AI search news ·
Do ChatGPT ads get your brand into the answer? 96.4% of advertisers aren't cited above their own ad
ChatGPT now shows a sponsored placement on 25.94% of commercial prompts — but paying for the slot almost never puts the advertiser in the answer above it. That is the sharpest finding in SE Ranking's study of 50,006 commercial prompts, published 2026-08-10: the advertiser's domain appeared among the answer's cited sources in just 3.63% of placements, the exact advertised URL in 0.09% (12 cases), and even a bare brand mention in only 4.44%. The ad layer and the answer layer are, on this evidence, two separate distributions — which matches what our own corpus sees from the organic side.
Primary source: SE Ranking — ChatGPT ads study, 50,006 commercial prompts across 20 US niches (published 2026-08-10)
What SE Ranking measured
OpenAI announced advertising in ChatGPT on 2026-01-16 — ads at the bottom of answers, US free and Go tiers only, with an explicit "answer independence" principle: "Ads do not influence the answers ChatGPT gives you… Ads are always separate and clearly labeled." SE Ranking ran 50,006 commercial prompts across 20 niches in the US — the same prompt set as its earlier Google AI Mode ads research — recording the generated answer, its cited sources, and any sponsored placement. 12,974 prompts returned an ad: 25.94%, close to the 29.45% the same team previously measured in AI Mode. Every observed ad sat below the answer, one sponsored offer per answer, no competing ad beside it.
The paid slot and the organic answer barely overlap
| Overlap between ad and the answer above it | ChatGPT | Google AI Mode (same team's prior study) |
|---|---|---|
| Advertiser's domain among cited sources | 3.63% | 11.53% |
| Exact advertised URL cited | 0.09% | 1.95% |
| Advertiser brand mentioned in the answer text | 4.44% | — |
In 96.37% of placements the advertiser is not cited in the answer they advertise under. Buying the sponsored slot buys the sponsored slot — nothing in this dataset suggests it buys retrieval, citation, or a mention. That is OpenAI's stated design, and SE Ranking's numbers say the wall is, so far, holding.
Our first-party check: the biggest ChatGPT advertisers are organically invisible in our corpus
The study names ChatGPT's most concentrated advertisers: BestMoney alone accounts for at least 13.56% of all observed ads, Hungryroot for 79.92% of Food & Beverage ads, Booking.com for 62.57% of Travel, Robinhood for 52.24% of Finance. We checked the ten advertiser domains named in the study against our own instrument — the three dated runs of 12 small-business buyer questions × 4 engines × 3 samples whose 2026-07-28, 08-03, and 08-10 cuts contain 969 distinct organically cited domains. Nine of the ten never appear in any run — including VistaPrint and Intuit, whose products squarely serve the small businesses our questions are about. The one hit: intuit.com, cited in exactly one answer, on 2026-07-28, by Gemini — not ChatGPT. Different prompt sets, so this is a consistency observation rather than a replication — but it points the same way from the opposite side: heavy paid presence in AI answers coexists with near-zero organic citation presence.
The measurement blind spot, stated plainly
Ads are shown only on the consumer free and Go tiers. The API — which is how our audits, and to our knowledge every automated AI-visibility tracker, read ChatGPT at scale — returns no ad layer at all. So a quarter of what an actual free-tier user sees under a commercial answer is invisible to API-based measurement, ours included. SE Ranking also notes the same query can produce an ad in one session and none in another, so 25.94% is a floor for how often the slot exists, not a per-user constant.
One tension worth watching
OpenAI's January post said ads are "not eligible to appear near sensitive or regulated topics like health, mental health or politics" during its test. SE Ranking measured ads on 28.69% of healthcare prompts (vs 2.64% in AI Mode) and 28.76% of News & Politics prompts (vs 6.8%). Commercial healthcare prompts — insurance, supplements — are not necessarily "sensitive health topics" under OpenAI's policy, so this is not evidence of a broken rule; it is evidence that the boundary is drawn differently than the announcement's language suggested, and 14.35% of all ads were semantically unrelated to their prompt besides.
What this means for businesses that care about AI visibility
- Paid and organic AI visibility are separate budgets now, exactly as they were in classic search: a ChatGPT ad puts you below the answer, while the recommendation inside the answer is still earned by being retrievable and citable. Check which engines actually cite your business before deciding which of the two problems you have.
- If you buy ChatGPT ads, the study's practical findings apply: one slot, no competing ad beside you (for now), no query-level reporting, and a 1-in-7 chance the matching lands your ad somewhere semantically unrelated — SE Ranking's own two-week test campaign drew 97,000+ impressions and 1,263 clicks but few sign-ups.
- Treat any AI-visibility report — including ours — as a reading of the organic answer layer only. The ad layer exists on the consumer surface, is invisible over the API, and varies session to session.
See your number
See which businesses AI names when your client's buyers ask.
Running this for clients? The $249 agency 5-pack audits five businesses, white-labeled.
Who runs this
- Built and operated by Sensara LLC, Atlanta, Georgia — about us and how the audit works.
- See what the report looks like before you run anything — score per engine, the competitors AI names instead of you, and a fix plan.
- We run the same audit on ourselves every week and publish the result: in the latest run AI named AskedAbout in 1 of 144 answers. We report our own numbers the way we report yours.