Arabic SEO: the keyword research most agencies skip

Most Arabic SEO work in the UAE is an English keyword list run through Google Translate and handed to a writer. It produces grammatically correct Arabic that almost nobody types into a search box, because it skips the three things that actually determine what UAE Arabic searchers type: how intent gets phrased rather than just worded, the large share of queries typed in Latin script instead of Arabic script, and the gap between formal written Arabic and the Gulf dialect people actually speak and search in.

Key Takeaways

• Translating an English keyword word-for-word produces phrases that are grammatically valid but rarely match how a native speaker would ask the same question.
• A meaningful slice of Arabic-market search happens in Arabizi (Arabic written in Latin letters and numerals) which pure Arabic-script keyword tools do not index at all.
• Gulf dialect and Modern Standard Arabic diverge in vocabulary, word order and phrasing; most site content is written in MSA while most everyday search happens closer to dialect.
• Keyword tools set to an English or generic Arabic locale routinely miss UAE-specific phrasing that a Google autocomplete check in Arabic, run from inside the UAE, surfaces immediately.
• The fix is procedural, not a better tool: native speaker review of every shortlisted term, manual autocomplete and People Also Ask checks in Arabic, and Arabizi variants tested alongside the Arabic-script ones.

Translation gets you correct Arabic, not searched Arabic

A translated keyword list is optimized for grammatical correctness, and grammatical correctness is not what a search engine matches against. The gap shows up constantly in business-services queries: an English phrase like "company formation cost UAE" translated literally into Arabic produces a stiff, formal noun phrase that reads like a government circular. A UAE Arabic speaker actually looking for that information is more likely to type something closer to a plain question (how much does it cost to open a company, what do I need to start a business) using everyday verbs and question words rather than the noun-heavy construction English keyword research trains you to expect.

This is a search-intent problem wearing a translation costume. English keyword research already teaches you that "best CRM software" and "how to choose a CRM" serve different intents even though they share a topic. That same discipline has to be reapplied in Arabic from scratch, because a literal translation preserves the topic and destroys the intent signal. It carries over the English phrasing pattern, not the Arabic one a search engine's query-matching actually rewards.

Arabizi is a real slice of demand your keyword tool cannot see

Arabizi: sometimes called Franco-Arabic: is Arabic written in Latin letters, with numerals standing in for sounds Latin script does not have (3 for ع, 7 for ح, and so on), a system that emerged with early mobile phones and internet chat when devices didn't support Arabic script well (Wikipedia, "Arabizi", retrieved 2026-09-11). Every Arabic-qualified lead missed because of a translated-not-searched keyword list still has a cost of acquisition attached to it, and running that math through the CAC/LTV calculator makes the size of the gap concrete rather than theoretical. It is common across the Gulf's bilingual population, where expatriates make up roughly 85% of UAE residents alongside a large English-speaking presence (Siemund, "Multilingualism and the role of English in the United Arab Emirates", World Englishes, 2021, retrieved 2026-09-11), particularly among younger users and anyone typing quickly on an English-layout keyboard rather than switching input methods mid-search. A search for a service or product can just as easily arrive as Arabizi as it does in Arabic script or English.

The practical problem is that standard keyword research tools, whether set to an Arabic locale or not, are built to index Arabic-script text. A term typed in Latin letters and numerals does not match an Arabic-script keyword database at all, and it will not show up as a variant of the English term either, because it is neither English nor standard Arabic. It is its own written form. An agency running keyword research only in English and formal Arabic script has a structural blind spot here, not a minor gap: an entire register of real queries that no standard tool surfaces, because no standard tool is built to look for it.

Gulf dialect and Modern Standard Arabic are not interchangeable for search

Modern Standard Arabic is what UAE news sites, government portals and most translated corporate content use. It is the correct register for formal writing, and abandoning it for on-page content would be a mistake in its own right. But it is not, on its own, what most people search in. Gulf Arabic, the dialect spoken across the UAE, Saudi Arabia, Qatar, Bahrain, Kuwait and Oman, differs from MSA in vocabulary, some grammar, and especially in the everyday phrasing people default to when they are typing a quick question rather than composing formal prose, the classic diglossia pattern where a formal written register and an everyday spoken one coexist without being interchangeable (Wikipedia, "Gulf Arabic", retrieved 2026-09-11). It also carries loanwords, from Persian and Hindi among others, that MSA content simply does not use.

