Who do search & AI cite in Arabic?
An analysis of 34 real Arabic queries and 168 sources — to see who actually feeds the answers, and where the citation vacuum is.
The summary
We examined 34 real Arabic SEO and business queries. An AI-style answer or snippet appeared in 67.6% of them — but the source is rarely a documented Arabic expert. Sales-oriented agencies make up 41% of sources, freelance marketplaces show in half the queries, while named-expert blogs are just 19%, and the top Arabic expert appeared only twice. The result: the seat for a trusted Arabic source is empty — and that's an opportunity.
67.6%
of queries show an AI-style answer/snippet
19%
only, sourced from a named Arabic expert blog
41%
of queries have no named expert at the top
140
distinct domains for 34 queries — high fragmentation
2×
the ceiling for the top Arabic expert in the whole study
168
top sources, manually classified
Where do the answers come from? — sources by type
Each type's share of the 168 sources appearing at the top of the 34 queries.
Most frequent sources
No single domain 'owns' a topic — the top frequency is just 5 out of 34 queries.
- 5×Khamsatfreelance marketplace
- 5×Mostaqlfreelance marketplace
- 4×webskeetagency
- 3×Teryaq Mediaagency
- 3×developers.googletranslated docs
- 2×learnwebseonamed expert
- 2×samarsaqrnamed expert
- 2×ahmadkhatabnamed expert
Live capture from AI (Perplexity)
Live sample (July 2026): we asked Perplexity five questions and recorded the sources it actually cited inside its answer — not just what ranks in search. The result confirms the citation vacuum clearly.
How do I rank in Google results?
citeddevelopers.google · google · youtube
Translated Google docs + YouTube — not a single Arabic expert.
How do I appear in AI answers?
citedseoorgeo · mostafasarhan · waystudio · islamway
Scattered experts + an irrelevant source (islamway) = entity confusion.
Salla store SEO
citedarabicfollower (+2)
A generic content aggregator — no documented Salla reference despite huge Gulf demand.
Best SEO practices 2026
citedseoorgeo · scribd · techsy · seoarabia
One blog recurs alongside low-authority sources (scribd).
Who is the best Arabic SEO expert?
citedemirates24 (+2)
The AI itself said “there is no single agreed name” — and pulled names from a news listicle, not the experts' own sites. The seat is literally open.
Four key findings
The native Arabic expert is rare and scattered
Named-expert blogs are only 19% of sources, the top expert appeared twice, and 41% of queries have no expert at the top at all. Expertise exists but never compounds into authority.
Agencies and marketplaces fill the void
Agencies alone are 41% of sources and appear in 88% of queries; freelance marketplaces in 53%. When real 'authority' shows up, it's usually translated foreign docs, not an Arabic expert.
The citation vacuum is real
Despite answers in two-thirds of queries, the source is a fragmented pool of 140 domains (only 21 recurring) with repetitive definitional content. No single Arabic reference is cited consistently.
GEO and the Gulf market are weakest
'Appear in AI' queries are the most ambiguous with the weakest competition, and Saudi commercial results are sales pages all claiming to be 'the best' with no neutral reference or numbers. No documented Gulf voice fills the gap.
What does this mean for you?
The answer exists in two-thirds of queries, but whoever feeds it isn't a trusted expert — it's a mix of sales pages, freelance-marketplace listings, and copied definitions. When search or AI assembles its answer, it looks for a clear, coherent source backed by experience, numbers, and case studies — and that source doesn't exist in Arabic today. Any expert who documents real results with numbers, covers the neglected angles (ranking-drop diagnosis, a neutral pricing benchmark, a native Arabic GEO guide, Salla/Zid SEO with screenshots), and builds a consistent entity — turns from 'absent' to 'the cited source' before the door closes.
Methodology (transparently)
The aggregate statistics are based on 34 observation rows from live search results as a proxy for the source layer AI Overviews and generative answers draw from; we also added a small live sample (5 questions) captured directly from Perplexity for illustration, not statistical generalization. The 'answer shown' figure (67.6%) means the result mix matched snippet/answer triggers, not a visual confirmation for every query. Source-type classification is judgment-based with margin for error, and the sample is small and SEO/GEO-specific, so it doesn't generalize. Next deepening step: expand direct capture across ChatGPT, Gemini, and AI Overviews with documented screenshots.
Frequently asked
How was the data collected?+
By running 34 real Arabic queries through live search and recording the top sources (168), then classifying each by type and Arabic-native origin. This is the source layer AI answers rely on, not a direct capture from inside them.
What is the 'citation vacuum'?+
Answers exist, but no single trusted Arabic source is cited consistently — just scattered domains with repetitive content. That leaves the 'reference' seat open for whoever fills it with expertise and numbers.
Does this mean Arabic SEO is bad?+
No. It means the market is mature in demand but poor in sources documented with experience and numbers. That's a supply gap — an opportunity for anyone offering real evidence instead of generic promises.
How do I benefit as a business owner?+
Target the neglected angles with a direct answer and documented evidence (numbers, screenshots, case studies), a citable structure, and a clear author identity — so you become the source AI cites before your competitors.
Want to be the cited source?
I build content and growth systems that make brands and experts the reference in their field — on Google and inside AI answers, with documented, numbers-backed results.