Region & Language-specific search queries

[safarsin]

Hey everyone! 👋

To make the automation look even more natural, I want to add a new feature: Region & Language-specific search queries.

Currently, the tool uses a generic dataset. If we add localized datasets (for example, 10,000 unique queries in Spanish, German, French, Hindi, etc.), the search behavior will look much more realistic for your specific location.

I need your help:
Before I start setting up the code and the dataset structures, I want to know which languages and regions you actually use and want to see.

📝 Please take 1 minute to fill out this quick form: Google Form

What's next?
Based on your votes, I'll prioritize the most requested languages. Later, I'll create a guide on how you can contribute your own query datasets (either via Pull Requests or by simply uploading a .txt file) and get a Contributor badge on the repository!

Let me know what you think about this idea in the comments below!

[JeromeM]

Done ;)

[safarsin]

Appreciate it, man! 🤝

[safarsin]

Quick update!

The datasets folder and initial templates (French, Italian, Czech, and regional English) are now live.

Check out CONTRIBUTING.md for a quick guide on how to help.

Let's make this happen!

[JeromeM]

I was already brainstorming on this subject.
I wanted to use https://trends.google.com/trends/ dynamically to avoid having a "static" file with queries.
It would be more natural for MS and probably be less dangerous for a ban.

The datasets might be used in a fallback if we can't catch the trending searches.

What do you think ?

[safarsin]

That is a good one, and it'd definitely make the behavior look more organic.

However, I've looked into this before, and my main concern is that Google Trends is difficult to scrape dynamically. They have very aggressive rate-limiting and bot protection. If users are running the app daily (especially with multiple accounts), Google might block their IP or throw 429 errors very quickly.

I've heard about Google Trends RSS feeds that don't trigger bot protection and have no limits. But the problem is that it gives 20-25 searches.

I think we can build a hybrid system RSS feed (for fresh, daily trends) + local datasets (as a fallback and to add the number of queries).

[JeromeM]

Maybe we can use something else.
Like https://wikimedia.org/api/rest_v1/metrics/pageviews/top/en.wikipedia/all-access/2026/06/28 ?

Just trying to figure out how we can automatize the searches and have localized random searches.
I will dig deeper this afternoon and try to have a solid solution based on the country of the client.

[safarsin]

I think Wikipedia is what people read, not what they search. It'd be no different with local datasets

Maybe we can use a hybrid system: RSS feeds + Wiki top pages.

This is what RSS looks like, I forgot to attach it in the last reply:
https://trends.google.com/trending/rss?geo=FR

The queries in the <title> selector

[JeromeM]

Yes I saw the google page. I've also tried some parameters to see if we can have more results, but nothing is working.

An hybrid system looks good. I will make a POC to see if it can works and be reliable.