.webp){kind=small}
OpenAI is connecting ChatGPT Ads to the measurement tools advertisers already run, naming its first set of third-party partners for attribution, conversion data and incrementality. The update, set out in OpenAI’s announcement on 5 October, also brings a visual ad format that will be tested inside image generation, and brand suitability pilots with DoubleVerify and Integral Ad Science (IAS).
The timing matters. ChatGPT Ads passed a $1 billion run rate less than 200 days after its February launch, and advertisers can now grade those campaigns in the same systems that grade the rest of their spend.
Who Is Plugging In
The partners fall into four groups. Hightouch, Tealium and LiveRamp send conversion data from an advertiser’s existing systems into ChatGPT Ads, alongside OpenAI’s own pixel and Conversions API (CAPI).
Its attribution partners support Conversions API, reporting and click attribution across web and app, and the named ones are AppsFlyer, Triple Whale, Adjust, DV Rockerbox, Northbeam, Branch, Singular, Kochava, Airbridge and Tenjin. Fospha, Measured and INCRMNTAL cover full-funnel and advanced measurement.
Incrementality is the least developed piece, and OpenAI says so. It describes that work as still in its early stages, with Haus, Measured and WorkMagic exploring geo-based experiments to estimate the causal impact of ChatGPT Ads. Kantar and Cint are in early-stage testing to measure brand resonance.
Measuring Beyond the Click
The OpenAI Ads Blog argues that people use ChatGPT to work through a decision, so clicking an ad is often not their next step. OpenAI says that in an early analysis of global campaigns optimized for deeper-funnel outcomes, 52.7% of eligible one-day view-through conversions happened within an hour of the matched ad impression. That is OpenAI’s own analysis, not a partner’s.
Reporting now follows that view. Advertisers can choose a click-through attribution window that fits their buying cycle and add a one-day view-through window, which counts conversions after an impression without a click. OpenAI notes that these settings change reports only, not how campaigns optimize.
Optimization has its own update. Advanced conversion optimization learns from eligible click-through and view-through conversions for the advertiser’s chosen goal. Event Quality Score (EQS) rates the pixel and CAPI signals feeding it, so advertisers can see where their data is weak.
OpenAI also published early results that partners measured for individual brands. They are vendor case studies from a young platform, and they say more about what each method can show than about typical performance.
A Visual Format Inside Image Generation
The new ad format uses images to show product inspiration, product usage or the experiences a product makes possible. OpenAI will first test it during image generation in ChatGPT, with ads labeled and kept separate from the image being created. Testing begins later this month in the US with an initial group of advertisers.
The placement was signposted. On 1 October, OpenAI’s Asad Awan named image generation as a place to test interest-based ads. It also follows the larger image and CTA button test on the existing unit, reported in May.
Brand Suitability Without Reading Conversations
OpenAI says its placement guardrails keep ads out of emotionally vulnerable, sensitive or otherwise unsuitable conversations. Qualifying advertisers can now add Negative Phrases for narrower requirements of their own.
The DoubleVerify and IAS pilots let both firms assess how those standards are applied in a controlled testing environment. Neither gets access to private user conversations, which is the constraint every verification method in ChatGPT has to work around.
ChatGPT is creating a fundamentally different advertising environment, where context is defined by the conversation itself rather than a page or video.
The Problem Measurement Does Not Solve
Digiday reports that advertisers have raised three complaints since ads launched in February: measurement, formats and inventory. This update addresses the first two.
Inventory is a supply question. Shamsul Chowdhury, executive vice president of paid social at Zeno Group, puts ChatGPT at about 2.5 billion prompts a day against roughly 15 billion Google searches, with one ad placement on ChatGPT where Google has several.
For advertisers, the practical change is that a ChatGPT test can now be judged in AppsFlyer, Triple Whale or a geo-lift study rather than in OpenAI’s reporting alone. The visual format, if it moves beyond image generation, is where the extra supply would have to come from.
Recap
What has OpenAI changed in ChatGPT Ads measurement?
OpenAI named its first third-party measurement partners. Hightouch, Tealium and LiveRamp send conversion data, attribution firms including AppsFlyer, Triple Whale and Kochava handle attribution, and Haus, Measured and WorkMagic are exploring geo-based incrementality tests.
How does view-through attribution work in ChatGPT Ads?
Advertisers choose a click-through attribution window and can add a one-day view-through window, which counts conversions after an ad impression without a click. OpenAI says these settings change reporting only, while advanced conversion optimization learns from eligible click-through and view-through conversions.
What does the update change for advertisers testing ChatGPT Ads?
ChatGPT campaigns can now be judged in attribution and incrementality tools advertisers already use, rather than in OpenAI’s reporting alone. Inventory is unchanged, and Zeno Group’s Shamsul Chowdhury notes ChatGPT has one ad placement where Google has several.