Artificial Intelligence
OpenAI plans image ads and broader measurement in ChatGPT
OpenAI said it would test image ads in ChatGPT in the US and widen measurement, attribution and brand-suitability partnerships as it tried to answer advertiser concerns about spend and inventory.
- OpenAI had sold ads in ChatGPT since February, and advertisers had complained about measurement, formats and inventory.
- Later that month, OpenAI planned to begin testing image ads in the US during image generation, with the ads labelled and kept separate from ChatGPT-created images.
- The company expanded its measurement tools and partner integrations with Kochava and 10 other partners.
- Geo-based incrementality remained in early testing with Haus, Measured and WorkMagic, while brand-suitability pilots involved DoubleVerify and Integral Ad Science.
- Shamsul Chowdhury of Zeno Group said ChatGPT had fewer prompts and only one ad placement, which made supply tighter than on Google.
Since February, OpenAI had sold ads in ChatGPT, and advertisers had raised the same three complaints: they wanted better measurement, more formats and more inventory. OpenAI's latest changes addressed two of those points.
The company planned to start testing image ads in the US later that month. The test would run during image generation, and the ads would be labelled and kept separate from images made by ChatGPT.
OpenAI also expanded its in-house measurement tools and partner integrations with Kochava and 10 other partners. Those tools were meant to let advertisers assess web and app outcomes, including results beyond last click.
Geo-based incrementality was still in early experimentation with Haus, Measured and WorkMagic. OpenAI was also running brand-suitability pilots with DoubleVerify and Integral Ad Science in controlled environments, and said those partners would not get access to real user conversations.
Shamsul Chowdhury, Zeno Group's executive vice-president for paid social, said ChatGPT handled about 2.5 billion prompts a day globally, compared with about 15 billion Google searches. He said the smaller volume and single ad placement created a gap between supply and demand, and OpenAI executives were due to discuss the product at Advertising Week New York.
Named in this story
People
- Shamsul Chowdhury
- said ChatGPT had less supply than Google for advertisers
Companies
- OpenAI
- said it would test image ads and expand measurement and brand-suitability work in ChatGPT
- Kochava
- was among the measurement partners added by OpenAI
- Haus
- was one of the partners in early geo-based incrementality testing
- Measured
- was one of the partners in early geo-based incrementality testing
- WorkMagic
- was one of the partners in early geo-based incrementality testing
- DoubleVerify
- took part in brand-suitability pilots with OpenAI
- Integral Ad Science
- took part in brand-suitability pilots with OpenAI
- Zeno Group
- employed Shamsul Chowdhury, who commented on ad supply and demand
Products and systems
- ChatGPT
- hosted the ad inventory and image-generation placement discussed in the story
How the source tells it
The piece was mostly neutral trade reporting, with only mild launch timing emphasis.
No emotive framing was found in the original.