CiteLens
Research Lab
Research

Turkish vs English: AI Cites Almost Entirely Different Sources

For the same question in Turkish and English, AI Overviews share only 22% of their cited sources — and Turkish gets an AI answer slightly less often.

CiteLens Research · 500 prompts · 2026-06-21
ATBy Alper Tekin · Founder, CiteLens

If your brand is visible in English AI answers, is it visible in Turkish ones too? It is tempting to assume AI works the same across languages, but our data says otherwise. We asked Google's AI the same 444 commercial questions in both English and Turkish and compared the sources it cited in each language.

The overlap is tiny. Turkish and English AI Overviews for the identical query shared only 22% of their cited sources, and English surfaced an AI Overview slightly more often (96% of queries) than Turkish (94%). In practice, the language a customer asks in changes who the AI cites and recommends — these are effectively two different visibility markets.

For any business selling in Turkey, or any multilingual brand, this has a direct consequence: your English-market AI visibility tells you almost nothing about your Turkish one. You must measure each language and market separately. Below are the full results and what they mean for local and multilingual GEO.

22%
source overlap between TR and EN answers
96% / 94%
AI Overview coverage (EN / TR)
444
prompts with an answer in both languages

Key findings

  • Turkish and English AI Overviews for the same query share only 22% of their cited domains — the language you ask in changes who gets cited.
  • English queries surfaced an AI Overview 96% of the time vs 94% for Turkish.
  • A brand cited in English answers cannot assume it appears in Turkish ones; these are effectively two different visibility markets.

What this means for you

If you sell in Turkey, your English-market AI visibility tells you almost nothing about your Turkish one. You must measure each language separately.

CiteLens pins market and language per prompt, so your Turkish audience is measured against Turkish answers — not a US-English proxy.

Methodology

Each prompt was run in English (gl=us, hl=en) and Turkish (gl=tr, hl=tr). Overlap is the Jaccard similarity of the two cited-domain sets, averaged over prompts that returned an AI Overview in both languages.

Download the data (JSON)

Measure your own brand in AI answers

Run the same engine that produced this data on your own brand — see where you're cited and where you're missing.

Frequently asked questions

Do AI answers differ between Turkish and English?

Dramatically. For the same question, Turkish and English AI Overviews shared only 22% of their cited sources in our study — the language changes who gets cited.

Does Google's AI work as well in Turkish as in English?

Slightly less often. AI Overviews appeared for 96% of English queries versus 94% of the same queries in Turkish.

If I'm visible in English AI answers, am I visible in Turkish?

Not necessarily. With only 22% source overlap, English and Turkish are effectively separate AI-visibility markets; you can lead in one and be absent in the other.

Should I optimise my AI visibility per language?

Yes. Because the cited sources differ so much, GEO has to be done per language and market. CiteLens pins the country and language for every prompt so your Turkish audience is measured against Turkish answers, not an English proxy.

Why do Turkish AI answers cite different sources?

Local-language content, local authority sites and different community discussion all shift the source pool, so the AI grounds Turkish answers in a largely different set of domains than English ones.