ChatGPT Ads Manager Update: The Performance Marketer’s Playbook 2026

The ChatGPT Ads Manager just got the controls you have been waiting for. On May 22, 2026, OpenAI confirmed three operational upgrades inside the ChatGPT Ads Manager beta, plus a quiet test of dynamic CTA buttons inside ChatGPT itself. The update was reported by Search Engine Land and detailed in a product email from the OpenAI Product Team to advertisers in the beta.

Three weeks ago the ChatGPT Ads Manager beta was a self-serve test with lifetime budgets and country-level targeting. Today it has daily budgets, state and DMA and ZIP code geo, and aggregate totals in reporting. The platform is starting to look like a real performance channel. Here is what changed and the six moves you should make in the ChatGPT Ads Manager this week.

What the ChatGPT Ads Manager update actually shipped

OpenAI opened the ChatGPT Ads Manager beta to all U.S. businesses on May 5, 2026, with no minimum spend and CPC bidding. The May 22 ChatGPT Ads Manager update layers three operational features on top of that base.

The first is daily budgets. When you create a new campaign in ChatGPT Ads Manager, you can now choose between a daily budget and a lifetime budget. A daily budget caps spend per calendar day. A lifetime budget spreads spend across the campaign window and gives the system more freedom to pace delivery. There is one limit worth noting. Daily budgets apply only to new campaigns at this stage. Campaigns already running cannot be switched retroactively.

The second is U.S. geo targeting at three levels. You can target by state, by designated market area (DMA), or by ZIP code. DMA is the same Nielsen-defined market unit buyers use in TV, streaming, and programmatic. ZIP code targeting lets you run hyper-local campaigns for a single store or a regional event. State-level is the coarsest, useful for national advertisers with state-specific messaging.

The third is aggregate totals in reporting. Table views inside the ChatGPT Ads Manager now display impressions, clicks, and spend totals at the campaign, ad group, and ad level. You no longer have to export data and sum rows in a spreadsheet to see portfolio performance.

The new ChatGPT ads experience: dynamic CTAs

Alongside the ChatGPT Ads Manager update, OpenAI started an early test of dynamic CTA buttons inside ChatGPT. A small subset of running ads may now display one of four CTAs: Shop Now, Book Now, Sign Up, or Learn More. The CTA is selected automatically based on the ad creative and the destination experience. OpenAI said advertiser controls for CTA selection could come later.

This is the first time ChatGPT ads have shown a clear action verb. Before this test, the ad experience was closer to a contextual recommendation than a direct response unit. Adding a verb changes the metric you should care about. Click rates can climb, but only if the CTA matches the destination. A Shop Now button pointing to a long blog post will train the model on weak engagement.

You cannot pick the CTA today, but you can influence it. Tighten your destination URL to the action the CTA implies. If the auto-selected CTA is Book Now, your landing page should open with a booking form, not a hero image and three scroll sections. The cleaner the match, the better the ChatGPT Ads Manager system trains.

Why the ChatGPT Ads Manager beta matters now

For most performance marketers, the ChatGPT Ads Manager has been a parking-lot test budget. You wanted to learn the interface, not commit real spend. The May 22 update changes that math. Daily budgets and DMA geo are the two features that made early Google Ads and Meta Ads usable for regional buyers. With both in place, the ChatGPT Ads Manager beta now supports campaign architectures that map directly to existing media plans.

Daily budgets let you run always-on campaigns without watching the lifetime cap burn through in three days. DMA targeting lets you A/B test creative in matched markets, the same way you do on linear TV or connected TV. The combination unlocks the read-and-iterate loop that performance teams already run elsewhere.

This is also the moment to test before competition stiffens. The ChatGPT Ads Manager beta is open to all U.S. businesses. As more advertisers join, CPCs will climb. The buyers who learn the platform in May and June will sit on a 90-day head start by Q4. You can read our broader take on AI search visibility to see how the same dynamic played out in organic AI search.

Six moves to make in the ChatGPT Ads Manager this week

If you are already in the ChatGPT Ads Manager beta, your next 30 days should look like this.

1. Move new campaigns to daily budgets. Lifetime budgets were a workaround. Daily budgets give you the pacing discipline you need to run learning campaigns. Set the ChatGPT Ads Manager daily budget at a level that lets the campaign exit learning, usually 30 to 50 clicks per day at the platform’s current CPC range.

