs 12-month window, different listing mix
Bayut AED 3,993,035 vs Bayut's DLD figure AED 3,285,997	asking vs sold	listing prices vs recorded transactions

The first is the more useful one to write about. Two portals reporting an AED 1.08 million spread on the same metric type, for the same community, on the same day, is harder to dismiss than a gap you can explain away as asking-versus-sold.

A buyer checking two portals sees that spread and gets no explanation from either. They don't need a third page reproducing a portal's number. They need someone to tell them which number fits the decision they're making.

The off-plan versus ready trap is the best worked example of this.

Across all DLD sales in H1 2026, ready property took AED 146.69 billion (USD 39.9 billion) across 27,160 transactions against off-plan's AED 139.75 billion (USD 38.1 billion) across 58,840. Read that alone and off-plan is most of the deals and under half the money.

Restrict it to residential and it inverts. Off-plan took 60,425 residential transactions, 73.8 percent of volume, worth AED 168.2 billion (USD 45.8 billion), or 74.5 percent of residential sales value. The ready market took 21,436 transactions worth AED 57.5 billion (USD 15.7 billion).

The all-sales figures include whole buildings and land parcels. H1 2026's 86,005 transactions spanned 71,570 units, 7,301 buildings and 7,134 land parcels, and those bulk deals sit in the ready column and drag its average deal size to about AED 5.39 million against AED 2.38 million for off-plan.

An apartment buyer reading the all-sales version gets the wrong picture of their own market. Residential totals also vary by provider: Engel and Volkers reported 79,281 transactions worth AED 221.4 billion, Cavendish Maxwell 79,300, and Reliant Surveyors 81,839, with reported off-plan share ranging from about 71 percent to 73.8 percent accordingly. REIDIN also has residential value down 16 percent year on year with ready transactions down 26 percent, which is context a buyer choosing between the two segments needs and neither portal's community page carries.

Explaining which denominator applies to which buyer is content a portal has no product reason to produce and a broker is well placed to write. The full version of that argument is on the off-plan versus completed page.

Tier 3: evergreen buyer education

Lower ceiling, still worth owning, and the base your Tier 1 pages link from.

Freehold versus leasehold versus usufruct, the 99-year cap outside designated zones, non-resident purchase without UAE residency, the AED 2 million (USD 544,588) Golden Visa threshold. The u.ae government page appearing first on that query caps you at second place, and second there beats tenth on a listing query.
Post-purchase operations: handover snagging, service charge budgets and the Mollak system, tenancy renewals, missed handover dates. Buyers ask these before they buy.

What not to make your primary acquisition strategy

Community listing pages, tower listing pages, and general area guides. Not because they're bad pages. Betterhomes, my example brokerage throughout, runs an area-guides hub, and pages like these still earn internal linking, entity clarity, branded search and conversion support.

The mistake is expecting them to win generic transactional queries against a page that already contains your own listings. Build them if they serve those other purposes. Don't fund them ahead of Tier 1.

One number on the portal side of the ledger, from Property Finder's own white paper of 10 December 2025: between 2022 and 2025, the number of active agents grew 30 percent annually while listings rose 34 percent and engagement per listing fell 36 percent across that window. Per the same white paper, leading agencies now run portfolios 118 percent larger than in 2022 with marketing budgets up nearly 190 percent, and 5 percent of agents generate more than 40 percent of platform leads.

Those are their figures, not mine, and they are the vendor's own platform data about its own platform, published without a methodology and not corroborated anywhere independent. Read them as the operator's account. The budget line is still the sharper argument for owning an audience somewhere the portal doesn't set the rules.

05AI, precisely

What AI changes, and what it doesn't

Two things get collapsed together in most Dubai marketing commentary, and they behave differently.

AI referral traffic is clicks arriving from a chatbot. Today that is negligible here. Similarweb's public AI Traffic Checker recorded 31,400 AI-sourced visits to propertyfinder.ae in July 2026, ChatGPT accounting for 96.54 percent of them, amounting to 0.61 percent of all referral traffic to the site, with ChatGPT-sourced traffic down 23.36 percent month on month. Referral is already a minor channel: Similarweb shows direct traffic driving 41.45 percent of desktop visits. So that 0.61 percent is a fraction of a fraction.

AI visibility is whether you get retrieved, cited, mentioned or recommended inside the answer, whether or not anyone clicks. That's a different surface, it isn't measured by referral analytics, and Bayut's ChatGPT app is a third thing again: a live inventory feed the model can call.

