Per-product AI visibility usually arrives in this category as a shopping module on the top tier, tracking a single assistant and fed by a catalog somebody uploads from Shopify or a spreadsheet. That is a fair way to track a store and a poor way to track a brand, since buyers ask about a need rather than a SKU. The tools differ more than their pricing pages suggest, and how to choose an AI visibility tool goes through them by name.
Reachd reads products off the brand’s own site instead, and measures them with the ordinary run rather than a separate shopping report. Until early August that measurement still rolled up into a single AI visibility score per brand, and the number belonged to whichever product the assistants already talked about. The strongest product carried the figure while another sat almost unnamed in a category ChatGPT answers every day.
The headline score is still one number per brand and per market, since that is the job of the card at the top of the dashboard. Underneath it, every tracked product now carries its own score, its own competitors and its own sources, and each row of the product panel opens into that world. For a brand nobody has measured yet, the free AI visibility analysis gives the whole-brand picture.
Product-level numbers here are not a shopping report. They are the same measurement the brand’s own score is made of, run once per product.
Which competitors your product loses to in AI answers
Rivals are ranked over the answers to that product’s own category, and your product sits in the same list at the place its score earns. Pooling every question used to hand the largest product eight of the ten slots, its second product one and the third none at all. The leader of a small category always loses to the mid-table of a large one, so each product now runs in the contest it is genuinely in.
What AI reads when it answers your product category
Under the contest sit the third-party pages assistants cite on that category, most cited first, and separately your own pages among them. An empty own list is the central finding rather than a blank section, since it means AI decides that product’s market reading nothing you published. Citations stay the raw material of every number here, counted the same way how AI visibility is measured describes it for the brand as a whole.
Where your product is missing while its rivals are listed
The last block is the shelves, meaning the platforms already carrying that product’s competitors, with the state of your own card on each of them. Where nothing is known yet, the row hands you the site: search that settles it in one click. The verdict is refreshed by the same AI brand monitoring cycle that watches ChatGPT, Google AI, Perplexity, Claude, and Grok.
Product-level data in the API and the MCP server
The same world reaches machines, so a product’s competitors, sources and shelves can travel into a client report without anyone opening the dashboard. Product World is a tool on the Reachd MCP server, which lets you ask Claude or ChatGPT about one product directly.
For scheduled reporting it is also GET /v1/businesses/{id}/products/{product} in the read-only AI visibility API, returning the same rows the panel renders. Businesses that resell other companies’ goods keep a single market instead, tracked by what they sell in, and the panel stays hidden where a business has no products of its own.
Frequently Asked Questions
How is AI visibility measured for a single product?
Questions are assigned to a product by the customer need they ask about, never by which product the answer happens to name. The score for that product is the share of those answers recommending it, read over the same window as the brand's headline number.
Can SKU level visibility be tracked across ChatGPT, Google AI, Perplexity, Claude and Grok?
Yes. Every tracked product is asked about on all five assistants in the same run, and the per-product panel reports each one over the same window, so the numbers stay comparable.
Why does ChatGPT recommend a competitor's product instead of ours?
The drilldown answers this in two blocks. The contest names the products winning that category, and the source list shows the pages assistants read to get there, which is where an absent brand usually turns out to be missing.
Can product level AI visibility data be pulled through an API?
Yes, through GET /v1/businesses/{id}/products/{product} in the read-only public API, and through the Product World tool on the MCP server for asking about a product inside Claude or ChatGPT.
Which plan includes product level AI visibility?
Every plan, starting at $99 a month. It is part of the ordinary measurement rather than an enterprise shopping add-on. Starter covers ChatGPT and Google AI, and Pro adds Perplexity, Claude and Grok to the same per-product view.