Why Google shows different results in Paris and Dubai

One query, several results pages

Type the same search from Paris, from Lyon and from Dubai: you will not get the same page. This is neither a bug nor an SEO rumour. Google says so in its own help pages: location is part of the context it uses to improve results, and search always uses the context of the query, even when personalised recommendations are switched off.

For an openly local query (“accountant”, “web agency”, “office rental”), the effect is obvious: the map and the business listings change from one city to the next. What is less visible is that the standard results move too. A competitor based in Lyon can rank above you in Lyon and vanish in Paris, on queries that contain no city name at all.

The four factors that make the page vary

Location is the strongest factor, and the one most often forgotten. Language comes next: Google looks for pages in the user’s language, which makes a French speaker in Dubai very different from an English speaker in the same city. The device changes the layout, and so the number of links visible before scrolling. Finally, context decides which blocks are shown: map, questions, videos, AI-generated summary.

What this changes for an SEO strategy

The first consequence is simple: a ranking does not exist in the abstract. “We are third for this query” means nothing without saying where, in which language and on which device. A report that gives a single ranking per keyword has made a choice for you, often that of an average point in the country that matches no real customer.

The second consequence is commercial. There is no point appearing first in Paris if your target customers are in Dubai, Geneva or Casablanca. The work therefore starts with a question that has nothing technical about it: where are the people who buy from you, and where do they search from? We go through this approach in building your SEO strategy where your customers search.

The third consequence is about method: rankings have to be checked city by city, at scale, and you read the gaps rather than the averages. That is the subject of our method for tracking rankings city by city.

Why this is handled as a system, not by hand

Anyone can check one query from three cities with a browser. Doing it for hundreds of keywords, in several languages and dozens of cities, every week, nobody can. That is where the method becomes a system: thousands of automatic checks by country, city, language and topic, sorted by AI to surface the gaps that matter, then decided by an expert who chooses what to do with them.

We run this system in our rank tracking on Google and in AI. And when the analysis shows that a market needs its own pages, producing them at scale is a job for programmatic SEO.

What about AI engines?

The question does not stop at Google. The assistants that answer questions directly also vary with language and location. We devote an article to it: local GEO, or how to track AI answers by location.

Frequently asked questions

Sources

1. Google help: personalisation of search results (new tab)support.google.com