Articles — Product data for shops
Product data for shops
Wrong price, wrong shop, fake discount
A confident answer can still quote the wrong price, the wrong stock, or a shop that never sold the product. Bad product facts destroy trust, so visibility has to mean more than getting mentioned.
People ask ChatGPT, Claude, Perplexity, or Gemini things like:
“Best running watch with contactless pay under €300 — and where can I buy it in my country?”
Sometimes the answer looks confident and is still wrong: the price is off, stock is listed when it is not there, the product is tied to a shop that never sold it or in the wrong country, the model name is discontinued, or a sale is treated as the normal price so “is this discount real?” gets a bad verdict.
The overall pattern for who gets named is in How AI chatbots build a shortlist.
For a product recommendation, a wrong price is the failure a shopper can check.
AI visibility is two jobs
For a product, showing up is two jobs.
| Layer | Question it answers | Examples |
|---|---|---|
| Trust / content | Why recommend this brand at all? | Reviews, comparisons, credible mentions |
| Data | Can the system describe this product accurately? | Clear names, identifiers, fresh price and availability |
You need both. Strong storytelling will not fix a wrong price. Clean product data will not invent brand authority you do not have.
Why the price or the shop is wrong
The answer is usually not a live lookup of every store’s stock.
The model can repeat an older price, because the text it learned from does not change when the shop changes the price. A search can also pull an old roundup, an affiliate page, or a listing for another country, and then the price comes from that page even when your product page is right.
When the stock or the shop is missing, the model sometimes fills the gap instead of saying it does not know, which is how a shop that never sold the product still gets named. Why that guess beats a hedge is in Why AI won’t just say it doesn’t know.
A product page is not a live stock check. ChatGPT can look up your stock only after you connect it is when that lookup works.
The same item can also appear under different names. The assistant names a product, a shop, or a place is that matching job.
Where the page, the block, and the feed fit
Your product lives in three places is where the page, the structured block, and the catalog file have to carry the same price and the same identity. Structured data publishes facts machines can read is what that block does, and what it does not do.
What brands and shops can do
- Ask real buyer questions in several AI tools and log wrong prices, shops, and stock claims.
- Trace likely sources, such as old articles, marketplaces, and your own outdated pages.
- Make the official facts clear on the product page, in the structured block, and in the feed.
- Fix or remove old listings that stay online after the product changed.
- Treat freshness as part of the product, not a one-time task.
Read next if: Your product lives in three places · Structured data publishes facts machines can read