3 Reasons Why Pmax Ads Show Up In Shoddy Places

As marketers know only too well, there are trade-offs with everything. And in the era of set-it-and-forget-it tools, that trade-off is one of spend efficiency for ads sometimes showing up in questionable environments.

Despite grumbles, a lot of marketers have accepted that trade-off. But with AI fuelled made for advertising sites springing up, the phenomenon will likely exacerbate.

Looking specifically at Google’s Performance Max campaigns, which offer broad reach and valuable placements across multiple channels, three key challenges persist: the unchecked rise of low-quality AI-generated sites, the difficulty of opting out, and the labor-intensive process of sifting through vast ad placement data.

“MFA is a complex and challenging topic—it’s also a topic that lacks consensus and definition at the moment,” said a Google spokesperson. “We continue to review these issues and work with our industry partners on potential policy adjustments.” Google took action against over 2.1 billion publisher pages and 395,000 sites in 2023—an increase from 1.5 billion pages and 143,000 sites in 2022. It takes action on sites that violate its policies but were missed by its enforcement systems, the tech giant said.

Here’s your primer on these issues:

A surge in AI-spun sites

Previously, advertisers could avoid ads being served on MFA sites by using block lists.

Now, advertisers are faced with AI-spun template sites, often falling under the MFA category. These low-quality domains prioritize ad revenue over content and can emerge overnight with no human involvement, rapidly eating into ad budgets.

These sites can be spotted by their cookie-cutter appearance: repetitive designs, AI imagery, and suspicious domain names like privatecaregiverfortheelderlyfinder.today or gametacticzone.top, and extensions like .click or .today.

“We call them template sites,” said Rachmiel Moss, co-founder and CEO, DeepSee.io. “They’re carbon copies of each other, with slight variations.”

DeepSee.io has tracked a sharp rise in these sites, with their numbers jumping from 20,000 to nearly 50,000 since the start of the year, ADWEEK previously reported.

“They can be spun up very quickly,” said Moss, “There’s no human touch to them. I don’t know how Google monetizes this.”

Game sites, often featuring ad.txt files but little legitimate content, are also becoming more common.

“A single look reveals it’s just nonsense,” Moss added, noting that thousands of these sites are created weekly and continue to qualify for ad spend, “especially through Google’s Pmax campaigns.”

Traffic going to low-quality domains is never going to be zero, according to an agency executive, speaking anonymously to preserve industry relationships.

“With gen AI it’s just easier to spin up an MFA site,” the executive said.

No easy way to opt-out

Google’s ad ecosystem lacks a dedicated exclusion tool for MFA sites, meaning advertisers must manually block individual sites as they appear, according to the first agency exec.

“There isn’t necessarily a blanket setting to keep you out of that inventory,” the executive noted.

Although advertisers can block partial domains before a campaign launches, the constant emergence of new sites makes it difficult to maintain control, according to a second industry executive.

In 2021, Google rolled out dynamic exclusion lists, enabling advertisers to block placements based on topics like politics or devices such as tablets.

Navigating data overload  

While advertisers have long had access to ad tracking data through the Google API, they needed a certain amount of technical expertise to extract it, like writing custom scripts, said the first executive.

Since March, Google has made this data directly available in its ad ecosystem, simplifying access. However, extracting meaningful insights remains a challenge

“They contain thousands of rows in an Excel spreadsheet, so making it actionable and combing through the volume is more difficult,” the executive said.