In September 2026 we scanned the website of every Dubai brokerage that registered with the Land Department between one and four years ago and declared a website on its licence. A hundred and forty-three of those sites responded. A hundred and eleven served no Arabic at all.

Not partial Arabic. Not machine-translated Arabic. None — no Arabic text, no language toggle, no Arabic URL, nothing telling a search engine an Arabic version exists.

Why these firms specifically

Dubai's brokerage register holds around 9,500 licensed firms. Most are very small and never declare a website at all, which tells you nothing useful. So we took the firms in the middle: old enough to have survived their first year, young enough to still be growing, and confident enough to put a website on their licence.

That last detail matters more than it sounds. Only 1% of firms under a year old declare a website. By twelve years, it is 43%. A young firm that bothers to record one is a firm that takes its own presence seriously — which makes what we found harder to dismiss as inexperience.

What we found

Of 173 firms in the cohort, 26 had a declared website that was down, redirected to a dead domain, or returned a 404. Of the 143 that loaded:

• 111 had no Arabic in any form
• 58 had no meta description, so search engines write their search listing for them
• 45 had no H1 heading at all on the homepage
• 9 still said "coming soon" — several of them more than three years after registering
• 3 still had Lorem ipsum placeholder text live on the homepage

We are not naming firms. Every one of these is a fixable oversight rather than a scandal, and the point is the pattern, not the embarrassment.

What the Arabic gap actually costs

The usual objection is that Dubai property is sold in English to overseas buyers, so Arabic is decoration. That is true of some of the market and badly wrong about the rest of it.

Emirati and wider Gulf buyers are among the highest-value clients in the market, and they are the segment most likely to read your site in Arabic and most likely to notice you did not offer it. Institutional and government-linked counterparties correspond in Arabic as a matter of course. And a firm with no Arabic quietly excludes itself from every introduction that runs through those networks.

There is a second cost that is easier to measure. When 78% of your competitors publish nothing in Arabic, Arabic property search in Dubai is close to uncontested. The firms in this study are competing with each other for the same crowded English terms while an entire language sits open.

Why it keeps happening

Almost none of this is a decision. Most of these sites are built on off-the-shelf property themes that ship English-only, and adding a second language is not a setting — it is a rebuild of the layout. So it becomes a later problem, and later never arrives.

That is the trap. Arabic is not a translation layer applied to a finished English site — right-to-left changes navigation, alignment, iconography, form design and typography. Retrofitting it costs several times what designing for it costs, which we have written about in more detail here. The firms in this study are not lazy. They are paying the retrofit premium in advance, without knowing it.

If you run one of these firms

Three things, in order of how cheap they are:

1. Check your own page title and meta description today. If the title is "Home" or your theme's default, that is your headline in every search result and every shared link. It is a two-minute fix and a third of this cohort has not made it.
2. Decide whether Arabic is in or out, explicitly. "Later" is the expensive answer. If it is out, say so and stop paying for the option. If it is in, it belongs in the design brief, not the launch checklist — our real estate website design page covers how we approach that.
3. Look at your site on a phone, in public, cold. Most of the failures in this study are invisible from a desk and obvious in thirty seconds on a handset.

How we checked, and what we might have missed

We requested each homepage once, as a browser would, and looked for any Arabic characters, an Arabic language toggle, an hreflang tag pointing at Arabic, or a link to an Arabic URL. Any one of those counted as having Arabic. A site containing a single stray Arabic character was counted as having it, so the real figure is likely higher than 111, not lower.

Two caveats worth stating. A site that renders entirely in the browser could hide an Arabic option from a plain request; we opened a sample of the most JavaScript-heavy ones by hand and found none, but the check is not exhaustive. And a handful of sites returned bot-protection responses rather than pages, so they are excluded rather than counted as failures.

The register itself is public. Anyone can repeat this.