Guides
Are your products recommended by AIs?
More and more purchases start with a question asked to an AI: "what is the best product for…", "which brand should I choose for…". The assistant answers by naming specific products. If yours never appear, you lose sales before the customer even opens a store — yours or a competitor's.
The AI has become a recommendation shelf
A hesitating customer no longer just compares product pages: they ask ChatGPT, Gemini or Perplexity for advice, as they would have asked a salesperson. The AI answers with a short list — three or four products, sometimes with a highlighted "best choice". That answer steers the purchase, often before any visit to a shop.
For a product brand or an online store, this is a new shelf: the one where the assistant places the items it recommends. Being on it means entering the selection; being off it means leaving the space to the products the AI knows better.
A known brand does not mean a cited product
A common confusion is to assume that if the AI knows your brand, it recommends your products. It is not automatic. A brand can be well described overall without any specific product standing out when a customer asks which one to buy. Brand visibility and product recommendation are two distinct things.
Conversely, a product can be named without the brand being highlighted — because it is described in comparisons, reviews or clear pages the AI could pick up. What matters for sales is what happens at the product level.
What makes a product stand out
AIs pick up what is clearly described and mentioned in several places. A product stands out more easily when its page is precise (what it is for, for whom, its concrete features), when it appears in comparisons or third-party reviews, and when its name is stable and unambiguous. A vague page, a too-generic name or a total lack of external mentions make the product invisible to the assistant.
Price, promotions or advertising play no part: the AI does not sell placement. It recommends what it understands and finds described — hence the importance of clear product information, on your site and elsewhere.
How to check for yourself
Do not ask "tell me about my product": ask the question a customer who does not know you yet would ask. "What is the best product for this use?", "which brand do you recommend for this need and budget?". See whether one of your products appears, in which position, and next to which competitors.
Do it across several AIs — their answers differ markedly — and repeat over time. This manual test, run across several products, phrasings and markets, quickly becomes heavy: this is exactly what our pre-diagnosis measures automatically, for free and with no account.
What it changes for your sales
When a customer asks for a recommendation and the AI names three products without yours, you are not losing a spot in a ranking: you are absent from the decision. This first selection increasingly happens at the assistant, ahead of your store and your campaigns.
Being in those answers, correctly described and at the right moment, puts you in the selection with no extra ad spend. Knowing whether your products are there is therefore a commercial metric, just like your conversion rate or your customer reviews.
Frequently asked questions
Does the AI recommend products based on price?
No. Unlike a marketplace, the AI does not push a product because it is the cheapest or because someone paid for it. It recommends what it understands and finds clearly described. Price may be mentioned if it appears in the sources, but it does not buy the recommendation.
Do you need a large catalogue to be cited?
No. A single, clearly described product picked up by a few reliable sources stands out better than a large catalogue of vague pages. Clarity beats volume.
My products change often, is that a problem?
It can be, if AIs rely on old references: a discontinued product may keep being cited, and a new one stay invisible until the sources are updated. One more reason to measure regularly.
Are marketplaces like Amazon enough?
They help, because AIs often read those pages and their reviews. But relying only on them makes you dependent on how your product is described there. Clear information under your own control remains the most solid basis.