The honest answer is both, split by platform and OS, and tested like any other variable. In the GCC, iOS audiences skew more English, broad campaigns should carry both languages, and on the account I run the top text asset was Arabic.
The mistake is not picking the wrong language. It is assuming a tilt you never measured.
I run paid user acquisition for a GCC fintech app across Meta, Google App campaigns and TikTok, in both Arabic and English, and the language question comes up in almost every conversation I have with GCC growth teams. Here is what the data on a live account shows, what general operator experience adds, and where I would be honest that the evidence is thin.
Should GCC app ads be in Arabic or English?
Neither by default. Language is not a creative preference. It is a targeting decision with a measurable cost, and it should be treated exactly like an audience or a bid: something you test, per surface, against your real KPI.
The GCC makes this genuinely two-sided. These are markets where a large share of residents are more comfortable in English, where nationals often move between Arabic and English within a single sentence, and where the mix shifts by country, by city, by age and by device. Any blanket rule, "always Arabic because it is the Gulf" or "always English because everyone speaks it", throws away performance on whichever half of the audience it ignores.
So the useful question is not which language wins. It is: on this platform, this OS, this audience, which language wins, and by how much?
How do platform and OS change the language answer?
Materially. The clearest pattern on the account I run is that iOS audiences skew more English than Android in the same market, which matches what I have seen across the GCC generally: the iOS base over-indexes on English-comfortable, higher-income users. Android campaigns on this account have historically been Arabic-first, and broad conversion campaigns take both languages.
The rule I hold hardest came from getting it wrong: if a campaign has no language filter set, deploy both languages and let delivery vote. Do not assume a tilt on the audience's behalf. The algorithm allocating spend between an Arabic and an English version of the same ad is a cheaper and more truthful survey than any internal debate.
| Surface | Typical language lean | What I deploy |
|---|---|---|
| iOS campaigns | Skews more English | Lead English, keep Arabic in the mix |
| Android install campaigns | Often Arabic-first | Lead Arabic, keep testing English |
| Broad conversion campaigns | Genuinely mixed | Both languages, always |
| Any campaign with no language filter | Unknown until measured | Both, and let delivery allocate |
Treat the table as starting priors, not laws. Your category and market will bend every row.
What did language testing show on a GCC fintech app?
Two findings stand out from the GCC fintech app I run. First, text carried more weight than visuals: in Google App campaigns, the hero text assets ran roughly 4x better than the visual assets on cost per in-app action. Second, the single best description asset was Arabic. On an account where plenty of polished English creative was live, the top-performing description line was the Arabic one.
That second finding matters because it cuts against the lazy assumption that English is the "premium" performance language in the Gulf. When the message met the reader in the right language, it beat everything else in the account.
The message itself mattered too. Salary-advance style framing, speaking directly to end-of-month cash pressure, resonated consistently in this market. Language amplified a message that already worked; it did not rescue ones that did not. One account, one category, one market, so treat it as evidence rather than proof. But it is real spend, not opinion.
Which Arabic should you write: Gulf dialect or MSA?
This part is operator experience across GCC accounts, not data from the account above, so weigh it accordingly. My working split: dialect earns attention, MSA earns trust. Gulf dialect belongs where the ad has a human face and a social feed context: talent-led video, UGC-style hooks, conversational captions. Modern Standard Arabic belongs where credibility is the job: product claims, financial terms, anything a regulator or a sceptical customer will read twice.
For a fintech app, that usually means dialect in the hook and the voice, MSA in the promise and the fine print. And within the Gulf, dialect is not one thing: each market hears its own, and every audience notices when the voice is not theirs. If you cannot produce dialect properly for each market, well-written MSA beats badly faked dialect every time.
Why does translated creative underperform?
Because translation preserves the words and loses the ad. Re-create; do not translate. The failures are structural, not cosmetic:
- Hooks do not survive. A hook works through rhythm and idiom. Rendered literally in the other language, it reads like paperwork.
- The layout breaks. Arabic runs right to left and often longer on screen, so text overlays, safe zones and visual hierarchy all need rebuilding, not mirroring.
- Voice casting shows. Subtitling an English speaker for an Arabic audience signals, in the first second, that this ad was not made for you.
- Cultural reference points differ. Salary cycles, family framing, what financial confidence looks like: the GCC versions of these are specific, and borrowed ones ring false.
The practical standard I hold creative teams to: an Arabic ad should be written by someone thinking in Arabic, from the same brief, not derived from the English master.
How do you test language as a variable, not a default?
- Start from one proven concept. Testing language on an unproven message confounds two variables; you will not know which one failed.
- Build native versions, not translations. Same brief, same offer, each language written from scratch by a native thinker.
- Deploy both wherever no language filter exists. Let the auction reveal the tilt instead of assuming it.
- Judge per surface, at the real KPI. A language can win on iOS and lose on Android in the same week. Judge on cost per registration or per funded action, never on clicks; the cheap-click language is not always the one that converts.
- Give each version room to learn. An ad needs roughly 50 conversions a week to exit learning cleanly, which is a platform norm. Starve a version of budget and you have not tested it; you have staged it.
- Roll the lesson forward. When a language wins on a surface, that becomes the lead language for the next bundle there, and it stays under review, because audience mix drifts.
FAQ
Do English ads work in Saudi Arabia?
Yes, for meaningful segments, especially iOS users, younger urban audiences and expat-heavy cities, but Arabic is the default language of the market and usually the volume driver. Run both, judge per surface, and expect Arabic to matter more in KSA than almost anywhere else in the GCC.
Should I run separate campaigns per language?
Only when you need different budgets, bids or audiences per language. Otherwise, both languages inside one campaign is the cleaner test: shared budget, same audience, and delivery tells you the tilt. Split them out once a language has proven it deserves its own budget line.
Is Arabic inventory cheaper than English in the GCC?
Often there is less advertiser competition for Arabic-first creative, since many international teams default to English. But cheaper impressions are not the point. Judge on cost per real outcome; on the account I run, the best-performing description asset was Arabic on merit, not because its media was cheap.
Can I just subtitle my English videos in Arabic?
As a stopgap, it is better than nothing. As a strategy, it underperforms, because the viewer can tell within a second that the ad was not made for them. Native Arabic hooks and voices consistently beat subtitled English in my experience. Re-create the ad rather than captioning it.
Does language matter more than the creative concept?
No. Concept first, language second. On the GCC fintech app I run, salary-advance style messaging worked in both languages because the underlying tension was real, while weak concepts failed in both. Get the message right, then let language testing tell you how to deliver it per surface.
Not sure whether your language mix is a strategy or a habit? I run a free App Growth Review: a short, no-pitch look at your live campaigns, your creative split, and where the budget is leaking.
Related reading: App creative testing that beats fatigue, Why GCC apps waste paid media budget and SKAdNetwork, explained for GCC app marketers.