The click numbers are small and may stay small. The retrieval surface is the one worth building for, and it rewards the same things Tier 1 does, which is current figures with dates and sources attached.

06FAQ

Questions

Can a Dubai brokerage outrank Bayut, Property Finder or dubizzle for "apartments for sale in Dubai Marina"?

Realistically, no, and the reason is aggregation rather than size. On 4 September 2026 Property Finder listed 3,554 Dubai Marina apartments, dubizzle 3,141 and Bayut 3,123, against 100 on Betterhomes.

Those 100 Betterhomes units are inside the portals' totals, because Betterhomes lists there. No agency can out-supply a page containing its own supply.

Which of the three portals is biggest?

The tools disagree and they change month to month. Semrush had dubizzle.com first in UAE Real Estate in April 2026 at 2.33 million visits, then bayut.com first in June 2026 at 3.61 million. Similarweb ranks bayut.com first but files dubizzle under Classifieds, so it is not in that comparison.

Both of those category pages overwrite themselves every month and neither is archived, so the states quoted here cannot be re-opened. On listing count in one community on 4 September 2026, Property Finder led at 3,554. Count inventory in your own communities instead of trusting one traffic estimate.

Are the portals already inside AI answers?

Bayut launched an app on ChatGPT on 1 June 2026 and added conversational AI Search on its own portal on 27 August 2026. On measurable clicks the picture is small: Similarweb shows 31,400 AI-sourced visits to propertyfinder.ae in July 2026, 96.54% from ChatGPT, which was 0.61% of that site's referral traffic.

The tool integration is structurally significant. The click numbers today are not.

Does listing on more portals fix visibility?

Property Finder's white paper of 10 December 2025 reported that between 2022 and 2025, active agents grew 30% annually while listings rose 34% and engagement per listing fell 36%. Per the same white paper, leading agencies' marketing budgets rose nearly 190% against 2022, and 5% of agents produce more than 40% of platform leads.

Those are Property Finder's own platform data about its own platform, with no published methodology and no independent corroboration. Read them as the operator's account rather than as a market measurement. Adding listings into that has a predictable effect on attention per listing.

Why do the two big portals report different average prices for the same community?

On 4 September 2026 Bayut reported an average asking price of AED 3,993,035 (USD 1.09 million) for Dubai Marina apartments from 6 months of listings, while Property Finder reported AED 2,908,161 (USD 792,000) from 12 months of listing data. Both are asking prices, so the AED 1.08 million spread comes from window length and listing mix.

Bayut's DLD-sourced sold average of AED 3,285,997 across 2,248 transactions is a third measurement again.

Is off-plan really most of the Dubai market?

Depends which dataset. Across all DLD sales in H1 2026, off-plan was 58,840 transactions worth AED 139.75 billion against ready property's 27,160 worth AED 146.69 billion, so under half the money.

On residential-only data it reverses: 60,425 off-plan deals, 73.8% of volume, worth AED 168.2 billion, or 74.5% of value, against 21,436 ready deals worth AED 57.5 billion. The all-sales figure includes 7,301 buildings and 7,134 land parcels. Residential is the relevant set for an apartment buyer.

Have the portals already written the advisory content?

Yes, and that is the point. Bayut's MyBayut service charges guide covers the DLD index and the AED 3 to 30 range, and its table of contents is still labelled 2024. Property Finder has an equivalent blog guide.

In the searches I ran on 4 September 2026, neither appeared, while 2026-dated brokerage guides did, including Luxhabitat's with Burj Khalifa at roughly AED 67.88 per square foot. The opening is in currency and unit-level specificity.

How long before a brokerage's advisory content shows up in AI answers?

No honest answer includes a date or a guarantee, and I do not give either. What I work from is documented instances in my own frozen-prompt runs, where specific published pages were quoted back by named engines on recorded dates.

Those are logged observations, not a forecast of what any new page will do. Anyone quoting you a timeline or a citation outcome is selling something they cannot control.

Sources

Related pages

All figures verified 4 September 2026. Currency converted at the UAE dirham's fixed peg of AED 3.6725 to USD 1. Page last updated 9 September 2026.

Want to know where your brokerage sits on these queries?

A 30-minute call. We will run your prompts live and you can watch what comes back.

Book a Call Free visibility check