OpenAI’s ChatGPT Ads is rolling out three updates that shift how campaigns are targeted, tracked and bid: an automated Maximize results bid strategy, platform‑level placement controls (iOS, Android, Web) and one‑day view‑through conversion reporting. The package also includes a WorkMagic integration for consolidated reporting and conversions API support, and it’s arriving in Brazil and Mexico.
What changed
Advertisers can now choose a Maximize results bid strategy that automatically sets and adjusts bids to pursue the campaign’s selected objective. Instead of manually editing individual bids, teams pick the campaign goal and let ChatGPT Ads’ automated bidding system optimize toward that outcome.
Campaign creation adds an Eligible platforms setting so advertisers can restrict delivery to iOS apps, Android apps or the Web. That control helps isolate surfaces where conversion rates or customer value differ materially.
ChatGPT Ads Manager will report one‑day view‑through conversions at the campaign, ad group and ad levels. A view‑through conversion is recorded when a user converts within 24 hours of seeing an ad without a credited click — a common attribution signal that typically increases reported conversions tied to impressions rather than clicks.
Finally, the platform now integrates with WorkMagic so teams can view ChatGPT campaign metrics alongside other channels. The WorkMagic connection can also send conversion signals back to OpenAI via a Conversions API, giving advertisers a server‑side route to feed outcome data into the ad platform.
Why this matters
These features answer two practical advertiser needs: more precise placement control and better capture of conversion signals — while also introducing more automation. Platform targeting reduces wasted impressions and simplifies surface‑level comparisons. Conversion API support and the WorkMagic integration can recover signals that client‑side tracking misses, notably for app events and server‑side conversions.
At the same time, view‑through attribution and automated bidding create familiar trade‑offs. View‑through conversions expand the pool of attributed actions because they credit conversions without a recorded click. That can paint a different performance picture from click‑through conversions and should be treated as a separate signal when assessing ROI.
Automated bidding reduces operational overhead but reduces granular control. The platform’s Maximize results mode is designed to generate the most outcomes from the budget, but it may favor volume over efficiency in some contexts. Marketers need to verify whether those incremental conversions match their value expectations.
Practical next steps for ad teams
1) Run controlled tests: Split‑test Maximize results against your current bidding setup. Track click‑through and view‑through conversions separately so you can assess incremental quality, not just raw volume.
2) Segment by platform: Use Eligible platforms to isolate Web and app audiences. If lifetime value or conversion behavior differs between surfaces, separating them will clarify where budget performs best.
3) Harden conversion tracking: Consider the Conversions API via WorkMagic to improve signal completeness, especially for app installs or server‑side events that often drop out of client‑side tracking.
4) Watch attribution metrics: When view‑through conversions appear, monitor cost per action, post‑conversion value and retention to detect inflated short‑term reporting that doesn’t translate into long‑term value.
What to watch next
Marketers should watch three outcomes closely: whether Maximize results consistently outperforms manual or conservative bidding across verticals; how view‑through conversions influence budget allocation and reported ROI; and whether OpenAI expands these controls and attribution options to additional regions and ad surfaces. For performance teams, the immediate work is pragmatic: test automation, use platform targeting to reduce noise, and treat view‑through figures as a distinct attribution input rather than a replacement for click‑based conversions.
