Arabic and localization
Arabic keyword research: a practical method for Gulf markets
Ilyes Massoudi8 min readThe short answer
Arabic keyword research starts from the customer's intent, not from a translated English list. Every intent has at least three written forms, Modern Standard Arabic, Gulf dialect and Latin-letter Arabizi, and their volumes differ widely. The working method: write down the intents that precede a purchase, collect their forms from autocomplete and Search Console, pull search volumes per country from Google Keyword Planner, then map each group to one page that answers it. A global number means nothing in the Gulf.
The mistake everything else grows from
Most Arabic keyword lists I see on Gulf websites were built one way: take the English list and translate it. The result is words that are grammatically correct and typed by nobody. A formal phrase for “landscaping services” is a clean translation, but the searcher in Riyadh types “landscaping company in Riyadh” in dialect, or “home garden landscaping”, and may misspell it in a way that carries more volume than the correct spelling.
Order is the whole issue. Someone who translates and then “optimises” has decided the words before looking at the data. The right method starts from the intent and ends at the word.
Three written forms for every intent
A Gulf user types one intent three different ways, and all three appear in the data:
| Form | Example for the intent “fix my air conditioner” | Where it dominates |
|---|---|---|
| Modern Standard Arabic | صيانة مكيفات (AC maintenance) | Institutional content, older searchers |
| Gulf dialect | تصليح مكيف، فني مكيفات قريب مني (AC repair, AC technician near me) | Local and urgent services |
| Arabizi (Latin letters) | tasleeh mukayef | Mobile, younger users, technical terms |
On top of that, four spelling patterns genuinely change the numbers: the hamza (أفضل versus افضل), the final taa marbuta versus haa (شركة versus شركه), the final yaa versus alef maqsura (مستشفى versus مستشفي), and spaces inside compound names. Search engines understand most of them, but tools report them as separate keywords, so you have to group them before judging the volume.
What differs between Saudi Arabia, the UAE and Qatar
The Gulf is not one search market:
- Saudi Arabia: the largest Arabic demand in the region. Searchers type Arabic first, with local vocabulary: “wesh afdal” for which is best, “abi” for I want, and “in Riyadh / in Jeddah / in Dammam” as a near-fixed location suffix.
- UAE: demand splits between English for residents and Arabic for nationals and Gulf visitors. Dialect leans to “shu ahsan” for which is best and “fi Dubai” for in Dubai, and the Arabic volume for a given term is typically a tenth of the Saudi figure.
- Qatar: small volumes in both languages, and much of the demand is institutional. A term carrying 880 searches a month in Saudi Arabia may show as 10 in Qatar. That does not mean the market is absent. It means the tool is rounding.
The working rule: pull volumes per country. A global figure for an Arabic word lumps Egypt, Morocco and Iraq in with the Gulf, and tells you nothing useful about a page that sells in Doha.
The method in six steps
1. Write intents, not keywords. What does the customer ask before buying from you? “How much”, “best”, “near me”, “is it worth it”, “the difference between”. Ten to twenty intents cover most service businesses.
2. Collect the forms. For each intent write the MSA, dialect and Arabizi versions, then open Google in a private window, type the start of the phrase and record the autocomplete. At the bottom of the page you get related searches, and in the middle, other questions people ask. Three free sources of what people actually type.
3. Pull volumes per country. In Google Keyword Planner, set the country and Arabic, paste the list. Record volume, competition and cost per click. The last one tells you where advertisers pay, which is where money is.
4. Look at the results page yourself. For every important query, look at page one: is it shops, articles or a map? That tells you what type of page Google wants for that query, and an article cannot compete for a query Google wants a product page for.
5. Map each group to one page. Every group of forms for one intent gets exactly one page that answers it. Two pages for the same intent compete against each other; one page for ten intents answers none of them.
6. Rank by opportunity. Volume matters, but so does whether you can compete, what advertisers pay, and how close the intent is to a purchase. A query with 50 searches close to buying is worth more than one with 5,000 typed by students.
If your site already exists, start from Search Console
The best Arabic keyword data does not come from a paid tool. It comes from the Performance report in Search Console: the queries your site actually appeared for, in the real forms people typed, with impressions and clicks per country.
Look for one kind of query: high impressions, few clicks, average position between 5 and 20. Google already considers you roughly relevant for those. One page written specifically for such a query can move it to page one with the least effort available anywhere.
Four mistakes I see constantly
- Chasing the generic head term. The bare word for SEO carries large volume, and most people typing it are students or curious. “SEO company in Riyadh” is ten times smaller, and the person typing it wants to pay.
- Ignoring common misspellings. The “wrong” spelling can be the larger one. Do not use it in the title, but make sure Google can tell your page answers it, which usually means both forms appear in the text.
- An Arabic page for every English keyword. The Arabic market may need pages the English version does not have, and vice versa. Structure follows demand, not translation.
- Trusting one number. No volume without a country, and no country without a look at the results page itself.
What the audit does with this
In the Digital Presence Audit I apply exactly these steps to your site: I measure what your customers type in Arabic, English and Arabizi, with volumes for each country you target, compare that against the pages you own, and hand you the list of demand no page of yours serves, ordered by opportunity size against competitive difficulty. It is usually the cheapest opportunity in the whole report, because closing it takes a well-written page rather than an advertising budget.