Last updated August 2026.
On Aug. 11, 2026, OpenAI expanded ChatGPT Ads to the UK, Mexico, Brazil, Japan, and South Korea, adding five countries to its shopping-ad rollout, according to Digiday and Modern Retail. The same week, Otterly.AI’s ongoing 2026 tracking reported a separate and more consequential number: more than three in four ChatGPT shopping-intent answers now surface a sponsored product carousel, a paid placement block that runs alongside the generated answer text.
That is the number that matters if you run a GEO program for a retail or DTC brand. A 76.4% sponsored rate means the “get cited organically” pitch for shopping queries now competes directly with a paid unit sitting in the same answer, on most of the prompts you care about.
This page breaks down what that split looks like by prompt type, and gives you a five-step audit to run before you decide how much of your GEO budget goes to organic content versus paid placement.
What a sponsored carousel actually is
A sponsored product carousel is a horizontal, paid unit of product cards (image, price, and a buy link) that ChatGPT inserts into a shopping-intent answer. It is distinct from the organic answer text, which still cites and links to sources the way it always has.
The carousel does not replace the organic answer. It sits above or alongside it, and it is the first thing a shopper sees. That is the shift: your organic citation used to have the top of the answer to itself. Now, on most shopping prompts, it shares that space with a paid unit.
Anatomy of a ChatGPT shopping answer in 2026
A typical shopping-intent answer now has three layers, in this order:
- Sponsored product carousel. Paid product cards, shown on 76.4% of shopping-intent prompts. Advertisers bid for placement; the position is not earned through content.
- Generated answer text with organic citations. The model’s synthesized answer, still built from retrieved sources and still where your content, schema, and third-party mentions earn a citation.
- Follow-up source links. A short list of linked domains the model drew on, which is where your brand’s organic presence shows up if it earned the citation.
Layer 1 is pay-to-play. Layers 2 and 3 are where organic GEO work still pays off, and where the rest of this page is focused.
Where organic citations still win, by prompt type
Not every query in a retail category behaves like a shopping-intent prompt. The sponsored carousel targets purchase-ready queries specifically. Here is how the split looks across the three prompt types that make up most retail GEO programs.
| Prompt type | Example | Sponsored carousel appears | What still wins organically |
|---|---|---|---|
| Shopping intent | ”best noise-canceling headphones under $200” | 76.4% of the time (Otterly.AI, August 2026) | The answer text and follow-up source links; your product page competes with editorial roundups for citation, not for the carousel |
| Informational | ”how do I clean suede shoes” | Rare; no purchase-ready signal for the ad system to target | Nearly the whole answer; this is where structured, direct-answer content earns citation share |
| Comparison | ”[Product A] vs [Product B]“ | Occasional, and usually smaller than on pure shopping prompts | The verdict paragraph and any comparison table the model can extract from a cited page |
The takeaway: your organic GEO effort has the most room to work on informational and comparison prompts. On pure shopping prompts, organic citation still matters, but it is no longer the only thing competing for the shopper’s first glance.
The 5-step category audit
Run this before you plan next quarter’s GEO spend for shopping queries.
- List your top 10 shopping-intent prompts. Use the phrasing real buyers use: “best [category] under $[price],” “[category] for [use case].”
- Run each prompt in ChatGPT and record whether a sponsored carousel appears. Note which brands bought the slot.
- Separately record what the organic answer text cites. This is the layer your content and third-party mentions can still win.
- Calculate your category’s sponsored rate. If it tracks close to the 76.4% industry figure, treat carousel space as a paid-media decision, not a content one.
- Score your organic citation presence in the answer text and source links. That score, not the carousel, is what your GEO content work should move.
Data box: sponsored rate, as of August 2026
The 76.4% figure is a snapshot, not a constant. OpenAI is still expanding ChatGPT Ads by country and category, and Otterly.AI updates its tracking regularly. Treat the number as directional for your planning and re-check it each quarter.
For a deeper on-page and editorial-placement playbook for DTC brands losing shopping citations to editorial sites, see DTC in the Age of Agentic Shopping, updated this month with this data.
Tools that can measure your own split
Otterly.AI is the source of the 76.4% industry figure and tracks sponsored-versus-organic placement at the prompt level, so it is the natural starting point if you want a like-for-like read on the same metric.
Profound adds citation-source mapping, showing which specific domains win the organic slot in your category once you know how much of the answer is paid. Temso covers citation tracking for shopping and informational prompts alongside content execution, from $89 a month, so the audit above and the fixes it points to live in one subscription rather than two tools. Peec AI offers a similar monitoring lens with strong coverage of buying-intent prompts, useful for cross-checking your own numbers against a second source.
None of these tools change the 76.4% number. What they change is whether you know your own category’s number before you spend against it.
For the full side-by-side of GEO platforms, including pricing and engine coverage, see /rankings/geo-tools. Definitions for citation share and related terms are at /glossary.
Run the 5-step audit on your top 10 shopping prompts this week. Then check the full GEO tool comparison at /rankings/geo-tools to pick the tracker that fits your category’s sponsored-to-organic split.