Vendor reportGraphite and Similarweb, 2026
Prompts that name a brand are much more likely to show an ad
Audience targeting
| Outcome | Names a Brand (%) | Names No Brand (%) |
|---|---|---|
| An Ad Appeared | 61% | 39% |
| No Ad Appeared | 36% | 64% |
ChatGPT Ads: Share of prompts naming a brand, split by whether a competitor ad appeared (%)
- Reader Note:
- The linked table splits ChatGPT prompts by whether they named a brand and whether an ad appeared. In the figure, among prompts where an ad appeared, 61% named a brand and 39% named no brand; among prompts with no ad, 36% named a brand and 64% named no brand. This is an observed association from Similarweb's panel, not a causal test of targeting rules, and the source calls the related pre-ad exchange analysis early and directional.
- Evidence:
- Observed behaviour
- Vendor interest:
- Moderate — This comes from a company with a commercial interest in this market, so it is not a neutral source.
- Sample:
- Similarweb user panel across 9 live ChatGPT ad markets (panel size not disclosed), plus Graphite analysis of 95k+ US ChatGPT ads from 500+ advertisers, 70k+ creatives, 50k+ headlines, and 8k+ landing pages from April to mid-August 2026.
Good for
Comparing yourself against a similar segment, and choosing what to test next.
Not for
- Saying that one thing caused another, or predicting what you would get.
- This comes from a company with a commercial interest in this market, so it is not a neutral source.
- The source never says how this number was measured.
- The source says how much it looked at, but not how any of it was chosen.