2. Build DMA-matched test markets. Pick two DMAs with similar audience profiles and split your creative test across them. The same approach Reddit and Snap buyers use for incremental reads works here. Our piece on last-click attribution and AI Mode explains why matched-market reads beat last-click in the AI ad era.

3. Run a ZIP-code retail pilot. If you have a brick-and-mortar location, set up a single ZIP code campaign around it. Compare CPC and conversion rate against your existing Google Local and Meta radius campaigns. Local intent inside ChatGPT looks different from a Google query, and the early read will tell you if the channel earns a permanent line item.

4. Audit your landing pages for CTA fit. The dynamic CTA test means your landing page first impression carries more weight. Open every page your ChatGPT Ads Manager creative points to. If the page does not deliver the action the auto-selected CTA implies, fix the page or change the destination URL.

5. Use aggregate totals to set a weekly review cadence. Aggregate totals at the campaign and ad-group level remove the spreadsheet step. Schedule a Friday 15-minute review where you scan totals and flag any ad group whose clicks-to-spend ratio is more than 25% off the account average.

6. Document your ChatGPT Ads Manager learnings. The platform is iterating week to week. Capture every change you notice, every CTA mismatch you find, every DMA test result. By the time outcome-based bidding ships (Digiday reported it as in development), you will have a benchmark library competitors will not.

The pattern across AI ad platforms

The ChatGPT Ads Manager beta is following a path you have already seen. Meta launched Advantage+ with limited controls, then added pacing, geo, and creative variant tools over 18 months. Reddit Ads launched with no daily budgets and country-level geo, then added DMA targeting last year. Snap shipped Unified Attribution in beta last week with similar early gaps. Each platform reached usable scale only after pacing and geo controls landed.

The ChatGPT Ads Manager is at the same inflection point. Beta access is open, CPC bidding is live, daily budgets and DMA geo are in, and outcome-based bidding is on the roadmap. The next 90 days are the cheapest the platform will ever be for early buyers. Our Meta Ads MCP operator guide covers the same playbook for the Meta side of the trend.

What the ChatGPT Ads Manager beta still does not have

Three gaps remain in the ChatGPT Ads Manager beta. First, audience targeting beyond geo and context hints is in gated rollout, with retargeting and lookalikes still to come. Second, outcome-based optimization (the feature that would let you bid against CPA rather than impressions and clicks) is in development without a public timeline, per Digiday’s May 21 reporting. Third, international expansion is staged. OpenAI has confirmed UK, Mexico, Japan, Brazil, and South Korea are next, but no date has been set.

These gaps tell you where to hold spend. Do not pour brand-defense budget into the ChatGPT Ads Manager until lookalikes are GA. Do not run pure performance campaigns expecting CPA bidding until outcome-based optimization ships. The features that landed on May 22 are the foundation, not the roof.

Our Google Ads budget pacing guide walks through the pacing audit you should run anywhere you spend. If you want to compare ChatGPT Ads against the established platforms, our Google Ads vs Meta deep dive is the right starting frame.

Final read on the ChatGPT Ads Manager update

The ChatGPT Ads Manager update is the first version of the platform that supports real performance discipline. Daily budgets, DMA geo, and aggregate totals are the trio that turns a beta into a buy-able channel. Dynamic CTAs hint at where engagement metrics are going next. The six moves above turn the ChatGPT Ads Manager update into a 30-day plan you can run inside any account that has beta access.

The window to learn the ChatGPT Ads Manager beta before competition tightens is short. If you are already in, run the audit this week. If you are not in yet, sign up. Read our generative AI features guide for context on how Google, Meta, and OpenAI are converging on the same playbook. The AI search terms report breakdown is the next stop, and our Performance Max alternatives guide is the long-read on paid-media architecture.

If you want a hand mapping the ChatGPT Ads Manager beta into your existing media plan, book a free consultation and we will walk through your account.

Source primary: Search Engine Land. OpenAI expands Ads Manager Beta with new budgeting and geo targeting controls. Source secondary: PPC Land. ChatGPT Ads Manager gets daily budgets, geo-targeting, and dynamic CTAs.

Let’s Grow!