Chapter 10
When language and place both changed, the names changed
Imagine two children in a toy shop. One asks a robot, in English, for the top toy shops in Europe. The other asks in Lithuanian about toy shops in Lithuania. A shop known only in Lithuania comes up for the second child.
Your customers can ask in any language, about any place. A local name can show up in one answer and be missing from another.
So I asked robots about well-known companies from eleven European countries, in twelve languages.
The short version. Local names came up far more when the question used the language of the company’s head office and named the area around it. Those two things changed together, so I cannot say which one mattered. When you test, change one thing at a time.
What I did
The robots could search the web.
Local names came up far more in the home version
I used two questions that did not name the company, asked in English and in the company’s home language. Local names came up far more in the home version. For big global names, the gap was much smaller.
Named means only that the answer said the name, in praise or in passing.
For a Lithuanian power company, the English questions named Europe or “International”. The Lithuanian versions named the Baltic states or Lithuania. Language and place changed together.
Picture it. Each box is one mix of language and place: English on top, home language below, a wide area on the left, home on the right. Green boxes are mixes I asked. Grey boxes stayed empty.
At Javier’s: like trying 2 of 4 mixes, old flour in the old oven and new flour in the new oven. If the pizza tastes different, nobody can say why.
A tone program saw gaps
A second program gave each answer a mark from minus one to plus one. I compared the English and home-language marks for each company.
Picture it. Each folder is one company. On a green folder, the marks for its two sets of answers sat more than 0.15 apart. On a white folder, they sat closer.
At Daniel’s: like a program reading 65 client files, each in German and in English, and giving the German copy the higher mark by more than a set margin on 29. The English copies run longer, and the program reads only the start of each.
Do not trust such a mark on its own. The English answers were the longest, and I have no count of what length does to the mark. So read the answers yourself.
Repeats differed more by robot than by language
In another part of the study, I asked a few questions five times each, one company at a time.
Picture it. Each parcel is one set: five answers from one robot, in one language, to one question. Green means a computer found the five very close in meaning.
At Marta’s: like 100 customers each ordering the same thing five times, and 87 of them getting five boxes that look almost the same.
How to check this for your business
Giulia asks a robot, in English, for the top accountants in Europe. The answer lists big firms. Then Giulia asks in Italian about accountants in Milan, and the list changes.
Was that the language or the place? Giulia changed both, so there is no way to tell yet. Next comes the mix Giulia skipped: English, and Milan.
Try this on Monday: fill the empty box
Take one question from your chapter 1 sheet. On one day, with the same town, ask it ten times for each robot in your sheet’s language, then again in English, in a new chat each time. If your sheet is in English, use another language your customers speak. A translation website can help.
Compare the two sets of ticks, and read what each answer says about you. You changed one thing, so this is cleaner than mine.
References
Chapter 10. When language and place both changed, the names changed
Preprint. Dmitrij Żatuchin (2026). The Language Blind Spot: How Query Language and Brand Recognition Tier Shape AI-Constructed Brand Reputation Across Twelve European Languages. arXiv:2606.23165. Data. 10.5281/zenodo.20794390.
What does AI say about your business?
Put the book into practice. Tell us your company and website to request an AI visibility audit. We’ll reply by email to agree what to check.
What the audit shows
- Where your company appears in the answers, and next to whom
- The pages AI cites in your market
- What to publish first on your own site
