Meta Ads that find buyers in the UAE.

Usually both, and as two separate creative tracks rather than one asset translated. Arabic tends to carry hospitality, food, retail and government-adjacent audiences here, which is why an Al Seef restaurant group on a DET mainland licence leads in Arabic and tests English second, while English carries expatriate property, ecommerce and much of B2B, so the two rarely want the same hook. We write the angles separately, test them separately and read the result per track, because a translated caption over a shared image tells you nothing about which audience responded. Where the budget will not fund Arabic creative and a right-to-left destination, we scope English only and write that down.

Technically yes, and it will not be worth much. Meta accepts a source audience far smaller than the one it recommends modelling from, and a few hundred names sits in that gap, so what comes out is closer to a broad audience with extra steps. We build the first reach from interest, behaviour and employment signals instead, and treat growing the file to a size worth modelling from as part of the work. Once it gets there the lookalike is the strongest targeting on the platform, and we move to it.

Yes, and it is a size problem rather than a creative one. The reachable pool in the UAE is small enough that a winning ad meets most of its audience quickly, so frequency climbs, response falls, and cost per result rises while the budget has not moved. A Dubai Production City ecommerce brand on a free zone licence sees it inside a fortnight. Raising the budget makes it worse, because it buys the same people again sooner. The fix is a queue: the next variant is built and tested before the current one tires, at a refresh cadence agreed up front rather than negotiated after the decay.

It is priced against the work the account needs, and three things move the number: how many creative tracks run, how fast they have to be replaced, and whether the audience is being built from scratch or inherited. A single English track for one offer is a different engagement from Arabic and English across a catalogue. On VAT, the media and the fee are separate: Meta charges 5% VAT on ad spend where the sold-to country is the UAE and no Tax Registration Number sits on the ad account, and with a valid TRN on the account that charge moves to the reverse-charge mechanism on your own return. Send the current spend and we will tell you what running it properly takes.

With exclusions written before the first creative, and a hook that names who it is for. Location, language, behaviour and employment exclusions are set against what the offer can actually serve, and reviewed whenever the offer changes rather than once at launch. A Yas Island leisure operator on a DCT Abu Dhabi tourism licence is paying for Dubai impressions until somebody writes that exclusion down. The second half is the creative itself: an ad that opens by naming its buyer disqualifies the rest of the feed faster than any setting does. We report wasted reach as its own line in the weekly review rather than leaving it buried inside cost per result.

Yes, and that is more often an argument for it than against it: with nothing to capture, Meta is where demand gets created rather than harvested. It is the wrong first channel in two cases. If the offer only makes sense to someone who already understands the category, the creative has to teach before it sells, and that budget usually does more on the proposition and the destination first. And if your buyer is a narrow, named few, the reach Meta is good at is the wrong shape for them. We say which of the three you are in before we quote. Where the answer is that the audience has to be built rather than bought, social media marketing is the slower half of the same job.

Yes, all of it, and the ownership is set on day one rather than at the end. The Business Manager, the ad account, the pixel and Conversions API configuration, the product catalogue and every custom audience are created in your name, with Webzenia added through partner access, so removing us is a permissions change and not a migration. The part worth checking with any agency is the custom audiences and the event history, because those are what the account has learned. An account rebuilt from zero pays to learn it a second time.