ChatGPT Ads Maximize Results: how it works, settings and how it compares to manual bidding

What Maximize Results actually does

Maximize Results is a bidding strategy, not a campaign objective: it layers on top of the CPM, CPC or oCPC choice already made at campaign level (details on our ChatGPT Ads campaign setup page). Once turned on for an ad group, Ads Manager sets and adjusts the bid on your behalf, within the budget you set, to get as many results as possible for the chosen objective: as many clicks as possible for a Clicks campaign, as many conversions as possible for a conversion-oriented campaign.

That is the reverse of manual bidding, where you enter a maximum amount per click or per thousand impressions yourself, and that ceiling, combined with the competition at the time, determines what you actually pay.

Selected by default: what that changes in practice

The most operational point of this update: for new eligible ad groups, Maximize Results is the starting setting, not an option you have to turn on. An account that creates a new ad group without reviewing the "Bid strategy" section ends up steered by automatic bidding, without having explicitly decided so.

That is not a problem in itself, but it changes how you read an existing account: if a campaign is spending differently from what you expected, the first thing to check is no longer just the bid you entered, but which bidding strategy is active. The setting lives in, and is changed from, the "Bid strategy" section of the ad group; manual bidding remains available at any time.

What it does not guarantee

OpenAI is explicit about this, and it is the nuance most often lost along the way: Maximize Results currently guarantees no target CPA, CPC or ROAS. It optimizes for the volume of results within the budget, not for a precise cost per result.

In practice, two accounts with the same budget can get a different number of clicks, at a different cost per click, without either result being abnormal: the algorithm looks for the best total volume, not a constant unit cost. If your business priority is to hold a precise cost per acquisition, that lack of a guarantee is the most common reason to switch back to manual bidding, where you, not the algorithm, set the ceiling.

Maximize Results and oCPC: two different mechanisms, often confused

Both target conversions, but they sit at different levels. oCPC is a campaign objective: it optimizes delivery toward a specific conversion event you designate, while still billing per click (full details on our ChatGPT Ads pricing page). Maximize Results is a bidding strategy: it determines how the bid itself is set, whatever campaign objective sits above it.

So it is possible to combine an oCPC objective with a Maximize Results bidding strategy: one chooses what to optimize toward, the other chooses how to bid to get there. Our comparison table lays out all four configurations on a single page: CPM vs CPC vs oCPC vs Maximize Results.

When to keep Maximize Results, when to switch back to manual

A simple decision grid, based only on what OpenAI documents:

Either way, the platform's general rule applies: OpenAI recommends waiting at least 24 hours after launch before concluding there is a delivery problem, and observing enough volume before changing strategy.

What is not publicly documented

Points we flag as unconfirmed for lack of an official source:

• The precise algorithmic mechanism behind the bid adjustment (frequency, signals used) is not described.
• The exact eligibility criteria that determine which ad groups get Maximize Results by default are not published in detail.
• No comparative figures between Maximize Results and manual bidding (volume, average cost) are published by OpenAI: any comparative number you come across is from third-party observation, not from the platform itself.