Pmax: How do top-performing campaigns work?

It's clear that algorithms want freedom. Google's top-performing campaigns are proof of this: they demand a variety of titles, descriptions, images, resources, and the ability to display them in multiple locations. It seems like the algorithm's motto: "Give me all the content, I'll take care of the rest." And the truth is, in many cases the results are outstanding. Are they really good, or is there a catch? How far does our control go? We'll explain how they work.

Powerful resources, diverse content

Google's top-performing campaigns leverage the power of AI to optimize your ads across all Google channels (Search, Display, YouTube, Discover, Gmail, and Maps). But unleashing the system requires multiple creative angles in addition to high-level resources.

Therefore, the recommendation is to use different creative lines in the images, different text and descriptions, and even different points of view. This may seem contradictory; the idea is to encompass tastes and affinities. This way, Google's AI will be able to find and segment based on the creative asset. The algorithm will then reproduce the pattern.

How do we analyze a Pmax?

These campaigns are so automated that Google offers little or no statistics about them. It's impossible to know, for example, which locations perform best. This poses a problem for analysis, as it's like going "blindfold."

To solve this problem, scripts were created, which allow you to break down and view all the information about the locations through scripts. All of this is done in Google spreadsheets, which are updated regularly and provide data on performance.

Only in this way can budget distribution and campaign performance be properly analyzed. Especially because performance by placement is never equal: sometimes display placement takes up a lot of budget, for example, while other placements are left behind. In short, without a script, no conclusions can be drawn.

What's up with the brand?

The main criticism of Pmax is undoubtedly the volume of conversions attributed to the brand. Can the brand be negatively attributed? Yes. However, for most brands, which are not yet fully established, this is not the best approach. This means that a large portion of the conversions generated by the campaign will be through the brand. If the brand's weight exceeds 20% of conversions, it's worth negatively assessing it and implementing an SEM campaign solely for branding.

For ecommerce, standard or Pmax shopping?

The shopping campaign has been the ecommerce sales campaign by nature. Thanks to Pmax scripts, it has been proven that this location does not typically receive the largest percentage of the budget.

As a trick, you can create a resource group with only the product feed, without images or titles. This way, the system will be forced to display only the feed, turning the PMAX campaign into a shopping campaign.

In short, Google's top-performing campaigns are an excellent option for any e-commerce site. With these recommendations, you can test resources and scale ROAS.