FoundRateHow to Check If ChatGPT Recommends Your Store (3 Ways)
If shoppers are asking ChatGPT what to buy in your category, there are only two possibilities: it recommends your store, or it recommends someone else's. You cannot fix a problem you cannot see, so the first step is finding out which one is true. Here are three ways to check, from quickest to most reliable.
1. Ask ChatGPT the questions your shoppers ask
Open ChatGPT and type the questions a real buyer would type, not your brand name. Brand-name searches ("tell me about [My Store]") always flatter you and prove nothing. Use buying-intent questions instead:
Good test questions
• "What's the best [your product category] under $[price]?"
• "Where can I buy a [material/style] [product] online?"
• "Recommend a [product] for [use case / who it's for]."
Run each a few times. Note whether your store is named, where it ranks in the list, and which competitors show up instead. That competitor list is gold: it tells you exactly who is winning the sale you're losing.
2. Test in a clean, logged-out session
Here is the trap most merchants fall into. If you're logged into ChatGPT, it may personalize answers based on your history, and if you've been visiting your own store, it can surface you for reasons a real shopper would never trigger. That gives you a falsely rosy result.
To get closer to what an actual shopper sees, test in a logged-out session or a private/incognito window, and turn off any memory or personalization. Ask the same buying questions and compare. If you appear when logged in but vanish when logged out, you are effectively invisible to new customers.
3. Test systematically, not once
The biggest problem with checking by hand is that a single question on a single day is noise, not signal. ChatGPT's answers vary run to run, shift as it retrieves different web results, and change over weeks. Asking once and seeing your store might be luck; asking once and not seeing it might be an off day.
A reliable read means asking many buying questions that cover your catalog, repeating them, and tracking the results over time so you can see a real Found Rate (the share of buying questions where you're named) instead of a gut feeling. Doing that by hand across dozens of questions, logged out, on a schedule, is tedious and easy to get wrong, which is exactly the gap FoundRate was built to close.
What to do with the answer
If you are being recommended: good, now protect it. AI answers drift as competitors improve their data and as models update, so re-check periodically. If you're not: the fix is almost always your product data, not your prices or your ad budget. Work through the seven highest-impact fixes, and see the nine checks that decide whether AI can read your products at all.
For the background on how ChatGPT chooses what to name in the first place, start with Does ChatGPT Recommend Shopify Products?
FoundRate runs buying questions against ChatGPT for your real catalog, in a clean session, on a schedule, and shows your Found Rate, your position, and exactly which competitors win each search. Run your free scan.