ChatGPT Shopping and Alexa product results: what to track and what to fix
Product answers in ChatGPT carousels and Amazon Alexa are a new shelf. A tracking guide for ecommerce teams: catalog hygiene, what to monitor, and where Shopping Insights helps.
Quick verdict
When someone asks an assistant what to buy, the answer can be a carousel of products or a spoken recommendation, not a list of links. Brands in those results get the sale, and brands outside them never see the shopper. Tracking this takes a different approach from mention tracking, because the unit is a product, not a brand. Promptwatch covers it with AI Shopping, which tracks ChatGPT shopping carousels and Amazon Alexa product results, and it lets you upload your own catalog.
Why product answers need their own tracking
A normal AI visibility panel asks "which brands are named?" Shopping answers behave differently:
- The result is a set of products with images, prices, and merchants.
- Position inside the carousel matters.
- The same brand can have several products, each with its own presence.
- Merchants, not just brands, compete for the slot.
We cannot give you a figure for how many purchases flow through these surfaces, and we will not make one up. What we can say is that you can measure your presence in them directly.
Step 1: define your product panel
Pick the products that matter: best sellers, margin leaders, new launches. Write the shopping prompts a buyer would use, such as a product type plus a constraint, like a budget or a feature.
Promptwatch's Shopping Insights, which starts on Business, lists shopping items and products over time, with product position analytics, top merchant domains, and top products. You can also add tracked products to follow specific items.
Step 2: upload or maintain your catalog
If your product data is wrong, the engines will show wrong data. Check:
| Field | Common problem |
|---|---|
| Title | Keyword-stuffed or inconsistent across channels |
| Price | Out of date on your page versus feed |
| Availability | Out-of-stock items still listed |
| Images | Low quality or missing |
| Attributes | Size, color, material absent |
| Identifiers | Missing or mismatched across sellers |
Promptwatch supports own-catalog upload for AI Shopping, which lets you compare what you intended with what appears.
Step 3: track the merchants
In shopping answers, a competing retailer can capture the slot for your product. Look at the top merchant domains for your tracked prompts. If a marketplace dominates, decide whether to compete there, ensure your listings are complete there, or build a stronger direct page.
Step 4: add Alexa
Amazon Alexa is among the engines Promptwatch monitors through the real interface. Alexa product results are spoken or on-screen answers, where a single recommendation can matter more than a list. Track whether your products are named, how they are described, and which competitors are named alongside.
Step 5: connect product data to your site
Your product pages need to be reachable and clear. Check that OAI-SearchBot and ChatGPT-User can fetch them. Agent Analytics crawler logs, from Professional, show fetches and errors. Use structured, accurate product descriptions on the page so the content agrees with the feed.
Step 6: measure traffic and sales
Visitor analytics records AI-referred traffic and conversions. For ecommerce, tie it to purchase events in GA4 or your store platform. Remember that some shopping journeys end in a store or a marketplace and never reach your site.
Ads and shopping together
Sponsored placements can appear beside shopping answers. Promptwatch's Ads Radar captures sponsored ads in ChatGPT Search, Google AI Overviews, and Google AI Mode answers to tracked prompts, with the advertiser, target URL, and triggering prompt. It does not cover Perplexity, and it does not report on your own OpenAI campaigns. Ads Radar starts on Business and is on every agency plan.
A short weekly routine
- Check position for your top ten tracked products.
- Note new competing products and merchants.
- Review catalog problems found.
- Compare AI-referred conversions to last week.
- Pick one fix: a feed field, a product page, or a missing listing.
Promptwatch's Action Board can propose evidence-backed actions each week from the Discovery agent, up to five per run.
Plan notes
Shopping Insights starts on Business at $579 a month, with 5 projects, 5 seats, 350 prompts, and 42,000 responses. It is not on Essential ($95) or Professional ($245), and agency plans do not include Shopping Insights. Ads Radar and state and city tracking are on Business and every agency plan. Explore is free, ChatGPT only.
Other tools in the ranking cover ecommerce angles. Our ecommerce roundup in the blog lists some options.
Who should pick which
| You are... | Recommendation |
|---|---|
| A small online shop | Fix your catalog data and track by hand |
| A mid-size retailer | Promptwatch Business for Shopping Insights |
| An agency with ecommerce clients | Agency plans for Ads Radar and local tracking, accept no Shopping Insights |
| A marketplace seller | Track top merchants, strengthen listings there |
| Selling spoken-commerce items | Track Alexa explicitly |
Start with clean product data, because it is the cheapest fix. Then monitor with Promptwatch Business, which is where product positions, merchants, ads, and visitor data meet.