Products monitored
How RankX AI chooses the products it watches for in AI shopping answers, what the Products screen tells you about each one, and how to act on a finding.
The Products screen shows which of your products appeared in AI shopping answers, on which shopper questions, and against whom. It is built from the products you monitor: a set you choose for each Website, up to your plan's allowance of ten products per shopping prompt, kept current against your store every night.
Choosing the products you monitor
RankX AI watches for the products you choose, which keeps every check pointed at what you most want to sell. A store with fifty products and one with a hundred thousand choose the same way. There are three ways to add a product, all in Settings → Products Monitored:
- Search your connected store. A connected Shopify or WooCommerce store can be searched by name, and you pick the products to watch.
- Accept a suggestion. When your own store has already been seen selling a product in an AI shopping answer, RankX AI offers it back to you as a suggestion, so a product that keeps appearing is one click from being monitored.
- Upload a product file. A Shopify product export is recognised as it comes.
Choosing and refreshing products costs nothing. RankX AI refreshes the chosen set against your live store every night, so prices and stock stay current, and the metered action in AI Shopping is the check itself. A product removed from your store stops being monitored and keeps its history.
Monitor what you sell on the questions you track
RankX AI recognises your product in an answer by matching each product card a surface showed against the products you monitor. So the useful set is the products a shopper could plausibly be recommended on your tracked questions: your best sellers, and anything you are actively working to place.
A card whose merchant names your own store is still recognised as yours when the product is not in your set. It appears under your store's name on Merchants, labelled as your store, and it is exactly what the suggestions above are built from.
What the screen shows
For each of your products that has appeared:
- Which shopper questions it appeared on. This is the useful column: it tells you which intents your product is already an answer to.
- How often, against how many answers showed products at all for that question.
- Which merchants shared those answers. Every carousel has several cards, so appearing is not the same as winning, and who you appeared beside is a real competitive fact.
Every share divides by that prompt's own carousels, not by a site total. A product that appeared on two of two carousels for a niche question has a different meaning from one that appeared on two of forty for a broad one, and dividing both by the site total would erase the difference.
Reading it
Start from the questions, not from the products. A product appearing on questions nobody asks is not a win, and a product absent from the one question your category buys on is the finding.
Absence on a question where the carousel fired is real. The surface had a product opinion on that question and your product was not in it. That is actionable.
Absence on a question where no carousel fired is not. The surface had no product opinion at all, and nothing about your products caused that.
Compare against Merchants on the same question. Who took the slot, and whether that merchant is a genuine rival or a marketplace listing your own product, is usually the whole answer.
Acting on a finding: listing-copy proposals
RankX AI can turn a finding into new listing copy for a monitored product. From the product's own page it proposes a new SEO title, SEO description and product description, written from the dimensions the assistants compared on, and shows each one beside your current words. Nothing is written until you approve it, and RankX AI then writes the approved fields to your store and reads them back.
Three things shape a proposal:
- It uses only facts already in your own product text. A specification that is not in your listing never enters the proposal, so approving one cannot put a new claim on your shop.
- It keeps the product title as it is. The title is what the assistants match a card on, so RankX AI improves the description and search fields around it.
- It writes to a connected store that RankX AI can write to. On WooCommerce the SEO fields are offered where the store's SEO plugin exposes them.
Each proposal is priced; the rate is on the credit cost reference. After a change, the shopping question's page tells you whether anything moved, and says "too early to tell" until there are enough checks to say more.
The diagnosis check
RankX AI can run a separate check for a specific product to work out why it is not appearing: whether the product is present on the shopping surface at all, and what the listing looks like from that side.
It spends credits and it says so before it runs. It is worth running for a product you believe should be appearing and is not, and the specific answer matters most for a product you are about to act on.
What this screen covers
The cards each surface showed. RankX AI records the cards and their order; the ranking behind them is the vendor's own.
Visibility, not attribution. Where RankX AI reports traffic and sales it comes from Conversions from AI, Search Console and Analytics, which are first-party and named as such.
The questions you track. Shopping prompts are the panel, and the surfaces are observed on them.
Where to go next
- AI Shopping, including how to switch the module on.
- Merchants, for who appears instead.
- Shopping surfaces, for what each surface exposes.
- Shopify and WooCommerce, for connecting a store.
Last updated
Shopping surfaces
The two AI shopping surfaces RankX AI observes, how they differ in what they expose, and the one difference that changes how every ads figure must be read.
Merchants
The merchants whose products were recommended in AI shopping answers alongside or instead of yours, each with its share of the offers, and how to read the list.