What Google’s query matching update means for future PPC campaigns
“Today, we’re announcing updates for Search ads query matching and brand controls.”
It’s been a while since one sentence from Google elicited such a positive reaction from the PPC community.
Last month’s changes to queries, match types and negative keywords were a welcome reminder that Google wants agencies and advertisers to trust it.
In this article, I’ll explain each of the changes to PPC advertisers and agencies and how they give us clues about Google’s future direction for paid keywords and targeting.
Dissecting Google’s improvements to query matching
Ginny Marvin’s tweet might be the first update in years that everyone has welcomed. Interestingly, these updates align with the idea of search themes and the future that Google Ads wants.
Brand exclusions and inclusions
We’ve been able to use brand exclusions – distinctly different from negative keywords – by selecting from a brand database and telling Google that you don’t want to show up for searches about those brands.
There is now the inverse of that, what Google is now calling “brand inclusions” for Search campaigns.
You’ll have the ability to shut off match types in your campaign and make it exclusively broad match, and then layering on brand inclusion lists to tell Google to only show your ad for broad match queries to anything, as long as it’s related to your brand.
This is a great example of Google moving away from fixed match types.
Wider coverage for negative keywords
Until now, we’ve had to add more negative keywords than ideal because they didn’t account for misspellings and close variants. Google is improving that “negative matching.”
I’m not sure why this wasn’t there before. It’s one of those things that feels like an obvious move from the beginning. Perhaps it’s because this will actually lose Google money with more keywords blocked automatically.
Whatever the reason, I’m glad to see them improving how negative keywords work.
Improved search term aggregation and reporting
Continuing the theme of streamlining match types into themes, brands are now becoming a theme that we can include or exclude. It’s the same with search terms.
Google is now aggregating more search terms, so they’re also blocking many keywords from the search term reports due to privacy.
People are upset about this, which is both understandable and debatable.
By grouping keywords into search themes, Google will show more data indirectly. Instead of having three different versions of the same keyword with misspellings, it might be rolled up into one that’s very close and has more visible data. It doesn’t change that themes in searches are set to overtake keywords bit by bit.
Perhaps the biggest piece of this is the Insights tab in Performance Max.
I’m grateful that Google is displaying more data around search terms in Performance Max, especially compared to how it started out. But it isn’t a pure search term report. It’s search themes grouped into categories.
And that is where it’s all headed.
My prediction: In the next 24 months, there won’t be match types
Google Ads is moving to a point where you’ll set up a Search campaign, add search themes, and then tell the system which brands to include and exclude.
Right now, only Performance Max has brand exclusions, but Google has announced plans to roll it out to Search.
Let’s say you have a generic Search campaign and don’t want to show up for competitor names. Instead of adding negative keywords, you’ll just exclude competitor brands.
I like this because it covers a lot of misspellings, close variants and things of that nature.
While it’s a good extra level, my hope is that they don’t take away negative keywords.
Search themes: The future of keywords
Google as a platform is evolving rapidly and significantly. It’s unrecognizable from where it was two years ago; two years from now, it won’t be anything close to where it is now.
You can see that in the redesign of the Google Ads UI – certain things are being pushed away and hidden from view under certain menus, while other things are being pulled to the front.
Search themes weren’t around when Performance Max first released; they were just like any other audience signal, requiring you to create a custom intent keyword list.
It appears that Google quickly realized that because Performance Max pulls in the dynamic search component of DSAs, they can incorporate it as a more standalone feature.
When you’re building out a Performance Max campaign today – and this is much more important for non-shopping Performance Max – the absence of a product feed makes audience signals critical to reaching the right audiences.
When there is a feed, Google uses the product data to find searches in the Shopping inventory and determine who your customer is. Without one, you use audience signals and search themes to indicate to the system who you believe will buy from you.
Outside of all the best practices, search themes tell the system what kind of traffic it’s best to show up for (i.e., specific search terms, themes and anybody who searches for “x” or “y”).
