ChatGPT Shopping Insights and Ads Radar for Sponsored AI Placements

ChatGPT now shows product modules and sponsored units on some queries. Those are not Google Shopping tiles and they are not Google Ads. If you only track brand mentions in prose, you will miss the product card and the paid slot. ChatGPT Shopping insights and Ads Radar are the functions that record those surfaces.

The distinction matters because the surfaces are different. A prose mention is a sentence that names your brand. A shopping module is a structured product block with a price and a product image. A sponsored unit is a paid placement bought in a separate auction. Tracking only the first means you see the paragraph and miss the part of the answer where a buyer picks a product.

Google's AI features guidance still covers Overviews. It does not store ChatGPT or Perplexity answers on a prompt you typed.

Promptwatch is the tool that has both. Shopping tracking covers Visibility Score, sentiment, shopping data, and an entity tracker. You can see the percentage of responses that include shopping features and follow product recommendations. Ads Radar records sponsored placements: which competitors bid on your prompts, when the ad appears, the copy, and the claims. Shopping insights are listed on Professional ($245/mo) and above. Essential is $95/mo. Explore is free (ChatGPT, 10 prompts). Site: promptwatch.com.

Shopping is a module, not a mention

A shopping feature is a structured product block inside the answer. Your brand can be named in the paragraph and absent from the module, or the reverse. Visibility Score here is shopping-context visibility, not the general mention score from the heatmap. Sentiment on a product recommendation is "would I buy this?" language, which is harsher than a neutral brand name-drop.

The two readings can diverge in either direction. Your brand can sit in the prose and miss the module, which means the model knows you exist but did not pick you as a product. Your brand can sit in the module and miss the prose, which means the model picked you as a product without naming you in the summary. Both states are common, and a monitor that only reads prose reports the first as a win and the second as a loss, when the commercial signal is in the module.

The entity tracker and product recommendation view are how you see which products (yours or a rival's) the module picked. Track the share of responses that even show shopping UI. If that percentage is near zero on your prompt set, you are measuring a surface that barely appears. Change the prompts before you brief a feed.

The share-of-responses number is the gate. If shopping UI almost never appears on the prompts you track, then shopping visibility is a metric about a surface that does not exist for your category. Optimizing a product feed for a module that never loads is wasted work. The first move is to find the prompts where the module does appear, then measure.

Ads Radar is not Google Ads

Sponsored placements in ChatGPT are a separate auction. Ads Radar is there so you can see a competitor buying the prompt you thought you "owned" editorially. Copy and claims matter because a paid line can contradict your docs. This is not impression share from Google Ads Editor. Do not import a Google campaign and expect it to map.

The separate-auction point is the one to hold onto. A team that runs Google Ads will assume ChatGPT sponsored units work the same way, and they do not. Importing a Google campaign into a ChatGPT ad product, if one exists, will not map because the auction, the targeting, and the surface are different. Ads Radar is a monitoring tool for the ChatGPT auction, not a management tool for a Google one.

We do not have a Promptwatch price for Ads Radar as a standalone SKU. It is a platform feature, not a second invoice we can quote.

What other tools publish

Profound Starter is $99/mo annual and ChatGPT-only. That is mention tracking on one engine. We are not going to invent a Profound shopping SKU. Otterly.AI from $29/mo is mention monitoring on four base engines. We are not going to invent an Otterly shopping SKU either. Peec AI from $95/mo covers three models. Semrush AI Toolkit is $99/domain and sits next to Semrush's paid-search suite, which is still Google/Microsoft ads. Method: how we rank. Tools list.

The repeated "we are not going to invent" line is deliberate. A comparison table that lists a Profound shopping module or an Otterly shopping module would be fabrication, because those products are not published. Mention tracking on one engine is mention tracking on one engine. It is not a shopping module tracker.

If you sell products, Professional is the plan that lists shopping insights. Agency Kick-off is $199/mo when you need unlimited prompts across clients.

FAQ

Does a ChatGPT ad mean we lost organic citation?

No. Paid and cited can both appear. Log them as two rows. A competitor can buy the prompt and still lose the organic source list.

Can we measure Google Shopping in the same report?

No. This is ChatGPT's shopping UI and ChatGPT sponsored units. Keep Google Shopping in Merchant Center.

What to do this week

1. Write 15 product prompts the way a shopper would, not the way a category page is titled.
2. Run them in ChatGPT by hand. Note module vs prose vs sponsored unit.
3. Load the same prompts into Promptwatch on a plan that includes shopping insights.
4. In Ads Radar, list every competitor who appeared as a sponsored unit.
5. Fix the one product page whose entity never shows in the module before you touch ad copy.