Google Ads can now apply a forecast to live campaigns in one click

Google Ads has shortened the path between a Performance Planner forecast and a live campaign change. In an August 20 announcement, Google said advertisers can review suggested bidding or budget changes and apply them directly to existing campaigns in one click. It also announced broader AI Max experiments for testing budgets, ROI targets, and campaign controls.

For a marketing team, the useful speed comes with a sharper operating question: when is a forecast ready to change live spend? This article separates what the planner predicts, what an experiment observes, what a reviewer approves, and what an undo action can actually recover.

What does the new one-click path change?

Performance Planner already lets advertisers model how campaign-setting changes may affect key metrics and overall performance. Google’s Performance Planner help page now documents an implementation step: after generating a forecast, a user can select Apply suggested changes, review budget or bid adjustments by campaign, deselect campaigns, and confirm. Google says confirmation updates the selected live campaigns immediately.

That removes a manual transfer. A user no longer has to read a recommendation, open each campaign, reproduce the setting, and risk a transcription error. It also compresses the time available to notice that a forecast has the wrong campaign scope, conversion goal, time horizon, or business constraint.

The button does not turn the forecast into observed evidence. Google says Performance Planner simulates relevant ad auctions from the prior 7–10 days, accounts for factors including seasonality and competitor activity, refreshes forecasts daily, and uses Google AI to fine-tune them. Those are properties of a forecast, not a guarantee that the suggested settings will produce the forecasted outcome in a particular account.

Why is an AI Max experiment different from a planner forecast?

An AI Max experiment creates a controlled comparison inside an existing Search campaign. Google’s AI Max experiment documentation says a percentage of traffic and budget stays in a control arm with AI Max off while the remainder enters a treatment arm with AI Max on. The test observes performance under both conditions rather than simulating a future result from recent auctions.

Google’s August 20 announcement adds two relevant capabilities. AI Max experiments can now run with brand or location controls enabled. Google also says a multi-campaign A/B test for different budgets and ROI targets will begin rolling out in September 2026. The announcement does not give a date when every account will receive that feature.

The two surfaces therefore answer different questions. Performance Planner asks what Google’s model forecasts under proposed settings. An experiment asks what happened to traffic assigned to control and treatment during the test. A team may use both, but it should not describe a planner recommendation as “tested” unless an actual experiment covered the same setting, scope, objective, and relevant operating period.

What should a reviewer inspect before applying changes?

Start with the objective used by the plan. Google says a forecast can be created for a specific conversion goal, and eligible conversion actions must be primary and attached to forecastable campaigns. A recommendation optimized against the wrong primary conversion can be internally consistent while still serving the wrong business outcome.

Then inspect the proposed scope at campaign level. The implementation flow allows campaigns to be deselected before confirmation. Use that control to keep an approval narrow: identify each campaign, current bid strategy or budget, proposed value, expected metric movement, and the reason it belongs in this change set. Do not treat “eligible for the planner” as equivalent to “approved by the business.”

Finally, record the forecast context before it changes. Preserve the time the forecast was generated, its horizon, selected conversion goal, included campaigns, proposed settings, and displayed expected outcomes. Because forecasts refresh daily, a later view may not be the same evidence a reviewer saw when approving the change.

BaristaLabs recommendation: the person approving material spend should not be the automation that generated or surfaced the recommendation. For low-risk accounts, that separation can be a second person reviewing the saved plan. For larger portfolios, it can be a role-based workflow with campaign and dollar thresholds. The public Google sources describe the product controls; they do not prescribe an advertiser’s approval policy.

What does rollback mean after the click?

Google says users can monitor and undo applied planner changes in the Google Ads Bulk actions interface. That is an important recovery path for settings. Before relying on it, confirm that the intended operator can find the change record and restore the prior values for every campaign in the approved set.

Undo is not rewind. Reversing a budget or bid setting does not recover spend already incurred, recreate auctions that already occurred, or guarantee that subsequent performance returns immediately to its prior pattern. Google’s cited pages do not make those promises. The safe rollback record therefore needs both the previous settings and a clear observation window for deciding whether to restore them.

AI Max experiments have a separate reversion behavior. Google says that when an experiment ends without being applied to the base campaign, AI Max settings return to their original pre-experiment state, with documented handling for brand inclusions and exclusions. Do not assume that this experiment behavior governs a Performance Planner change; they are different workflows.

How can a team use the faster workflow safely?

Begin with one plan that contains a small, intelligible set of campaigns. Before anyone applies it, save the forecast context and have a named reviewer approve the exact campaign-level changes. Record prior values, the application time, the observation window, the metric that can trigger rollback, and the person authorized to undo the settings.

After application, verify the live settings rather than treating the confirmation message as evidence that every intended value landed correctly. During the observation window, compare actual results with the recorded business metric and note external changes that could complicate interpretation. If the rollback condition is met, restore the settings through the documented change history and verify the resulting live state.

For the September multi-campaign A/B testing rollout, first confirm that the capability exists in the intended account and that its brand or location controls match the campaign’s real guardrails. Use the resulting experiment data as experiment evidence. Keep it separate from the Performance Planner forecast, even when both point toward the same budget or ROI target.

Faster application needs a stronger decision boundary

The release makes Google Ads planning more operational: a modeled recommendation can move into production without manual re-entry. That can reduce routine work and setup mistakes. It also makes campaign scope, conversion-goal choice, approval, and rollback evidence more important because the final step is easier.

Use the button after the proposed values, campaign set, forecast context, and recovery path have a named owner. If that handoff is still informal, BaristaLabs can help turn it into a repeatable review through process automation and integration, or you can bring one campaign change path to a focused review.

Sources

• Google Ads, “Make AI Max work for your business with new testing and planning tools”, published August 20, 2026.
• Google Ads Help, “About AI Max experiments”, accessed August 20, 2026.
• Google Ads Help, “About Performance Planner”, accessed August 20, 2026.