These don’t necessarily have to be immediately related to what you’re selling. It could be terms people search for that indicate they would also buy your product.
If you look at the evolution of the different features around keywords in the past couple of years, custom intent largely became search themes. They’re more prominent and in asset group creation instead of one step down in audience signals.
Tread with caution: How broad is too broad?
Much of the resistance to Google’s changes aren’t so much about technology, but agencies and brands feeling like they have less transparency and reduced control.
A lot of the noise around reverting the system doesn’t account for how search itself has changed. Demands like wanting keywords to be one-for-one exact matches typically come from a small portion of Google Ads professionals.
However, there’s also some validity to the criticism.
There are (and always will be) certain industries where you need that specificity and regulation. Some of the biggest ones are pharmaceutical and legal, especially on-the-job injury and automobile accidents. In these domains, the system can be way too liberal.
It doesn’t hurt as much for the larger advertisers spending million-dollar budgets. But if you’re running campaigns for a local lawyer with a $10,000 budget, these broad matches don’t make sense.
Once broad match learns what’s what with your account, it becomes quite good at pulling in new opportunities. It will cost you time and money to let it learn what works, but what happens when the volume isn’t there?
In that situation, there are only three levers you can pull to make a search campaign show up in more auctions:
- Widen your target location. If you can’t service beyond a specific area, you can’t target outside of it.
- Increase your bids or reduce bid targets. This is expensive and not accessible to everyone. Plus, at a certain point, you’ll have increased your bids enough to capture up to 90% of what’s available. There’s not much more you can do.
- Add more keywords or match types. This will certainly expand your reach, though not always with relevance.
Combine a relatively common search (like “car accident lawyer”) with a low-volume location (like a small town with 10,000 people), and it’s tough to see broad match ever getting to a point where it works in that type of situation.
I hope Google has a plan to treat that scenario differently from people across the U.S. looking for sneakers. Otherwise, I can see the platform becoming prohibitively expensive for several brands and businesses.
A use case in balance and relevance
Historically, Google has looked at what’s best across all advertisers – that’s millions and millions of customers with varying budgets, setups and levels of expertise. They have to think about it from that approach.
But as agencies, we’re only concerned with taking care of our clients.
My agency works extensively with healthcare clients, and one of those verticals is called applied behavior analysis (ABA) – a very specific type of therapy for children with autism.
If I apply either phrase or exact match to that keyword, it will bring in very targeted traffic from people looking for ABA therapy for their children. The second we switched to broad match in a test, we got clicks from people searching for cognitive behavioral therapy and general therapy for people with autism.
Broad match moves it away from ABA to more generic treatment, from children to adults and teenagers. For my clients, none of that is relevant. But at the same time, not all traffic is irrelevant – only around 50%.
When your budget isn’t as deep as the Fortune 500s, 50% of your searches being irrelevant gets expensive quickly.
For smaller and growing accounts, it’s typically a good idea to start with phrase or exact match and layer in broad match once you capture at least 80% of impression share if you need to increase lead volume. In doing so, you should understand that lead quality may decline.
What you should be doing to prepare
You’d be hard-pressed to find a PPC marketer who doesn’t think Google could be doing some things better.
Google is changing from an analytics-charged, action-driven platform to more of an informational exchange platform.
It’s important to stop thinking in terms of clicks and conversions and instead consider audiences, search behavior and creative assets.
We’re almost moving back to a traditional marketing or advertising mindset. To be fully truthful, you should have been making this transition at least four or five years ago.
I often think that someone starting out with Google Ads today is in a better position because they’re free of preconceived notions about how things used to be done and “should” be done or pining for the good old days.
Mike Ryan from Smarter Ecommerce recently shared a presentation called “Decoding PMax,” the intro to which resonates deeply with what I have always advocated: People often say that you can’t “optimize” Performance Max because you have to set it and forget it, but the reality is there’s plenty to do. It’s just not necessarily in the campaign.
Continuing to brute force your way through Performance Max optimization is a losing battle.
So focus on what truly matters: