Performance Max asset groups gain a "Save changes as an A/B test" toggle

What changed

The panel says the experiment can measure changes to headlines, descriptions, images and videos. It also warns that the asset group cannot be edited while the experiment is running. Those details come directly from Bidner’s screenshot, rather than an inferred rollout announcement.

Google’s current asset-best-practice documentation corroborates the workflow: it describes creating an A/B experiment through Save and test asset group changes in Edit assets. The help page’s revision date and the feature’s first release date are not established. Search Engine Roundtable covered Bidner’s observation on 21 October, two days after the date displayed on the original post.

Preparing a meaningful creative comparison

Define the question before enabling the test. Comparing a price-led message with a convenience-led message requires a coherent alternative set of assets, not just a headline swap surrounded by contradictory imagery. Record which assets belong to each approach so the outcome can be interpreted against the intended change.

The editing restriction also matters operationally. Check scheduled offers, expiry dates and any creative that may need an urgent correction during the proposed test period. A test should not start with material the team already expects to replace immediately.

Keep the success measure and review period agreed in advance. Looking for a winner after each short fluctuation can turn an experiment into repeated reactive edits. The evidence here supports an available testing workflow and an observed interface; it does not establish a guaranteed performance improvement or availability in every account.

First contributor credit goes to Joey Bidner for the account observation identified in the reviewed sources. It is not a claim that he discovered asset-group experimentation before anyone else.