Brands that sell in several countries often run one marketing playbook everywhere. That rarely holds up, because buyers in each market ask different questions, trust different sources and compare different brands.
AI search turns out to work the same way. The newest Semrush AI Visibility Index, published in October 2026, added a UK benchmark next to the US for the first time. Two markets that share a language and use the same AI models still got noticeably different answers. The release puts it directly: there is no single, global approach by which AI visibility is achieved.
What the study measured
The Index ran fixed monthly sets of 2.5 million UK prompts and 25 million US prompts across ChatGPT, Gemini and Google AI Mode, plus a separate set of Google AI Overviews results for real search keywords, from January to July 2026, across 22 categories. It compares brands that AI answers mention with the domains they cite as sources.
Four ways the UK and US differ
The platforms agree less in the UK. On average, any two platforms shared 63 to 70 of their top 100 mentioned brands in the UK, against 70 to 78 in the US. The UK platforms agreed less on both brands and sources in all seven months measured.
Fewer UK brands hold all four platforms. PPC Land reports that 16.7% to 26.0% of the UK domains that reached at least one platform's top 100 held all four, against 23.5% to 36.5% in the US.
Each market has its own always-visible brands. Semrush calls them the Universal Core: brands in the top 100 mentions on all four platforms in every month. There were 24 in the UK and 36 in the US, and only 15 were shared. Six of the nine UK-only domains were ecommerce or retail brands, including Tesco, Argos and Asda.
The source shelf changes by market. The Next Web reported that news and information sites took 29.5% of citations among ChatGPT's top 100 UK source domains on average, against 19.6% in the US. In sports, the UK share of source positions held by brands that were also mentioned was at least 35 points lower than in the US in every month.
Local domains count separately
One detail matters a lot for brands that run several country sites. The Index records domains as they appear, so one company can show up twice. Amazon appears as amazon.com in the shared core and as amazon.co.uk in the UK-only list. Ford qualified for the report's famous but uncited group through ford.com in five UK months and ford.co.uk in the other two.
If you run a .com alongside a .in, .ae, .sg or .co.id site, AI engines may treat them as different sources, and your strength on one does not automatically carry to the other.
What I expect in India, the Gulf and Southeast Asia
The study covers only the UK and US. Nothing below comes from it. This is my view, based on how these markets work.
If two English-speaking markets already diverge, markets with more languages and more local platforms are likely to diverge further. Buyers in India mix English and Hindi in the same question. Buyers in the UAE and Saudi Arabia move between English and Arabic. In Indonesia and Malaysia, Bahasa and Malay prompts sit next to English ones. Each language can pull a different set of sources, and each market has its own marketplaces, review sites and publishers that AI engines may lean on.
That is why we report results for India, for the Gulf and for Southeast Asia separately, and by language, rather than as one regional score.
How to test whether your market translates
- Write the same buyer questions for each market. Keep the intent identical, then localise the city, currency words and product names.
- Run each set in the languages buyers actually use. Log English and local language results separately.
- Record mentions and cited URLs per market. Note which of your country domains, if any, gets cited.
- Compare the source shelves. List the third-party domains each market leans on. Expect them to be different, and plan outreach market by market.
- Re-run on the same dates each month. Single readings move around, so look for what keeps recurring.
Our use cases by industry show example buyer questions you can adapt for your category.
Reading the study fairly
Semrush, now owned by Adobe, sells AI visibility tools, and the report promotes them. The study measures how often a result recurs across months, not growth or trends, and it does not claim that visibility changes caused commercial results. Its value is the size and consistency of the market differences it found.
For anyone selling across several countries, the practical point is clear. Do not copy one market's AI visibility plan into the next one. Measure each market first.