Beyond Keywords: SERP AI Evolution and the Future of Paid Search
AI has taken the marketing world by storm, becoming the go-to buzzword for all things automation and machine learning. However, it's important to note that AI’s impact on the search landscape is unmistakable, with both Google and Bing adopting new features like Search Generative Experience (SGE) and the like. This integration is actively reshaping the realm of paid advertising, highlighting the substantive changes AI introduces.
Let’s cut to the chase, what is SGE & how is it impacting search?
SGE’s main purpose is to provide users a better search experience with an AI-generated display in the search engine results page (SERP). The display shares key information relevant to the search and is typically summarized in a more visually appealing format to give users more useful information. See an example of this at play below:
After initial search:
AI-generated response:
The first thing any marketer would notice is just how much space the generative result takes up. After clicking, the SGE content expands covering the whole screen, making it so one has to scroll down to see the entire content. The average height of the generative result is over 1,760 pixels, taking up almost the entire search result page. The second is that the first organic position has now dropped below the fold, again forcing the user to have to scroll to get to any organic link. Thus, making a significant impact on click-through rates and organic traffic.
On the other hand, how do paid listings come into play with SGE?
It is safe to assume that Google and Bing have no intention of losing out on paid advertising. Ads can be expected to be continually integrated with SGE, as it becomes increasingly prevalent. For now, ads can appear both above the SGE result and directly below it for Google search results. For Bing, ads can appear in line with the response. See an example below:
Now, it is important to note that SGE has not rolled out en masse for Google, as it is currently only available in an experimental labs form. However, it is already ingrained and live with Bing search results and likely will be fully rolled out for Google soon. What we can expect with full rollout in Google is around 86% of all search queries to display a search generative element. So, make no mistake, we will see SGE in the majority of our search results very, very soon.
Adapting your paid search strategy to prepare for SGE.
Traditionally, paid search is a “keyword” centric channel, where we base our ad efforts strictly on query targeting. In the past, this has led to the majority of marketers adding thousands upon thousands of keywords to try to capture as much search volume as they could. However, this is a very mundane, brute force method that is very trying to manage effectively.
With recent engine updates, particularly the relaxation or “loosening” of match types such as exact match, and the introduction of SGE and other algorithmic changes, there is now less certainty regarding how to effectively position paid search targeting. This is why we need to adapt our paid search thinking to focus not so much on the keyword, but on the user. At Rise, we look to have an audience-focused approach and try to understand what stage the user is at in their conversion journey. We can look to implement this strategy at scale in several different ways to try to “future-proof” our strategy for SGE:
Audience Layered Structure
One way we can utilize the search engines to our advantage is by leveraging their audience data to help categorize users based on where they are at in their purchase phase. We can then use a distinct campaign structure to target those audiences directly and show different messaging to guide them further down the purchase funnel. At Rise, we do this by splitting up our campaign into two different buckets:
Audience-Linked Campaign:
These campaigns layer on the search engine’s “in-market” audiences as well as normal keyword targeting to provide an extra layer of intent, reaching users who are further in their journey. Specifically, this campaign targets users that have been matched with keyword targeting as well as being “in-market” in the audience segment. We can then utilize more conversion-focused messaging to drive them to make that purchase. For example, if we are targeting a user looking for a mountain bike, we can target the keyword “mountain bike” and can also target those in-market for “Bicycles & Accessories”. So, we know, through Google data, they are already in the market for the product, and through keyword targeting, they are actively searching for mountain bikes at this moment. Thus, showing them limited-time deal messaging with a call to action can push them over the finish line to purchase their new mountain bike.
Non-Audience-Linked Campaign:
This campaign would firstly negate the audience we have targeted above, to negate any overlap in targeting. We would only target keywords, such as “mountain bike” for those who may be starting their research into determining which product is best for them. As such, for this group, we may want to show more brand-focused, discovery-type messaging to introduce the user to the brand and its differentiations.
Broad Match Testing
Another way we can adapt to more of an SGE environment is through broad match testing. With the rollout of AI, people are searching more conversationally, almost as if talking to a friend to get recommendations or get more information. As a result, search queries are becoming longer, often twice as long as traditional searches, as people treat their search as more of a conversation with the search engines through SGE. The longer the query the more permutations that may come off it, which becomes too daunting to cover with just exact match keyword targeting.
However, broad match keywords give us more ability to cover each relevant search that may come through and a better opportunity to more easily scale, as opposed to brute-force adding many new keywords singularly. For example, when searching, with the introduction of SGE, a user may begin with “what are some good recommendations for biking near the rocky mountains”. A simple “mountain bike” exact match keyword may not cover this query. However, a broad match “mountain bike” keyword could give the ability to show mountain bike product ads after SGE begins to knead interest by showing relevant information that the user was looking for. Then the user may feel more persuaded to purchase a mountain bike for their journey.
Now, while broad match is of higher utilization, it is also a bit of a polarizing topic for marketers. This is especially true for those who have had bad experiences with broad match casting too wide of a net and wasting ad dollars by showing ads to irrelevant searches. The search engines have maintained that broad matching has improved over the years, but there are still understandable reservations for marketers to fully leverage this tactic. At Rise, we balance this hesitation by utilizing a robust negative keyword strategy, as well as layering on our audience-based campaign structure, to funnel out irrelevant queries and protect spending in the wrong areas. We also utilize our media optimization platform, Connex®, to automatically monitor search term reports and ensure that we are only targeting the right users.
The search landscape will continue to evolve and change with more and more reliance on machine learning and artificial intelligence, especially with the upcoming rollout of SGE. Having an audience-focused paid search strategy that not only adapts to the new SERP changes, but can also be resilient to the constantly changing search landscape is becoming even more crucial for future success.
Ready to capitalize on AI to reach your audience? Contact us today!