The 'Other' Bucket: What is Google Ads Hiding?

The ‘Other’ Bucket:
What is Google Ads Hiding?

"As much as 75% of [my client's] search ad spend is going to “Other” search terms that they can't even see."

The Mystery of Google Ads’ “Other Searches”

Every other week, our Google ads correspondent Jyll Saskin Gales walks us through the latest platform changes. Jyll spent six years at Google in a senior ad role, and today runs the Inside Google Ads training program.¹

• Google's Marketing Live event claims more transparency and control for advertisers, but actual control has decreased.

• Many search terms are hidden — ostensibly for privacy reasons, impacting visibility and optimization of ad spend.

• Ethical concerns arise over Google's practices, likened to changes in Google Analytics and Gmail advertising.

• Effective AI targeting relies heavily on accurate and complete data, which is often not available for small business owners.

It's been a couple of weeks since Google's big Marketing Live event. And one of the things that they talk up at these events is just how much transparency and control they're giving advertisers. And then out the back door, they're removing controls, putting more things into the black box. Is that your sense too?

I do share the same sense with you.

Google announced some more transparency and control features coming to Performance Max. For example, more asset group reporting, more opportunity to add exclusions.

But on the whole, if you're a Google Ads practitioner in 2024, you have way less transparency and control over your campaigns than you even did in, let's say, 2020.

Can you give me a couple of examples of that? How have we been losing control in Google Ads over the last few years?

I think one of the big ways is actually the “Other” search terms. So when you run a search campaign, you enter your keywords into Google Ads, and then you check your search terms report to see which user searches actually match to your keywords. And usually, you would be able to see those search terms and add negatives and optimize based on it.

But now, due to privacy, according to Google, a lot of those search terms you're advertising on, you can't actually see. And with some of my coaching clients, I can see that as much as 50% or even 75% of their search ad spend is going to “Other” search terms that they can't even see. So they don't even know what they're advertising on.

Is this similar to when Google Analytics, back in the day, you'd be able to see every keyword, everything that came in? And then one day, they just basically deleted all of them and put them, what was the keyword they used, like [not provided] or something like that? Is this kind of the same thing on the ad side?

It's a similar kind of idea. I'm not familiar with the specifics of the Google Analytics example. But in the last five or so years, Google has become much more conscious of user privacy, thinking about privacy thresholds.

Another example of this is back in 2017, you could advertise to people based on words in their Gmail inbox. So you could keyword-based advertise based on the emails people had. You cannot do that anymore.

{laughs} Was it effective though? Privacy issues aside.

Highly effective. The clients I worked with at Google who used this, you could advertise to people who you know were getting emails from your competitors. There were lots of really creative ways to use that when Gmails were their own separate campaign type. Now, Gmail's been brought under the umbrella of Demand Gen and Performance Max, so you can use audience targeting. But it's not quite the same as targeting ads based on what people have in their inbox.

Is it unfair for me to call this a scam? Google has you paying to advertise on searches that you've picked, but you can't actually see those searches?

Yeah, there are definitely folks in the industry who, especially after all the news Google has had recently about the way they may have misled SEOs or the DOJ case about bid floors and things.

So some people think it's just a way to get you to advertise on garbage and ignore your negative keywords. We have no way to prove or disprove that.

What I can say as someone who worked there is that's not the Google I know. And I think on the whole, Google does intend to do right by its advertisers.

Is it because of privacy or some other reason? I don't know.

But if you haven't checked your search terms report in a while, or maybe you have, but you've never scrolled down to the bottom to see that other search terms line, you're going to want to check.

I was on a coaching call yesterday and my client, actually their other search terms performed way better than their search terms. So not so concerned there. But then a client call I had last week, their other search terms performed way worse than the search terms we could see. So it's definitely worth looking into. And ultimately, you have to rely on something like your bid strategy to guide Google, whether it's search terms you can or can't see, towards the outcomes you want to achieve.

I wonder if this is in the long run, a good thing, because one of the debates in the digital marketing space is that when AI first entered the media buying space, it was terrible. Now most people have accepted that AI, especially around targeting, it actually performs a little bit better than when the humans get our hands in micromanaging stuff.

Do we see that in this kind of big, mysterious, lumped “other” search terms?

Yes, with an asterisk.

The automation, whether it's optimized targeting or the search terms you can't see, or running a PMAX campaign, the AI and automation can do better targeting than humans if given the right information.

And that's the gap that I think Google product managers have a blind spot around because in many cases with your small business owners or your small budget accounts, they don't have the correct data feeding in or complete data feeding in.

If you're a lead gen advertiser, you may not have the full funnel leading in. You may not have conversion tracking set up. You may not be sharing first-party audiences. So when all of those foundations of the account are in place, the AI has the data it needs to drive good results for you.

But so frequently for so many different reasons, we don't live in an ideal world. And so you don't have all those fundamentals in place. So the AI doesn't have the correct data to work off of. And as a result, it's not able to drive the best results for your business.

I don't think that's Google's fault. AI does exactly what we tell it to do. But at the same time, it's not the business owners fault either that they can't properly implement these things that are really complicated.

Be sure to check out Jyll’s Inside Google Ads training program¹

Confirmed: Google Pulls Way Back on AI Summaries

New data from BrightEdge confirms what we reported yesterday — that Google has quietly pulled way back on its AI summaries.

Those summaries sit at the top of the search results page, and were the big star at Google’s recent I/O event. But since launch, users found it giving nonsensical (and sometimes dangerous) answers, and marketers said it was all but plagiarizing their web pages with only a tiny credit link.

‘Working as intended’

Google responded by saying, essentially, ‘working as intended,’ and blaming media outlets for reporting on outlying searches that don’t get a lot of queries.

That said, clearly the attention spooked Google, with BrightEdge finding that now fewer than 15% of queries have that AI Overview on it. Compare that to when they rolled it out, that number was almost 85%.

When do the summaries show up now?

BrightEdge says there are a few things that make it more likely for AI Overviews to trigger:

• They’re almost 3x as likely to show up when Featured Snippets are also on the page

• They’re also more likely to show up when the search query is a question

As for what keeps them away:

• So far, questions about local topics or businesses are the least likely to have AI Overviews

• And Sitelinks are also less likely to trigger an AI Overview, presumably trying to keep the risk of providing inaccurate information about brands down

Industry breakdown

BrightEdge also looked at how often AI Summaries now show up based on industry:

• 63% of keywords in healthcare will trigger an AI Overview

• In B2B Tech, 32% of keywords will

• in Ecommerce, 23%

• Restaurants and travel tend to barely show any