ChatGPT Shopping Insights and Ads Radar for Sponsored AI Placements
How we turn on Promptwatch ChatGPT Shopping and Ads Radar after a mention baseline, and how we keep sponsored placements off the organic slide.
ChatGPT can name a brand in prose, show a shopping unit, or show a sponsored placement. Those are three rows. Mixing them is how ecommerce QBRs go wrong. We turn on Promptwatch ChatGPT Shopping and Ads Radar after the prompt list is stable, not instead of it.
We do not publish Shopping conversion rates we did not measure. Visitor analytics (script or GTM) is the only place we will call a session.
Shopping insights
ChatGPT Shopping tracking covers Visibility Score, sentiment, shopping-specific data, and an entity tracker. Shopping Insights shows the share of responses that include shopping features and whether your own products appear. Product recommendation tracking is "who got named in a shopping context," which is not the same as a citation on a how-to prompt.
We load the products the PIM already sells. We do not let an agent invent SKUs. Content Agents can draft a category explainer into Webflow or Framer and wait in the review inbox. Specs come from the catalog or they do not ship. WordPress publish is still coming soon.
Professional at $245/mo lists shopping insights on the brand side. Business is $579/mo. Essential at $95/mo is the mention-and-citation program we start on. Explore is free ChatGPT, 10 prompts, and is not a shopping stack. Kick-off at $199/mo is how we hold several catalogs (unlimited projects and prompts, 10 seats). G2: 4.7/5.
Paid engines remain ChatGPT, Gemini, Claude, Perplexity, Grok, Llama, DeepSeek, Mistral, Copilot, AI Overviews, and AI Mode. Shopping is a ChatGPT surface. We do not pretend Gemini has the same unit.
Ads Radar
Ads Radar tracks sponsored placements in AI search: which competitors bid on your prompts, when they appear, ad copy, claims, and positioning. That is paid competitive intel. It is not organic GEO.
We put Radar on a separate slide. If a competitor bought the unit, the fix might be a media conversation, not a rewrite of the FAQ. If nobody is buying and you are still absent from the shopping module, that is a product-feed and entity problem, not an H1 problem.
Otterly at $29 (Gemini add-on, lag) will not give you this split. Peec at $95 is mentions. Profound Starter at $99 annual is ChatGPT mentions without this commerce pair. We will not invent a BrightLocal or GeoRanker shopping SKU to fill the cell.
Daily paid checks. No instant alerts. No $29 Promptwatch plan.
What we will not do
We will not call a sponsored unit a citation. We will not promise that appearing in Shopping equals revenue. We will not turn Radar on in week one while aliases are still wrong. We will not autopublish a comparison page that invents a competitor price we saw in an ad.
FAQ
Do we need Shopping if we are B2B and have no cart?
Usually no. Keep prompt tracking and citations. Turn Shopping on if ChatGPT already shows a product module on your category prompts. If it never does, Radar may still matter if someone is buying the query.
Can we see Shopping in Explore?
Explore is 10 ChatGPT prompts for a demo. It is not the shopping insights SKU we run on a catalog. Use Essential or above, then Professional if shopping insights are on the plan you need.
How we run it
- Freeze prompts and aliases. Baseline mention vs citation for two weeks.
- Turn on Shopping Insights for the SKUs the PIM lists. Turn on Ads Radar on the same prompt set.
- Split the monthly slide: organic mention, shopping module, sponsored unit.
- Measure clicks in visitor analytics if a session arrives. Do not invent a CVR.
- hello@1001seomedia.com if you want the commerce views on the retainer. Bring the product list. Do not bring a promised Shopping rank.