Although you write in Spanish, language models can expand searches to multiple sources—many of them in English—to get better results. In this article, we explain what fan-out queries are, why this language shift occurs, and how to optimize your content so you don't lose visibility in this process.
If your content strategy is designed for Spanish-speaking markets and you are betting on appearing in ChatGPT Search answers, there is a data point that changes how you should be working on your content.
Peec AI, an AI search analytics company, analyzed more than 10 million prompts and 20 million fan-out queries from ChatGPT and found something that is not yet on the radar of most marketing teams: 43% of the background searches generated by ChatGPT are done in English, even when the user wrote their question in another language.
For markets like Spain, Mexico, Argentina, or Colombia, this has direct implications for which brands and which content end up being cited in the answers.
What Are Fan-out Queries and Why They Matter
When a user asks a question in ChatGPT Search, the system doesn't just reply with what it already knows. It generates a series of background searches—fan-out queries—that it sends to search providers to find up-to-date sources.
First, it reformulates the original query into one or more specific searches. Then, depending on the initial results, it may generate additional, more focused searches.
OpenAI's public documentation describes this process but doesn't explain how the system decides which language to use for those intermediate searches. That is exactly what Peec AI investigated.
The critical point is this: fan-out queries are the filter prior to citation. Before ChatGPT decides which sources to cite, it decides which sources to consult.
If the background searches are done in English, the sources being considered are predominantly in English, and local sources are left out of the process before selection even begins.
What the Peec AI Analysis Found
Peec AI filtered its dataset to include only cases where the IP location matched the prompt language.
Polish prompts from Polish IPs, German from German IPs, Spanish from Spanish IPs. Cases with mixed signals were excluded.
The results within that filtered dataset showed that 78% of non-English prompts included at least one fan-out query in English. No non-English language in the dataset fell below 60%.
Turkish was the most affected, with 94% of its prompts generating fan-outs in English. Spanish was the lowest of the analyzed group, at 66%, but it is still a significant proportion.
The pattern emerging from the analysis is consistent: ChatGPT tends to start fan-out queries in the prompt's language but adds English searches as it builds the response.
Concrete Examples of How This Affects Results
Peec AI included several cases in its report that illustrate the effect in practice.
A Polish prompt from a Polish IP asking for the best auction portals produced an answer that omitted or sidelined Allegro.pl—which according to Peec AI is the dominant ecommerce platform in Poland—in favor of eBay and other global platforms.
A German prompt asking for German software companies produced a response without any German companies.
The most illustrative example is the Spanish prompt about cosmetic brands. Peec AI showed the actual fan-out queries generated by ChatGPT. The first search was done in English. The second was done in Spanish, but it added the word "globales" (global), a qualifier the user never used. The system interpreted a Spanish prompt from a Spanish IP as a request for global brands, not local ones.
These are examples from Peec AI's tests, not representative data of all ChatGPT Search behavior. But the pattern is consistent with the English bias shown by the aggregated numbers.
Why This Is Different from Traditional Citation Factors
Over the past year, several studies have analyzed what factors predict whether ChatGPT cites a source. SE Ranking published an analysis on citation signals. The Tow Center investigated attribution accuracy. These studies assume the content is already under consideration and analyze what makes it more or less citable.
What the Peec AI analysis shows is a preceding problem: the language of the fan-out queries can filter which sources are even considered before the selection starts. If your content is in Spanish and the background search is done in English, you never make it to the selection phase.
This changes the diagnosis for content teams in Spanish-speaking markets. It's not just a matter of domain authority, content structure, or citation signals. It's a matter of whether the content's language excludes you from the process from the very beginning.
What Content Teams Can Do About This
Evaluate Whether English Versions of Strategic Content Make Sense
For brands operating in Spanish-speaking markets but competing in categories where English content dominates globally, having English versions of the most important articles could be the difference between entering or not entering the fan-out process.
It's not a universal strategy, but for categories where ChatGPT clearly favors global English sources, ignoring that channel comes at a cost.
Monitor What Fan-out Queries ChatGPT Generates for Your Key Topics
Tools like Peec AI allow you to see what background searches ChatGPT generates when asked about topics relevant to your business.
Understanding in what language those searches are being performed and what sources are appearing is the starting point for knowing if there is a structural visibility problem or if your content is already being considered.
Include English Terminology Within Spanish Content
An intermediate alternative is to ensure that Spanish content includes technical terms and proper names in English where appropriate.
If ChatGPT does a fan-out in English looking for "best auction portals Poland" and your Polish content only mentions "mejores portales de subastas," the semantic match is lower.
Using the terminology the system is likely searching for increases the chances of being considered.
Don't Abandon Local Search Optimization
Peec AI's analysis describes a system bias, not a rule.
Spanish was the language with the lowest proportion of English fan-outs in the dataset, at 66%. That means the remaining 34% of Spanish prompts generated background searches exclusively or mostly in Spanish.
Well-optimized content for Spanish-speaking audiences remains relevant, especially for queries with clearly local intent.
What We Still Don't Know
OpenAI has not publicly explained how ChatGPT decides the language of its fan-out queries. It is not clear if the bias toward English is an intentional design decision or an emergent behavior of the system.
It is also unclear if OpenAI has plans to adjust this to better reflect local markets.
The Peec AI dataset comes from its own platform, not from a representative sample of real ChatGPT user sessions. The analyzed prompts reflect the use cases of Peec AI clients, who are primarily SEO and marketing teams. This may overrepresent certain types of queries and categories.
What is clear is that fan-out queries are a real and documented mechanism of ChatGPT Search, that the language of those searches affects which sources are considered, and that content teams in non-English speaking markets have an additional factor to monitor in their AI visibility strategy.