Being named is not the same as being recommended

Say someone asks an AI assistant where to buy good kitchenware. The answer names five stores. You are one of them. Good news, you would think.

But read how you are mentioned. 'There are also smaller sellers such as yourstore.nl' is very different from 'for pans that last, yourstore.nl is the best choice'. In the first case you are a footnote. In the second case you are the advice.

Four levels

It goes up roughly like this. You do not appear at all. You are named as one of several options. You are named with a reason. Or a specific product of yours is recommended.

That last one is worth by far the most. Someone who is told that a specific product at your store is good does not need to look any further. Someone who sees your name in a list of five usually clicks on the first one.

Why stores get stuck here

Being named often happens by itself once an assistant can read your site and understands what you sell. Being recommended takes more: somewhere there has to be something that shows why you are a good choice. A clear reason, a specialism, a product description that goes further than size and colour.

Assistants do not rely only on your own site for this. They also pick up what has been written about you elsewhere: comparisons, blogs, forums, reviews. If you appear nowhere outside your own site, there is little for a recommendation to rest on.

What to do with this

Do not just check whether you appear, but how. Find the sentence in the answer where you are named and read it word for word. Is there a reason given? No? Then you know what to work on, and that is something other than producing more content.

The Krillo Index on krilloai.com counts this difference separately. For each store you see how often it was named and how often it was actually recommended, based on around 30 buying questions per category, put to ChatGPT and Gemini every month.