ChatGPT Ads reporting controls expand while product-feed bulk uploads remain unresolved

What changed

Reporting controls are documented

OpenAI’s Ads API documentation describes selectable attribution windows and reporting time bases for Insights. These settings affect the basis on which results are counted and dated, so two systems using the same conversion name can still produce different totals. Reporting controls do not, by themselves, establish which bidding, optimisation or catalogue capabilities are available to an advertiser.

Save the selected attribution window and time basis with each export or analysis. When reconciling ChatGPT Ads results with analytics or an internal reporting system, compare definitions and event timing before treating a difference as a tracking fault. No particular setting can be recommended from the available information.

Product-feed workflow needs verification

September reporting described spreadsheet creation for product-feed campaigns and product-level review status. OpenAI’s current campaign quickstart, however, explicitly says bulk upload does not support product-feed campaigns. Those accounts conflict. Product-feed bulk upload should therefore be treated as unresolved, not as an assured capability or a basis for catalogue planning.

A useful evaluation separates the two evidence streams. First, verify the current help guidance and the workflow visible in the relevant account. If an apparently eligible product-feed option is present, test it with a small non-production example where permitted and record the result, date and account context. The existence of product-level review information would not independently prove that bulk creation is supported.

For a reconciliation test, hold the conversion definition constant and compare results under explicitly recorded time and attribution settings. For a catalogue test, define success as an accepted and reviewable upload rather than simply finding a spreadsheet control. These different tests prevent a confirmed reporting feature from being used to validate a disputed campaign-creation claim.

Why it matters for advertisers

Recorded attribution choices make reporting differences easier to explain. The unresolved bulk-upload conflict is operationally separate and makes it premature to design a product-feed workflow around the reported capability.