Google Ads uses separate experiment workflows for AI Max in Search and asset optimization in Performance Max. For a Search campaign, an AI Max experiment compares a control with AI Max off against a trial with it on within the existing campaign. For Performance Max, an asset-optimization experiment compares asset combinations against a base set. Choose the workflow for the campaign type, then judge the outcome against your own business metrics—not a platform-wide promise.
Choose the experiment for your campaign type
| Campaign and test | What is compared | What you need |
|---|---|---|
| Search: AI Max experiment | A control portion with AI Max off and a trial portion with AI Max on, within one existing campaign | An eligible Search campaign; restrictions apply |
| Performance Max: asset-optimization experiment | Different asset combinations against a base asset set | An existing Performance Max campaign and a base asset group |
These are not interchangeable procedures. In particular, the Search experiment’s traffic-and-budget split is not the setup for a Performance Max asset test.
As an Amazon Associate I earn from qualifying purchases.
How an AI Max Search experiment works
Google Ads Help describes an AI Max experiment as a test run. In the single-campaign flow, Google assigns a share of traffic and budget to the control, where AI Max is off, and the remainder to the trial, where it is on. After reviewing the results, you can apply the experiment setting to the campaign.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Scan for outdated or missing drivers - takes under a minute3Repair Windows errors before they cause bigger problemsYou can start the experiment from Recommendations or configure it through Experiments. Google says the Recommendations route can simplify setup, while the Experiments route offers more granular settings when the campaign meets the selection requirements. Google also says the experiment can reduce some setup errors and shorten the learning period; those are stated benefits, not guaranteed outcomes.
#1 Best Overall
Know what the treatment may include
AI Max is a bundle rather than one isolated switch. Google identifies search-term matching and asset optimization as its two main features. The experiment setup flow also includes text customization and final URL expansion by default. Review the treatment configuration so you know which changes the comparison is testing; some treatment features can be deselected.
Brand inclusions or exclusions configured during setup apply to both the control and trial arms for the experiment. Initial setup does not offer certain ad-group-level changes, including disabling search-term matching or adding URL inclusions. Google says settings can be changed later by editing the new campaign.
Rank #2
Check eligibility before planning around the flow
The newer AI Max experiment flow has restrictions. Google lists campaigns with Display Network targeting, shared budgets, portfolio bidding strategies, active experiments, and certain existing text-customization settings among the cases that can prevent use of the flow. Check Google Ads Help’s current eligibility list and interface instructions before scheduling a test; availability and requirements can change.
Free tools Windows power users keep installed
One-click scans. No signup required.
A custom experiment remains an alternative when you need distinct control and treatment campaigns: it uses a copied campaign structure, unlike the newer AI Max experiment that stays within one campaign. Select the method that fits the account’s eligibility and the comparison you need rather than assuming every campaign has the same one-click option.
How Performance Max asset tests differ
For Performance Max, Google’s developer documentation describes asset-optimization experiments as a comparison of different asset combinations against a base set. Setup requires an existing Performance Max campaign and a base asset group. The question here is how asset combinations perform relative to that baseline—not whether a Search campaign should have AI Max enabled.
Decide what success means before launch
Set the primary business outcome and efficiency measure before interpreting results. Compare conversions or conversion value alongside cost per acquisition (CPA) or return on ad spend (ROAS), as appropriate to the campaign goal. Also inspect whether the resulting search terms, generated text, and selected landing pages remain relevant to that goal.
- Write down the primary outcome metric and the efficiency measure you will use.
- Set the comparison period before reviewing results, and use a period that makes sense for the campaign’s conversion cycle.
- For Search, examine whether query matching, text customization, and landing-page selection support the intended audience and offer.
- For Performance Max, assess the tested asset combinations against the base asset set.
- Decide in advance what evidence would justify adopting the setting, continuing to observe, or keeping the control.
Google’s setup and reporting guidance does not establish a universal experiment duration, account-level sample-size rule, or statistical-significance threshold. Do not treat a short-lived fluctuation or a platform benchmark as proof that a setting will work for your account.
Put Google’s reported result in context
In an April 2026 announcement, Google reported an average of 7% more conversions or conversion value at a similar CPA or ROAS for AI Max Search campaigns using the full feature suite—search-term matching, text customization, and final URL expansion—compared with campaigns using search-term matching alone. This is Google’s company-published comparison, not an independent study or a forecast for an individual advertiser. It also compares a particular feature bundle with a particular baseline, so it should not be read as a guaranteed lift from enabling any one AI Max setting.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
