Arabic or English Ads in MENA? The Wrong Question, and What to Ask Instead

The short answer: stop choosing. In the Gulf the same person searches in both languages and switches based on what they are trying to do, so the winning structure is not "Arabic or English" but separate campaigns per language, each with creative and a landing page in that language. The expensive mistake is not picking the wrong language. It is running one language and paying for traffic that lands on a page it cannot read.

The question every MENA account gets wrong

Sit in enough campaign reviews in this region and you hear the same debate. Should we run Arabic or English? Someone cites the official language. Someone else cites the expat share. A decision gets made, usually on instinct, and the account runs that way for two years.

Both sides are answering a question the buyer does not ask.

Take the UAE. Over two hundred nationalities, and functionally two languages that matter commercially: Arabic is official, English runs business. That is not a split you resolve by picking a side. It is a split that lives inside individual users, most of whom operate comfortably in both and choose between them situationally.

The same person will search أفضل قهوة في دبي on a Friday morning and "best coffee in Dubai" on a Tuesday at work. Not because they changed. Because the context changed.

What actually differs between the two

They are not interchangeable, and treating them as a translation problem is where budget dies.

Arabic pulls on trust, culture and mobile behaviour. Arabic-language creative reaches Emiratis and Arab expats in the register they use socially rather than professionally. It reads as addressed to them rather than at them. In our own campaigns Arabic creative has consistently outperformed the English equivalent on mobile click-through in consumer categories, sometimes substantially. Treat that as directional, not as a benchmark to hold your account to: the gap varies enormously by category, and in some verticals it disappears entirely.

English pulls on reach and B2B intent. For professional services, SaaS, commercial real estate and anything with a considered purchase and a procurement process, English carries the query volume and the decision-maker. It also carries the expat professional segment almost entirely.

The pattern underneath: Arabic tends to win where the decision is personal and emotional, English where the decision is professional and evaluative. Most categories contain both moments, at different points in the same funnel.

Localise, do not translate

This is the single most common failure, and it is visible in the ad copy from across the room.

Run "Best quality, best price!" through a translator and you get أفضل جودة، أفضل سعر! It is grammatically correct. It is also flat, generic, and reads exactly like what it is: an English ad wearing Arabic. It competes on the one axis where you have no advantage.

The localised version does not translate the sentence. It replaces the appeal. Something closer to أسلوبك يعبر عنك, meaning "your style speaks for you", with supporting lines about quality that lasts, design that reflects you, fast and easy delivery. Same product, different persuasion, built for how the audience actually evaluates.

The test is simple. Show the Arabic to a native speaker in your target market and ask them, without context, whether it was written in Arabic or translated into it. They will know within one line. So will your CTR.

This is also where machine translation quietly costs you money. It will produce something correct and lifeless, and correct-and-lifeless loses to a competitor who wrote natively.

The bilingual user is not an edge case

Here is the structural insight most accounts miss. You are not choosing between two audiences. You are frequently choosing between two halves of one audience, and often two halves of one person.

Run the same commercial query in both languages and compare the results pages. Different businesses rank. Different ads serve. Different competitors appear. If you are only present in one, you are absent from a results page your buyer will see, on a query they will type, in a market you claim to serve.

Dual-language presence is not double the work for the same result. It is coverage of a market you are currently seeing half of.

The tactical setup

This part is mechanical, and getting it wrong is why bilingual accounts underperform even when someone did commission the translation.

Separate campaigns by language. Not separate ad groups inside one campaign. Separate campaigns. You need independent budgets, independent bidding, and independent performance data, because the two languages will behave differently enough that blending them hides both.

Match the landing page to the ad language. This is the one that silently destroys accounts. An Arabic ad pointing at an English landing page is a paid click you have already lost. The user arrives, the page does not speak to them, they leave. You paid for the click, you got the bounce, and your conversion rate quietly tells you Arabic "does not work" when what actually failed was the handoff.

If you have Arabic ads and no Arabic landing page, pause the Arabic ads until you do. That is cheaper than learning the same lesson across a quarter of spend.

Mirror the funnel, not just the headline. The form, the error messages, the confirmation, the follow-up email. A journey that switches language halfway through reads as unfinished, and in a trust-sensitive market that costs conversions.

Set the language and location targeting deliberately. Language targeting in Google Ads keys off the user's interface and content settings, not the language of your keywords. An Arabic keyword in a campaign targeting English speakers will serve to people who will not convert. Check this setting on every bilingual account you inherit; it is wrong more often than not.

Where this lowers cost

The reason to do this is not tidiness. It is that both languages are usually cheaper than the one everybody defaults to.

English commercial keywords in the Gulf are contested by every regional and international advertiser in the category. Arabic equivalents for the same intent are frequently less contested, which shows up directly in cost per click and cost per acquisition. You are bidding into a thinner auction for a buyer with the same wallet.

The same asymmetry runs through organic and AI search. Arabic answers are being assembled from a much smaller pool of well-structured Arabic content than their English equivalents, which is exactly the gap we cover in how to get cited by AI. Less competition for the same intent, in a language most of your competitors have not bothered to do properly.

That window is open now and it will not stay open.

The audit

Run this against any MENA account, in order.

1. Pull last quarter's search terms and sort by script. If one language is under 10% of impressions in a market where both are spoken, that is a coverage gap, not a preference.
2. Check the language targeting setting on every campaign. Confirm it matches the language of the keywords and creative.
3. Click every ad in a bilingual campaign and note the landing page language. Any mismatch, fix or pause today.
4. Show your Arabic creative to a native speaker in the target market. Ask whether it was written or translated. Act on the answer.
5. Split any campaign currently mixing languages into two, and give each its own budget so the data separates.
6. Check that the whole funnel holds the language, not just the first page.
7. Compare CPC and CPA between the two languages for the same intent. If Arabic is materially cheaper, that is unclaimed inventory.

The honest summary

The language question is really a matching question. Ad language, landing page language, funnel language, and audience intent all need to agree, and the campaign structure has to keep them separable enough to see what is working.

Do that and both languages get cheaper, because you stop paying for clicks that were never going to convert.

Pick one language and you have not simplified your account. You have halved your market and called it a strategy.

Sources

The judgements about register, localisation and bilingual behaviour come from running MENA accounts, not from published research; treat them as testable claims rather than findings. The platform mechanics are documented and linked below.

• Google Ads, about language targeting. Language targeting keys off the user's settings, not your keyword language.
• Google Ads, about location targeting
• Google Ads, about the search terms report. For the script-split audit described in the article.