The practical consequence for keyword research is that a keyword list built entirely from MSA vocabulary will match how your service page reads and miss how your buyer actually searches. This is not an argument for writing site content in dialect: MSA remains the right register for the page itself, the same way formal English is the right register for an English service page even though nobody searches in fully formal English either. It is an argument for building the keyword list from real dialect-inflected search behavior, then writing the page in a register that answers it clearly, rather than assuming vocabulary parity between the two automatically exists.

What proper Arabic keyword research actually involves

None of this requires an exotic toolset. It requires treating Arabic keyword research as its own research task rather than a translation step tacked onto the English one. A workable process looks like: start from the same intent map used for English (informational, comparison, transactional) then build the Arabic keyword list against that intent map independently, rather than translating the English terms one by one. Check Google autocomplete and People Also Ask with the interface language and location set to Arabic and the UAE, not just the keyword input; the suggestions that surface there are drawn from real query behavior and routinely differ from anything a translation tool would generate. Test the same core queries in Arabizi as well as Arabic script, since both can carry meaningful volume and neither substitutes for the other in a keyword tool's index. And treat every shortlisted term as a draft until a native Gulf Arabic speaker (ideally someone who fits your actual buyer profile, not just any Arabic speaker) confirms it reads as something they would naturally type, not something that reads as translated.

This sits inside the same acquisition-planning logic covered in the UAE growth and go-to-market guide: Arabic content is comparatively uncontested space in most UAE B2B categories, Arabic accounts for only around 0.6% of the content language of websites tracked worldwide despite the size of its speaker base (W3Techs, usage statistics for Arabic content language, retrieved 2026-09-11), and that advantage only converts into traffic if the keyword research underneath it reflects how people actually search rather than how a translation tool renders English into Arabic.

Checklist: building an Arabic keyword list that reflects real search behavior

1. Map intent in English first, then rebuild the Arabic list against that same intent map. Don't translate term by term.
2. Set autocomplete and People Also Ask checks to Arabic language and UAE location, and record what actually surfaces rather than what you expected to see.
3. Test Arabizi variants of your top terms alongside Arabic-script versions; treat a zero-result Arabizi check as inconclusive, not as proof the term has no volume.
4. Route every shortlisted keyword through a native Gulf Arabic speaker before it goes into a content brief. Ask specifically whether it reads as searched or as translated.
5. Keep on-page copy in Modern Standard Arabic even when the keyword research is dialect-informed; the register mismatch between formal content and colloquial search terms is normal, not a defect to fix by writing dialect copy.
6. Re-check autocomplete quarterly. Search phrasing shifts as UAE's bilingual population skews, and a keyword list validated once does not stay valid indefinitely.

Frequently asked questions

Do we still need Arabizi keywords if our target audience reads Arabic script comfortably?

Yes, because Arabizi use is a typing-speed and habit pattern, not a literacy gap. Fluent Arabic readers still default to Latin-script typing on English-layout keyboards for quick searches. Skipping Arabizi assumes it only exists where Arabic literacy is weak, which is not how the behavior actually works across the UAE's bilingual population.

Should our website copy be written in Gulf dialect since that's how people search?

No. Keep on-page content in Modern Standard Arabic. It reads as professional and is the expected register for a business website, the same way formal English suits a service page even though searchers rarely type in fully formal English. Dialect belongs in the keyword research process, not necessarily in the published copy itself.

How do we check Arabic autocomplete without an Arabic keyboard or fluent Arabic staff?

Set your browser and Google's interface language to Arabic and your location to the UAE, then type queries using an on-screen Arabic keyboard or a transliteration input method, most operating systems include one. Screenshot the results and send them to a native speaker for a quick read, rather than attempting to judge the phrasing yourself.

The bottom line

Arabic keyword research fails silently, not loudly: the translated list looks complete, the page publishes, and it simply never matches what anyone actually types. The fix is not a better keyword tool; it is treating Arabic search behavior as a distinct research task with its own intent mapping, its own check for Arabizi and dialect variants, and a native speaker in the loop before anything goes into a content brief. Agencies that skip this usually aren't being careless: translation is the visible, billable step, while checking what people actually search is the invisible one that determines whether the content ever gets found. If Arabic search is going to be part of a paid or organic UAE acquisition push, fold the keyword work into your broader advertising and acquisition planning so budget and content decisions move together instead of being sequenced as separate projects.

Figures were verified on 11 September 2026 against Wikipedia's linguistics references, W3Techs' website-language usage statistics, and published research on multilingualism in the UAE. Language-behaviour figures shift gradually; re-check before citing them again in a year or more.