How to check whether Google's AI recommends your Shopify store

Checking whether you rank on Google is easy: search, scroll, count. Checking whether Google’s AI recommends your store is harder. There’s no position number, answers change between asks, and the answer may name three brands and stop. But you can get a clear enough picture by hand in about fifteen minutes. Here’s how.

Step 1: Write ten questions a shopper would ask

Use the words your customers use, not your internal product names. A mix that covers the main situations:

• Three “best” questions: “What are the best linen bedding brands?”
• Two with a budget: “Good linen sheets under $150?”
• Two with a need: “Best sheets for hot sleepers”, “linen bedding that ships to Canada”
• One comparison: “[Your brand] vs [the competitor you lose to most]”
• Two about you by name: “Is [your brand] a good brand for linen sheets?” and “[your brand] reviews”

Write your brand name the way a customer would type it, with spaces and capitals, not the way it appears in your web address.

Step 2: Ask them in the right places

• Google search. Some questions show an AI Overview at the top, some don’t. Note which do, and which sources the Overview links to.
• Google AI Mode (the AI Mode tab in search) and the Gemini app. These always answer, so you’ll see who gets named even when search shows no Overview.
• ChatGPT, if your customers are likely to use it. Many do for shopping research now.

Step 3: Write down what you see

One row per question per place. A spreadsheet is fine. These example rows are made up:

Question                         | Where     | Named? | Who instead            | Sources cited
best linen bedding brands        | AI Mode   | no     | Parachute, Brooklinen  | 2 roundups, Reddit
linen sheets under $150          | Gemini    | no     | Amazon, Target         | -
Is [brand] good for linen sheets | AI Mode   | yes    | -                      | your site, 1 review

Ask the same questions again on another day. One answer is noise; a pattern across ten questions on two days is something you can act on.

Step 4: Read the pattern

• The AI doesn’t know you by name. If the brand questions get “I’m not familiar with…”, the problem is awareness. You need to be written about somewhere credible before category questions are realistic.
• It knows you, but you’re missing from the category questions. That’s the most common pattern. The fix is getting into the roundups and threads the AI cites for those questions. The “sources cited” column is your outreach list.
• It gets your facts wrong. Wrong materials, prices or shipping regions point to messy product data. Clean up product types and descriptions, and check your product schema.
• It sends shoppers to a big retailer. If the answers say “try Nordstrom” or “check Amazon”, you’re competing with the store rather than another brand. That’s what happened in more than half the answers in my study of 42 Shopify stores.

When a tool is worth it

Doing this by hand once is worth it, because reading the answers yourself teaches you a lot. Doing it every month across several AI engines gets tedious. That’s what tracking tools are for. I compare the main ones, including mine, in AI visibility tools for Shopify stores.

If you want a quick first read without the spreadsheet, the free Geomint check below builds twelve questions from your own catalog, asks Gemini, and shows every answer word for word. It covers Gemini only and runs once, so it’s a starting point, not a tracker.