OpenAI transitions ChatGPT Ads to automated average cost bidding models marking the end of manual max bid caps

The advertising landscape within ChatGPT is undergoing a fundamental structural shift as OpenAI moves to deprecate manual bidding controls in favor of automated, average-cost-based strategies. According to updated documentation released by the company this week, the platform is phasing out "Max Bid" and "Max CPC" (cost-per-click) settings, replacing them with "Target CPA" (cost-per-acquisition) and "Target CPC" models. This transition signifies a broader industry move toward algorithmic bidding, prioritizing system-wide optimization over the granular, manual constraints that have historically allowed advertisers to strictly cap individual auction bids.

The Shift from Manual Caps to Algorithmic Averages

The core of this update lies in the distinction between a "hard cap" and an "average goal." Under the legacy system, advertisers could set a specific maximum amount they were willing to pay for a single interaction. This offered a sense of budgetary safety for performance marketers wary of high-cost spikes. Under the new Target CPA and Target CPC models, advertisers define an average cost they wish to maintain over the life of the campaign.

The system now operates with more flexibility, intentionally bidding higher or lower than the target in individual auctions. The objective is to leverage the AI’s real-time data processing to identify and win higher-value traffic, with the expectation that the aggregate cost will balance out to the advertiser’s set target. This approach mirrors the sophisticated automated bidding strategies employed by industry incumbents such as Google Ads and Meta, which have spent the last decade shifting users away from manual bidding to maximize total conversion volume.

Critical Chronology of the Transition

OpenAI has outlined a deliberate timeline to ensure that advertisers and API partners have sufficient lead time to adapt their workflows. The phase-in process is structured as follows:

• November 9, 2024: The new Target CPA and Target CPC options become available within the ChatGPT Ads Manager for all new conversion and click-based campaigns.
• November 18, 2024: A significant backend change occurs for API partners. Any legacy Max Bid or Max CPC values transmitted through the API will be automatically interpreted by the system as an average target rather than a hard limit.
• December 31, 2026: The final sunset date for all legacy fields. After this point, the platform will no longer support manual maximum bid inputs.
• January 1, 2027: The formal deprecation of Max Bid and Max CPC for all remaining active campaigns.

This extended timeline is designed to mitigate the operational shock for enterprise-level advertisers who rely heavily on automated bidding scripts and third-party management tools. By providing over two years of lead time, OpenAI is allowing for a gradual migration of historical data and campaign structures.

Implications for Campaign Performance and Delivery

One of the primary concerns for digital marketers regarding this transition is the risk of under-delivery or cost inflation. OpenAI’s documentation explicitly warns that setting a target below recent historical performance may result in a significant drop in ad delivery. If an advertiser sets a Target CPA of $20 when their actual historical cost has been $40, the system may struggle to find sufficient inventory that meets the lower criteria, effectively silencing the campaign.

The shift also complicates the billing process for some agencies. The documentation clarifies that the bidding goal and the billing event type remain distinct settings. This means a campaign can be optimized toward a Target CPA while the actual financial transaction remains tied to clicks or impressions. This decoupling requires advertisers to be highly vigilant about their performance metrics, as the "Target" is an objective for the algorithm, not a guarantee of spend efficiency.

Furthermore, the disappearance of the "relevance-weighted, second-price auction" terminology from official support pages suggests a tightening of the platform’s pricing mechanics. By aligning its pricing rules with the broader "final winning auction bid" standard, OpenAI is moving toward a more transparent, yet potentially more volatile, bidding environment.

Broader Industry Context and Competitor Alignment

OpenAI’s move is not an isolated event; it follows a well-established industry trend. In August 2024, Microsoft Ads announced the removal of Max CPC for new non-portfolio campaigns, citing the need to reduce interference with automated bidding systems. The rationale provided by Microsoft—that manual caps prevent machine learning models from achieving peak performance—is identical to the justification now provided by OpenAI.

Industry analysts observe that as AI-driven ad platforms mature, they inevitably favor automation. Manual bidding, while intuitive for humans, often lacks the millisecond-level reaction time required to capitalize on high-conversion opportunities. By removing the ability for advertisers to "micromanage" bids, platforms can better utilize their own internal signals—such as user intent, historical search data, and engagement propensity—to allocate budget more effectively.

Strategic Adjustments for Advertisers

For those managing campaigns on the ChatGPT Ads platform, the transition necessitates a shift in strategic focus. Advertisers can no longer rely on setting a "ceiling" to protect their margins. Instead, they must:

1. Baseline Current Performance: Before switching to Target CPA or CPC, marketers should analyze their last 30 to 90 days of data to establish a realistic average. Setting a target that is too aggressive will inevitably choke campaign growth.
2. Monitor Automated Learning: After moving to the new bidding models, the "learning phase" will be critical. Advertisers should expect some volatility in the first two weeks as the algorithm recalibrates its bidding logic to match the new objective.
3. Audit API Integrations: For agencies and developers using API partners, immediate attention is required for the November 18 update. If an automated script continues to push a legacy "Max" value, the system will interpret that number as an "Average," potentially leading to unintended budget depletion if the legacy number was set higher than the intended average.
4. Shift toward Portfolio Management: With manual controls vanishing, grouping campaigns with similar objectives into portfolios may help maintain stability. This allows the AI to balance high-performing and low-performing ad groups to maintain a singular, stable average across a broader set of data.

Evaluating the Long-Term Impact

While the loss of granular control is often met with resistance by veteran media buyers, the transition to Target-based bidding generally correlates with increased scale. By allowing the AI to bid more for high-intent users and less for low-intent users, the platform can theoretically deliver a higher volume of conversions for the same total budget.

However, the lack of transparency in "black box" algorithms remains a persistent point of friction. As ChatGPT becomes an increasingly central node in digital information retrieval, the cost of advertising on the platform will likely rise. The removal of Max Bids gives OpenAI greater control over the auction clearing price, which, in a competitive market, typically favors the platform’s revenue growth over the individual advertiser’s cost-savings.

Ultimately, this update is a maturation signal. It indicates that the ChatGPT Ads platform has reached a scale where its internal machine learning models are confident enough to handle bid optimization without human intervention. For advertisers, the era of "set it and forget it" manual bidding is ending, replaced by an era of "define the goal and monitor the outcome." The success of future campaigns will depend less on the ability to write effective bids and more on the quality of conversion data provided back to the system to inform its decision-making. As the January 2027 deadline approaches, the distinction between those who successfully pivot to automated strategies and those who struggle with the loss of control will likely define the leaders in this emerging advertising channel.