How ChatGPT picks which stores to recommend

Every month we ask ChatGPT and Gemini hundreds of real buying questions and record which stores they name and which sources they cite. This is what we see.

The Krillo Index today: 3043 stores measured in 65 categories, in the Netherlands and Belgium.

It searches, then summarises

For a buying question the assistant usually searches the web first, reads a few pages (often comparison articles, review sites and large retailers) and names the stores those pages mention.

Named is not the same as recommended

Some stores appear in many answers but are never the tip. Others are named less often but recommended every time. The difference is usually what the sources say about them: a clear reason to choose that store.

The same question, a different answer

Answers vary between runs. That is why one check proves little and we count over many questions and runs, per category, per month.

Check whether AI names your store

Free, no account. Your rank in your category, who AI names instead, and 13 checks on your site.

Run the free check See the rankings

Questions

Can I pay to be recommended by ChatGPT?

Not in the organic answer. Being named comes from what the assistant reads about you.

Does a bigger store always win?

No. In many categories a specialist store is named before a large marketplace, because the sources say it is the specialist.