Google has announced new testing and planning tools for Search campaigns using AI Max. From September, advertisers will be able to compare budget and target-ROI changes across more than one Search campaign in a single A/B test. AI Max experiments will also gain brand and location controls, while Performance Planner will make the projected impact of bid and budget changes more visible. The announcement matters because it moves the conversation away from switching automation on and waiting for a result, toward connecting a change to an account-level controlled experiment.
The real value of this announcement is that it does not reduce the question to “is AI Max better?” A budget move, a target-ROAS change, query composition, creative variation and seasonality can all blur together when several campaigns are moving. The new experiment surface may make the principal variable easier to isolate. It will not design a sound control group, success metric or stop rule for the advertiser; those responsibilities remain with the team.
What did Google announce?
Google’s official announcement highlights three elements. First, multi-campaign Search tests will be able to examine budget and ROI-target changes together. Second, AI Max experiments will include brand and location controls. Third, Performance Planner will provide a planning view of likely effects from bid and budget decisions, with some recommendations moving closer to one-click application. Google places the first capability in a September rollout, so an option that is not yet visible in an account should not be treated as an existing setting.
Brand and location controls deserve particular attention. AI Max can expand the intent space available to a Search campaign. That flexibility can also make a result look stronger if branded queries or a particularly responsive geography skew the comparison. Controls are intended to keep the experiment legible for that reason. Confirm which controls are available to the specific account and market when access arrives.
Who could benefit?
Advertisers with sufficient Search conversion volume may get a more realistic portfolio view from multi-campaign experiments than from a tiny single-campaign test. If campaigns serving the same product group, service line or location compete for budget, changing one in isolation can hide the total effect. For agencies, it offers a more disciplined alternative to an informal before-and-after comparison in a client account.
For low-budget accounts or accounts with irregular conversion volume, however, the new tool is not automatically an advantage. A test cell without enough signal can measure chance more than a setting change. Launching new creative, a promotion, a bid strategy and an AI Max adjustment at the same time makes the outcome impossible to interpret. Start with one hypothesis before expanding the scope.
Confirmed details, unknowns and risks
It is confirmed that Google is preparing a broader test and planning layer for AI Max, with the multi-campaign A/B testing window identified as September. Brand and location controls are specifically named. Performance Planner will provide a forecast-oriented surface for budget and bid decisions. Together, these elements can make target-ROI and budget movements more systematic to observe.
What remains uncertain is how closely a forecast will match business reality. A planner is directionally informed by historic signals; it does not independently manage stock, margin, sales-team capacity, returns or demand shocks outside the campaign. Even a statistically visible test result is not automatically a durable profit result. Platform conversions need to be completed with CRM, revenue or sales data.
What should brands in Türkiye do now?
First, group candidate campaigns by comparable business purpose. Putting brand campaigns, broad demand generation and products with very different margins into one test pool will mislead the comparison. Write down each group’s conversion definition, budget flexibility and acceptable cost threshold. A performance marketing team should align that grouping with sales operations, not only with the media plan.
Second, save a baseline. Export the previous four to eight weeks of spend, conversions, qualified leads, net revenue, search terms and landing-page data. Record brand-query, geography and device splits separately. This reference is what makes it possible to identify where a positive or negative change came from later.
Third, limit the principal variable. State clearly whether the test changes budget, target ROAS or an AI Max setting. If creative or landing-page work is needed, schedule it in a different period. Starting an automation change in the same week as conversion-focused web revisions makes attribution much harder.
Fourth, add a quality-control layer. Ecommerce teams should pair net revenue with return rate and new-customer share. Lead-generation teams should read form volume together with qualified leads, time to sale and cancellation rate. Make the stop rule explicit before launch: rising irrelevant query intent, a broken landing-page tag or sales quality below the control group should trigger review rather than automatic scale.
Fifth, allow the decision meeting to wait a week. An early jump in clicks or impressions is not a reason to expand spend. Scaling before the conversion window, sales cycle and return period have matured can turn a short-term signal into a permanent choice. A slower but interpretable test builds an account that learns faster over time.
Fark Studio view: automation earns its value through a better experiment
AI Max testing is a useful step toward governable account automation. From Fark Studio’s perspective, the right question is not “which setting can we enable most aggressively?” It is “what can change while preserving the business outcome we care about?” The frame also touches SEO and landing-page quality: the demand found by advertising, the promise made by a page and the value recorded by measurement must support one another.
When the new tool is available in your account, begin with a small documented test. If the budget cap, success threshold, quality metric and rollback rule are written before launch, Google’s new surface does more than create data. It makes the next budget decision more defensible.
Sources
Search Engine Land — Google adds new AI Max testing and planning tools, August 20, 2026
Social Media Today — Google Adds Simplified A/B Testing for Search Ads, August 20, 2026
