There's no dashboard for this. No Shopify analytics tab that says "ChatGPT recommended your products 34 times last week." No notification from OpenAI. No automatic reporting.

The only way to know whether ChatGPT is recommending your Shopify products right now is to ask it yourself, systematically, with a process that gives you consistent, comparable data over time.

Here's the audit I run for Shopify stores. Takes about 30 minutes. Tells you exactly where you stand.

Why Doesn't a Native Tracking Tool Exist Yet?

ChatGPT doesn't expose recommendation data to merchants or platforms. Shopify can see referral traffic from chatgpt.com in your analytics, but that only captures the clicks that came through, it doesn't tell you how many times you were mentioned, recommended, or skipped.

Third-party tools like Metricus and Sight AI are building AI visibility tracking capabilities. But they're measuring the same thing this manual audit measures, query-level brand appearance frequency. The manual version is just more work, less automated. Right now, it's also more reliable for seeing real patterns.

What Queries Should You Run in ChatGPT?

Three categories. Run all three to get a complete picture.

Category queries. These test whether you appear when someone is browsing a space without a specific brand in mind, the highest-value recommendation scenario.

• "What are the best [your product category] to buy right now?"
• "Where can I buy [your product type] online?"
• "What brands make the best [your category]?"

Branded queries. These test whether ChatGPT has meaningful brand knowledge about you at all.

• "Tell me about [your brand name]."
• "Is [your brand name] a good brand?"
• "What do customers say about [your brand name]?"

Intent queries. These test whether your products surface when a shopper describes a problem your product solves.

• "I'm looking for [use case your product addresses], what should I buy?"
• "What's the best product for someone who needs [specific benefit]?"
• "I have a budget of $[price range] for [category], what would you recommend?"

Run 10-15 total queries. Use ChatGPT with browsing enabled (the default in GPT-4o) so you're seeing real-time web results, not just training data.

What Are You Looking For in the Results?

Document everything in a spreadsheet. Four columns: query text, brands that appeared, did your brand appear (yes/no), and your brand's rank if mentioned.

After running the audit, look for three things:

1. Brand recognition without product recommendations. ChatGPT knows your brand exists but won't recommend specific products. This is a product data gap, your feed doesn't have enough structured data for ChatGPT to recommend SKUs with confidence.
2. Zero brand mentions across all queries. ChatGPT has no knowledge of your brand. This is a broader visibility problem, missing third-party mentions, no editorial coverage, no review presence. Product data alone won't fix it.
3. Competitors appearing where you don't. This is your optimization roadmap. Every competitor who appears tells you something about what signals you're missing.

How Do You Close the Gap When Competitors Are Appearing and You're Not?

Run the same queries for your top 3 competitors and document what they surface. Then find the gap.

Common patterns I've seen:

• Competitor has 200+ Trustpilot reviews. You have none. That's trust signal gap.
• Competitor has editorial coverage in five industry publications. You have none. That's authority gap.
• Competitor's product descriptions are specific, detailed, and recently updated. Yours aren't. That's data freshness gap.
• Competitor's structured data includes brand, GTIN, and product category across their catalog. Yours is missing fields. That's coverage gap.

Most stores have two or three of these gaps, not all four. Pick the one that's fastest to fix and start there.

The Metricus analysis of ChatGPT Shopping recommendation patterns identifies editorial coverage, Trustpilot presence, and statistic-rich content as the consistent differentiators between brands that appear and brands that don't. Those aren't quick fixes, but they're directional.

How Often Should You Run This Audit?

Monthly is the minimum. Weekly if you're actively optimizing and want to see whether your changes are working.

The key is consistency. Use the same queries every time you run the audit. That's the only way to see whether you're moving in the right direction, a single snapshot tells you where you are, but the trend over 60-90 days tells you whether your work is paying off.

According to Sight AI's guide on competitive AI visibility, the most useful signal is recommendation frequency relative to competitors, not absolute appearance rate. If you're appearing in 3 out of 10 queries and your competitor is appearing in 8 out of 10, that gap is your benchmark. Close the gap, not just the absolute number.

What About Tracking in Google Analytics?

You can track ChatGPT referral traffic by filtering sessions with source containing "chatgpt.com" or "openai.com" in GA4. But this only captures clicks, mentions that didn't result in a click are invisible in analytics.

Use analytics as a corroborating signal, not a primary one. If your manual audit shows increased recommendation frequency and your GA4 ChatGPT referral traffic is also climbing, you know the work is translating. If one moves without the other, investigate.