MUMBAI: AI chatbots can recommend different brands to consumers asking the same buying question depending on their city, language and the platform they use, according to a new study by marketing company FTA Global.
The findings are part of “The Context Report”, which analysed 1,200 AI-generated answers across 10 business categories, 12 cities and seven languages on ChatGPT and Gemini. Each question was run five times to assess how consistently recommendations changed.
The study found that 159 of the 206 brands identified across the research appeared in only one country, suggesting that AI-generated brand discovery can vary significantly between markets rather than reflecting a single global view of visibility.
The difference was particularly pronounced in payment gateways. When asked which payment gateway to use in New York, ChatGPT and Gemini recommended brands including Stripe, Square and PayPal. The same question asked in Mumbai resulted in recommendations including Razorpay, Cashfree and PayU, with the two city-level results sharing no common brand.
The report found that nearly three of the top five recommended brands changed when the buyer moved away from New York and English. Financial categories showed the highest geographic variation, with an average of 4.4 of the top five brands changing between markets, compared with 1.5 for categories such as video conferencing and cloud storage.
In HR software, none of the top US brands identified in the study appeared in any of the other countries tested.
The study also found differences between AI platforms. ChatGPT and Gemini shared only about 3.2 of the top five brands for the same question asked in the same city.
AI recommendations also varied across repeated tests. The same question produced a consistent top brand in only 35 per cent of repeated tests, even when the location and language remained unchanged.
Local content also played a role. In Germany, 42 per cent of the sources cited by AI systems were German web pages, highlighting the influence of local-language content on answers to location-specific buying questions.
FTA Global founder and managing director Senthil Kumar Hariram said the findings indicate that brands need to look beyond a single AI visibility score and assess where, in what context and how consistently they are being discovered.
Alongside the report, FTA Global has introduced Context Authority, a metric designed to assess how consistently a brand appears in buying-related AI responses across relevant cities, languages and AI platforms.
The company said the approach moves away from treating AI visibility as a single global measure and instead evaluates brand presence across multiple contextual variables and repeated AI responses.
The findings come as AI tools increasingly become part of consumers’ research and consideration journeys. FTA Global said visibility in one market or one AI response may not necessarily translate into visibility across the different contexts in which consumers make purchase decisions.