Skip to content

AI shopping visibility: which products AI names, and which are ready to sell

AI shopping visibility in Citeon AI reads a Shopify shop's products and collections every week and shows, for each one, whether AI answers name it, cite its page, send visits to it, and touch it through your tracked questions. On Optimize it also checks each product page against the fields Google and OpenAI ask of a product.

In short
  • Your catalogue, read every week from your Shopify shop's own product data.
  • Four signals per product: named, cited, AI visits, asked. Never added into one score.
  • Shopping readiness on Optimize: each page checked against Google's and OpenAI's product fields.

What it does

Your visibility score says whether AI names your shop. A shop also needs to know which products it names, which product pages it reads, and whether those pages carry what shopping feeds ask for. Citeon AI reads up to 2,000 products and 200 collections each week, and marks which collections your menu links to. With the Citeon Shopify app connected, it reads up to 5,000 of each every morning through Shopify's Admin API, password protected stores included.

A shop's products, as they look

Products · in AI answers, 28 daysSample data · Acme Running
3collections
4products
6with a signal
2 of 4ready for shopping
  • Acme Glide 3 road shoeworking
    /products/acme-glide-3
    named
    9
    cited
    0
    AI visits
    22
    asked
    77%

    ReadyGoogle: all required fields present; every recommended field present. OpenAI feed: all required fields present.

  • Acme Ridge trail shoenamed only
    /products/acme-ridge-trail
    named
    3
    cited
    0
    AI visits
    0
    asked
    54%

    Not readyGoogle: missing offers.availability; recommended fields missing: aggregateRating, gtin. OpenAI feed: all required fields present.

  • Merino running socksasked, absent
    /products/merino-running-socks
    named
    0
    cited
    0
    AI visits
    0
    asked
    62%

    Not readyNo Product markup on this page: the surfaces have to infer price, stock and brand from the text.

  • Acme Tempo racing flatnot measured
    /products/acme-tempo-racer
    named
    0
    cited
    0
    AI visits
    0
    asked
    –

    ReadyGoogle: all required fields present; recommended fields missing: gtin. OpenAI feed: all required fields present.

Sample data: Acme Running is a fictional shop in Citeon's public demo, shown the way the Products page shows a real one. Named, cited, AI visits and asked are four separate signals, never added into a score; readiness is checked against Google's merchant listing fields and OpenAI's product feed.

Four signals per product

SignalWhat it counts
NamedAnswers that name you and write the product's own title
CitedAnswers that link the product's page
AI visitsSessions from AI assistants that landed on the page; not measured when visits are not tracked
AskedYour tracked questions that share a word with the product, and how often you are named on them

Each product gets a state from them: working, named only, asked but absent, or not measured, with the reason under it.

Shopping readiness

Once a month the site audit reads a sample of your product pages and checks their product markup:

CheckWhat it looks for
Google, requiredname, image, offers.price, offers.priceCurrency
Google, recommendedbrand, sku, gtin, mpn, description, url, aggregateRating, availability, shippingDetails, hasMerchantReturnPolicy
OpenAI product feeditem_id, title, description, url, brand, seller_name, image_url, availability, price
Texta title over 150 characters, a description over 5,000, a seller not named in the offer

Ready means product markup is on the page and every required field is present. A page with no product markup says so: the surfaces then have to infer price, stock and brand from the text.

Shopify's AI channels

Shopify can list a shop's products in ChatGPT, Google AI Mode and Gemini, Microsoft Copilot and Meta through its agentic storefronts, active by default for eligible stores, and set in its admin under Sales channels, Agentic (Shopify Help Center). Citeon AI cannot see that setting, so the Products page gives you the steps to check it once. Readiness is a separate check: it reads the markup on each product's own page. How the two fit together, and what the app writes to your store: the Citeon Shopify app.

Plans

PlanAI shopping visibility
Visibility, 59 euro a monthThe catalogue and its four signals, on three engines
Optimize, 199 euro a monthAdds shopping readiness and AI product descriptions, on six engines
EnterpriseThe same, with Enterprise limits

Prices and the rest of each plan: pricing. See it in the public demo, on a fictional shop.

Questions people ask

Which shops does it work for?

Shopify shops today: Citeon AI reads the product and collection data the shop already publishes, so there is nothing to install. Other platforms are detected but not read yet.

Does Citeon change my product pages?

It does not add product markup: readiness tells you which fields a page is missing, and your theme, an app or your developer adds them. With Citeon's Shopify app connected, the descriptions, SEO titles, meta descriptions and alt text you approve are written to your store, with undo.

Is shopping readiness the same as Merchant Center approval?

No. It checks the product markup on your own page against the fields Google and OpenAI name. Your Merchant Center account and your feed are separate and are not read.

Why four signals and not one score?

Because they mean different things. A product can be named without its page ever being cited, or bring visits from AI without being named. Adding them up would hide that.

What does not measured mean?

Nothing is known yet: no tracked question touches the product, no answer named or cited it, and no AI visit landed on it in the window. It is never shown as a zero score.

Sources

  1. Google Search Central, Merchant listing structured data. Product needs name, image and an offer with price and priceCurrency; brand, gtin, availability, shipping and returns are recommended
  2. OpenAI, product feed specification. item_id, title, description, url, brand, seller_name, image_url, availability and price are required; titles up to 150 characters, descriptions up to 5,000