Skip to main content
Product Here you can see detailed product-level insights: how products perform, how they’re grouped within the categories you’ve set up, and your full uploaded catalog.

​
Product table

Image The Products table is where the underlying signal appears. Each row is a product in your catalog, with the metrics Peec AI aggregates from the chats it tracks. Start here to understand what AI Shopping measures. The table shows these metrics:
Position is the metric to watch. In a chat, users usually see the first two or three results before they have to scroll, and on mobile that scroll is horizontal. A product that appears often but sits in position eight is barely seen.

​
Attributes

When ChatGPT compares products, it builds a table of attributes based on the user’s question and the context available to it. Peec AI captures those attributes and tracks how often each appears across your chats. Use this to:
  • Optimize your product detail pages: If the AI consistently evaluates products on attributes you don’t mention on your own pages, it has less to work with when deciding whether to recommend you.
  • Cover the attributes that matter: Covering the attributes that matter most makes it easier for the model to choose your product. For the full breakdown, see how each attribute scores against competitors in the detail view.

​
Product detail page

Image The product detail view is where AI Shopping becomes most actionable. Open it by clicking any product in the table. It narrows every metric to a single product for the time range you select, giving deeper insight into that product over time.

​
Main KPIs

A summary of the selected product performance over your chosen time period:
  • Visibility: Percentage of AI responses where this product appears.
  • Win rate: The share of chats in which the product is the number one recommendation.
  • Position: Average ranking when this product appears in AI responses (lower is better).
  • Appearances: The absolute number of chats the product appeared in.
  • Rating: Average star rating (0–5) across AI answers that rate this product.
  • Mentioned price: The price the AI cites for the product. Compare it to the catalog price to see whether assistants are quoting you higher, lower, or accurately.
  • Catalog price: The price from your uploaded catalog.
Change indicators show how each metric moved compared with the previous period of the same length.
Use the filters at the top to analyze more granularly.
As you scroll, you’ll find graphs, each telling a different story:
  • Visibility & win rate: Share of AI answers mentioning this product, and how often it is the #1 recommendation, over time.
  • Competing products: The products ChatGPT recommends alongside this product in the same answer. This shows your competitive set at the product level, helping you understand which specific products are being compared with or recommended alongside yours, rather than only looking at competition at the brand level.
  • Position versus competitors: A box plot showing where this product typically ranks compared with the products it competes against, sorted by visibility. Beyond the median position, it also shows how much rankings vary and highlights any outliers, giving you a fuller picture of how consistently each product ranks.
  • Merchants: Where AI is sending shoppers for this product.
  • Sources: The websites AI retrieved while answering about this product.
  • Shopping queries: The total number of shopping searches for the selected timeframe. (Product carousels are built from Google Shopping category searches rather than web searches, which makes these queries the most important factor in improving your product’s carousel visibility.)
  • Fanout queries: The web searches triggered alongside those chats, which the assistant uses to generate recommendations, reasoning, and comparisons. (Unlike shopping queries, which determine which products appear in the carousel, fanout queries influence the content and context surrounding them.)
  • Prompts & chats: The exact prompts and chats this product shows up in, and where it wins or loses.
A product that doesn’t appear often but ranks well when it does looks very different from one that appears constantly in a weak position, and that difference is what tells you where to focus.

​
Attributes

When the AI compares products, it evaluates them on attributes. The Products table shows which attributes come up and how often; this section shows how your product scores on each one against its competitors. Attributes are split into three tabs by how the AI measures them:
  • Characteristics: These are attributes with text values. Expand one to see how often the AI mentions each value for you versus your competitors.
  • Facts: These are yes/no and numeric specs that the AI states about each product.
  • Ratings: These are attributes the AI scores on a scale, averaged per brand.
The competitor columns are fixed per product, ranked by how much each brand overlaps with the attributes the AI evaluates yours on. An empty cell in your column where competitors have values is a signal: the AI couldn’t find that information about your product. Cover it on your product pages, and you give the model more to work with.

​
Prompts & chat table

  • Prompts: The exact prompts this product appears in, along with visibility, position, and win rate for each. This gives you the clearest view of where you’re winning and where you’re losing. Sort by the metric that matters most to find the prompts worth optimizing.
  • Chats: The actual AI conversations where this product appeared. Open any chat to read the full answer — what was asked, how the product was presented, and what else was recommended alongside it. If a metric surprises you, this is where you’ll find out why.