Social Rocket
For some advertisers, measurement remains a drag on ChatGPT Ads. Digiday reports on 1 October 2026 that agencies, despite running campaigns, are seeing missing conversions, delayed reporting or discrepancies between platform and in-house click tracking. The problems don't affect all testers equally, but they are enough to keep larger budgets partly in test mode.
Agencies see different measurement problems
According to Digiday, the UK agency Accuracast reports a lead campaign where form submissions came in, yet ChatGPT Ads still showed no conversions two weeks later. In another case, the platform showed around 100 clicks, while the agency's own measurement registered only about 20.
Other agencies such as AdRoll and Jellyfish reported no comparable click discrepancies, according to the report. At Web Guide Partner, conversions reportedly arrive rather late: roughly 24 to 36 hours after the event.
Above all, this shows how immature comparability still is. For performance teams, a single platform figure is not enough when campaigns need to be judged against Google, Meta or other established channels.
Data access is part of the problem
OpenAI has significantly expanded its advertising platform over the past few months. The company itself writes that CPC- and outcome-optimised bidding now make up the majority of campaigns, and that the pixel and Conversions API are key foundations for measurement and optimisation. OpenAI also says it works with more than 50 technology and measurement partners.
Even so, not all advertisers enable conversion tracking or custom audiences. Digiday reports legal and privacy-related reservations among companies that would have to connect their own customer data to the platform for this. Without these signals, it becomes harder to optimise for CPA or ROAS.
This very dependence between platform data and advertising effectiveness is also the topic of our overview of AI in media buying.
For German advertisers, the question is already practical
ChatGPT Ads are now also available in Germany. Since the end of August, OpenAI has offered self-service access via the Ads Manager in Europe. This means the measurement question is no longer just a US pilot problem.
If you test campaigns, you should therefore cross-check platform data against your own analytics, CRM or server-side measurement, and define before launch which metric will decide on scaling. The already introduced ChatGPT Ads features around Sponsored Agents, HubSpot and Shopify make this point even more important, because the funnel increasingly plays out within ChatGPT.
OpenAI's Help Center still lists aggregated performance data such as views and clicks for user reports and announces additional measurement options step by step. For large performance budgets, what remains decisive is how reliably the pixel, Conversions API, partner measurement and independent attribution work together in day-to-day practice.