The UAE is one of the few advertising markets where two languages routinely sit inside one campaign, and where getting the second one slightly wrong is worse than not running it at all. A clumsy Arabic headline does not read as a foreign company trying its best — it reads as a scam. This is about what AI tools genuinely handle, what they quietly get wrong, and how to structure a bilingual campaign so it works.
The failure that shows up on the image, not in the copy
Arabic is written right to left, and the characters join. Latin text does neither. When an AI image tool draws text onto a generated scene, it is usually doing so with a rendering pipeline built and tested on Latin script. Three things go wrong, in order of how often we see them.
Characters unjoined
Arabic letters change shape depending on their position in a word. A renderer that places glyphs individually produces text that is technically the right letters in the right order and completely wrong to read — the equivalent of English printed one letter per box.
Word order reversed
The renderer applies left-to-right layout to right-to-left text. The sentence comes out backwards. Native readers spot this instantly; nobody else does, which is why it ships.
Font substitution
The chosen display font has no Arabic glyphs, so the system silently falls back to whatever it has. The headline renders in a default face that looks nothing like the brand and often clashes with the layout.
💡 Test this before you trust any tool with Arabic creative: generate one image with an Arabic headline, and have someone who reads Arabic look at it. Not a translator checking the words — a reader checking whether it looks like something a real company published. The two are different tests and only the second one catches rendering faults.
Translation is not localisation
Even when the rendering is correct, a translated headline usually underperforms a written one. English ad copy leans on compression and wordplay that does not survive the trip. "Book now, pay later" translates cleanly. "Stop guessing. Start selling." does not — it becomes two flat imperatives with none of the rhythm.
The practical approach is to give the AI the offer and the audience rather than the English sentence, and have it write in Arabic from that. You get copy built for the language instead of copy dragged into it. Any tool that only offers a "translate" button is doing the weaker of the two jobs.
Should you split the campaign by language?
Usually yes, and for a reason that is about measurement more than delivery.
- 1Run both languages in one ad set and Meta will optimise toward whichever creative gets early traction. That is often the English one, simply because it tends to reach a broader slice of the UAE's mixed population first. The Arabic creative then starves before it has been fairly tested.
- 2Split into two ad sets and each language gets its own budget and its own learning. You find out what Arabic actually costs you per result instead of assuming.
- 3The cost of splitting is Meta's minimum daily budget, which applies per ad set. On an AED account a conversion objective needs roughly AED 11.07 a day per ad set — so two languages means budgeting above AED 22 a day before Meta accepts the campaign.
If your daily budget cannot carry two ad sets, do not split it into two underfunded ones. Pick the language your customer is most likely to buy in, run it properly, and add the second when the budget supports it.
How Meta language targeting actually behaves
This is misunderstood often enough to be worth stating plainly. Meta's language targeting filters on the language a person has set on Facebook or Instagram, not the languages they speak. In the UAE, a large number of Arabic speakers use their phone in English. Target "Arabic" as a language and you will exclude a substantial part of your actual Arabic-speaking audience.
💡 For most UAE campaigns the better structure is to leave language targeting off and let the creative do the filtering. An Arabic ad is self-selecting — people who do not read it scroll past. You reach everyone, and you pay only for the ones the creative speaks to.
What about the landing page?
A common and expensive mismatch: Arabic ad, English landing page. The click is bought and then thrown away. If you cannot produce an Arabic page, this is a real argument for running Arabic ads into WhatsApp rather than a website — the conversation continues in the language the ad started in, with a human on your side of it.
This is why click-to-chat is so much more common in the Gulf than in Western markets, and why a tool that treats WhatsApp as an afterthought is a poor fit here regardless of how good its creative is.
A workable structure
Two ad sets, one campaign
One Arabic, one English, each with its own budget above the AED minimum. Same objective, same geography, different creative and copy.
No language targeting
Let the creative filter. You reach English-interface Arabic speakers you would otherwise exclude.
Matched destination
Arabic ad to an Arabic page, or to WhatsApp. Never Arabic ad to English page.
Judge separately
Compare cost per result per language, not per campaign. They will differ, and the difference tells you where to put the next dirham.
Copy written in the language, not run through a translator
Adyft writes ad copy in the language of the market rather than translating it, and builds Click-to-WhatsApp campaigns with UAE dial codes handled properly. Fourteen-day free trial.
Start free — no card needed