We Took Google’s SGE for a Test Drive—5 Learnings for Paid Search Advertisers


My experience reviewing ad content in the SGE landscape has left me Googling who said “the more things change, the more they stay the same.” Turns out it was Jean-Baptiste Alphonse Karr for those of you, like me, with an inability to ponder something random and not immediately look it up. So while our friends in the SEO industry are understandably watching the SGE experiment with a keen eye, because it is a monumental shift in the way people will engage with online content, I’m taking a more laissez faire approach and focusing on controlling what we can.

Paid Search Is Too Big To Fail

To paraphrase Robert Walser, “money runs the world.” As cynical as it sounds, from a business perspective, he’s not wrong. At their core, businesses exist to make people money, so it’s easy to see why no matter how much the search landscape shifts, we can remain confident Google will find a prominent place for search ads in it. They’re not going to just disrupt $40.4 billion in revenue without a well thought out plan. Initial testing has demonstrated that best practices for ads still apply 10/10 times:

Use the right bidding strategies and budget, create engaging ad content, land on a page optimized for the terms you’re bidding on and designed to convert the user.

Let’s look at some real world examples. In this query for kids trampolines (that’s totally unrelated to shopping for my son’s birthday) we have five different advertisers occupying the top 3 search results. The most prominent spots are taken up by product ads from Jumpflex USA, Walmart and Jumpzylla. The search ad spots went to Trampolines.com and Amazon.

If my intent was strictly to purchase right now, these ads would all be extremely helpful. If I were just looking for info about kids trampolines, then not so much. I’d have to scroll down, modify my search, click and visit several different sites before being ready to buy. That’s the traditional method to search advertising in the standard google search interface.

  1. Identify a query someone would likely use when wanting to buy your product
  2. Bid on it
  3. Hope they were wanting to buy when your ad showed and that you had the best offer for their needs
Source: google.com

Automated bidding strategies for conversion value have improved the process dramatically by factoring in the consumer’s likelihood to purchase. You can now only bid on auctions when the signals suggest I’ve already done my research and my kids trampoline query is because I put off ordering it and now I need 2-day rush shipping.

PMax audience signals take automated bidding even further by incorporating the ability to target specific audience qualities. You could add in contextual targeting of sites your ideal shopper visits, demographics like “parents of toddlers” other search terms like “are kids trampolines safe?” TouTube videos of “backyard trampoline kids parties” etc. to create a profile of the exact type of shopper you’re looking for. With PMax your ads will be served to that consumer across all Google properties as they search and move closer to purchase.

Even with PMax, search is still extremely complex. Users still need to conduct queries across channels and sites, but your ads can keep up with them better and your messaging / product stays top of mind throughout. If SGE takes hold as the new standard for search though, that complicated path to purchase may get a lot more linear.

The new SGE experience has the potential to completely streamline the online shopping experience and make marketers lives a little bit easier along the way. Instead of worrying about all the various placements you might need and queries you’ll have to bid on, if you’ve done a good job identifying your audience you just need to show up in shopping and let the generative AI do the rest.

Source: google.com

As Lincoln pointed out in his blog, this new experience can mean less traffic to websites from organic search. It could also mean fewer ads are shown per search. In combination, we could be looking at fewer opportunities to remarket, less data to build audiences from and most importantly an increase in the significance of consolidated single search events. Phrase/broad match keywords may soon carry more value both as Google continues to push letting AI “do its thing” and as the search landscape changes. Where a consumer previously may have searched one exact phrase and clicked to a site, with the expanded use of natural language models, consumers will be free to have search conversations with the AI system. This will likely lead to more complicated search journeys, including more follow up queries all happening in the SGE landscape before a click to a site. If that’s the case, a greater variety of keywords will carry value and Phrase and broad match will allow you to stay a relevant part of the conversation in ways exact match may not.

  1. People will still click ads

People have needs. It’s why they search. They want to learn, be surprised, thrilled, impressed, reassured and helped. That isn’t changing. AI may do a little bit of all of that for them right in the SGE interface, but businesses and advertisers are also getting better at understanding those needs and curating products and experiences to address them. Quality ad content shows consumers a good or service in ways that educate, thrill, surprise, impress, reassure or help them and it gets clicked.

As paid search experts our job will still be to understand those audiences, segment them, creatively approach content & copywriting for each individual’s need and ensure our ads are getting placed in front of the audience at the right times and in the right places.

  1. The traffic you get from ads will likely be higher quality

Google’s purpose for existing is likely never to change. Their mission is to “organize the world’s information and make it universally accessible and useful.” To that end, generative AI is a logical trajectory they’ve likely always been on. It’s an evolution of features like “people also search for” and structured snippets in the knowledge graph. It’s a positive disruptor to the way we think about solving our problems and learning new things. Instead of clicking an ad just to get to a company website and learn about a new product, an SGE search will give that consumer all the information they would have had to hunt down themselves. Now when they click to your site, they’ll be that much more likely to purchase or engage.

  1. Paid KPIs will more or less stay the same

We’re still going to be responsible for driving cost-effective conversions. Efficient paid search has always been reducing spend on keywords that target “early in discovery” shoppers and instead focusing budget on “later in discovery shoppers.” In the new world of SGE though, those lines blur and we’ll likely be presented with fewer opportunities and a smaller funnel. It will be that much more important to be paying attention to the details of search keyword reports, impression share comparisons, landing page experience and CVR.

  1. “Basic” Ad Strategies Just Won’t Cut It

As the search landscape becomes increasingly complex and the value of a single search is potentially on the rise, the ability to “set and forget” a campaign will cease to exist. You’ll need to be mining paid search data to inform other marketing tactics, keep up with trending products and ensure a relevant content strategy is in place that continues to allow you to maximize your investment in the paid space.

  1. Ad Audiences Will Matter More Than Ever

Expanding on the fact there will be no further place for “basic” strategies, your ability to understand and guide user intent by leveraging audience targeting, will be significantly more important. You’ll need to bring in additional relevant websites these users may be visiting outside of your own and increase the complexity of the search phrases used to distinguish between them so that you can better separate informational top of funnel shoppers from the ready to buy right now ones, even if they’re using the same keyword.

Google has been investing heavily in search and PMAX and has recently announced new AI powered features for both. These features work best when used with well curated audiences. You tell Google everything you can about your ideal customer, the AI takes over and figures out the right ad content to serve them as part of their search experience and across all Google properties.

Feel Like You Don’t Know Where To Begin?

With so few known quantities and a lot of guesswork having to be done to try and stay ahead of the curve, it can feel overwhelming to figure out the best place to dive into the world of search AI. So don’t start with that. Start with your customers. Make sure you understand the type of information they’re looking for today, where they’re sourcing it and how you can be a bigger part of that conversation with your own content strategy. Then take portions of that content strategy and turn it into an ad strategy.

After that, enroll in Bard, try out SGE to find everything you need for your 4th of July party, ask ChatGPT to curate a meal plan for you, or even chat with Bing about the weather. You could also consider following one of the million Tiktok and Instagram accounts that have popped up promising to keep you up to date on all things AI. 

The best thing you can do is try to collect relevant data and experiences on these tools so you can start to form your own opinions about how they might intersect with your desired customer journeys. After all, the question isn’t has the search landscape changed, it’s what are you going to do to keep up?