Unless you’re a centipede living under a rock, you know that AI tools like ChatGPT and other large language models (LLMs) have completely changed how many businesses and marketing professionals operate.
If you’re a business owner, you’re confronted by a thousand tech startup gurus who want to sell you their AI integration technology. If you’re a marketer, you’re bombarded with options for new AI tools and software, and tips for how to optimize your clients’ content for AI-driven platforms, like Google’s AI Overview or Amazon’s voice-operated Alexa. And both business owners and marketers have to decide which AI tools to integrate and which ones to ignore—all while the competition is making up its own mind on these things.
Indeed, the most powerful thing about these new AI tools seems to be not the tools themselves but the hype and controversy that surrounds them. At Beyond Blue Media, we see the hype. We love the hype (sometimes). But we also don’t let it direct our every move. As data-driven marketers, we’re proud to offer both our clients and colleagues a level-headed assessment of current AI trends and how they are currently affecting organic search, and what it might mean for the future of organic search.
Terminology: Understanding “AI” and “LLMs”
Before we dive into how AI is affecting search, let’s get some of these terms out of the way. While “AI” is now a common shorthand to describe things like ChatGPT, it’s important to note that ChatGPT is a Large Language Model (LLM), not a true “artificial intelligence.” So what’s the difference?
LLMs can process a ridiculous amount of information quickly and can even adapt their “thinking” on the fly. But this “thinking” is still limited to the boundaries of whatever the LLM has access to—namely, content that has already been written and fed to them. They “learn” to properly restitch and mimic our language by downloading and analyzing content from billions of web pages. So while LLMs may “create” things like jokes, stories, or vivid descriptions, their creations are never wholly unique and are always a composite sketch of what’s already been published on the internet.
In examples like Google’s Gemini or Amazon’s Alexa, we find LLMs that are essentially overlaid with voice recognition technology and programmed to interpret the basics and intricacies of human communication. These platforms offer what are essentially spoken versions of conversations one could have with ChatGPT; but again, they are LLMs, not full-blown artificial intelligence (they can still eavesdrop, though).
From “Search Engines” to “Answer Engines”
For the last 30 years or so, businesses looking to get in front of more users online would engage in “search engine optimization,” or SEO—the practice of improving your website for both users and the search engine’s algorithm, so it is more likely to include your site in results for relevant queries.
Now, with companies like Microsoft and Google integrating LLMs with their search engines, you get things like Google’s AI Overview results and Bing’s generative search. While they don’t entirely replace search results, these LLM-derived “answers” appear at the top of the results page and offer users a direct and instant response to their query as opposed to sifting through the results.
With the advent of Google’s AI Overview and LLM-assisted search, many businesses and marketers are wondering whether it makes more sense to optimize for the new AI-driven search models—dubbed “answer engines”---because of their ability to answer a query directly and succinctly without analyzing an entire list of search results.
Answer engines are not entirely new. Apple’s Siri, launched in 2011, was an early version of an answer engine, capable of reporting the weather and other mundane informative tasks. An earlier answer engine, the website Wolfram-Alpha, has been around since the late 2000s, offering users direct answers to common questions by scouring a massive database of academic content.
However, newer, LLM-assisted answer engines are far more dynamic and have access to far more information than these early examples.
So is Google an “Answer Engine” Now?
Not exactly. The way it currently incorporates LLM technology, Google can be thought of as a hybrid search-answer engine. The AI Overview does function as a new type of “answer engine,” but the rest of the organic search results are still presented below it, and the signals Google uses to determine which information to include in the AI overview are essentially the same signals it uses to rank content in regular search results.
To see this in action, take for example a search for “where is the Taj Mahal?” Google’s AI Overview delivers the answer immediately, along with links to the websites where it got the information:
However, below the direct answer, we still have the standard set of internet search results—including photos, videos, FAQs, and links to informational sites—that allow users to delve deeper or get other questions answered.
The value of this hybrid approach is obvious when you think about it: while the technology now exists to provide immediate, direct answers to queries, not all users are asking the same type of questions, and not all of them are even asking questions. Sometimes users are looking to compare prices for a product, sometimes queries do not have direct, singular answers, and sometimes users just want to know more than one thing.
Google, Bing, and their competitors have always claimed that their algorithmic adjustments were always made to improve the user experience; thus, it makes complete sense that Google and Bing are now effectively functioning as both search engines and answer engines—those who want their answers right now can have them, and those who need something else or want to dive deeper after getting an answer can still easily get what they want.
Can I Get Conversions from AI Overviews?
This depends on a few things, most importantly the type of business and the type of content you’re creating. For example, if you’re a local HVAC company and you want to be cited in the AI Overview answer for the query “how does a furnace work?”—you’re probably out of luck. There are simply too many pages answering that question for the engine to consider you, a local HVAC company in Ohio (or wherever), as the penultimate authority. In this case, your best bet is to stick with SEO best practices, such as Google Business Profile optimization, to get more local business.
However, if you own a company that serves a highly specific niche or has rare expertise, such as how to use liquid nitrogen to cool concrete, then your informational content would naturally stand a better chance of being cited in Google’s AI Overview, along with the link to your site.
So the short answer to “can I get conversions from the AI Overview?” is, “yes, you can”—but your content must be excellent and it must carry some kind of significant authority signal to the search engine. For my Taj Mahal query, Google brought up the UNESCO site, as that is a world-renowned institution for preserving historical places; it did not bring up, for example, a local restaurant that might have its own history of the Taj Mahal on its website.
Will I Miss Out if I’m Not in the AI Overview?
The examples above also bring up a common assumption about the AI Overview and answer engines more broadly: that they will divert traffic away from the websites in search results that would typically hold the information in the answer. While it is not technically a myth—click-thru rates to websites for informational searches have been declining for years, and more rapidly since the advent of AI-integrated search—the quality of the lost traffic is rarely discussed.
In truth, most traffic that went to a site from informational queries did not result in a traditional “conversion,” i.e. a purchase, form fill, or phone call; users more often retrieved the information they were after and then left the site. So while these searches had previously generated “traffic,” what is being lost with the shift to answer-engine behavior isn’t as valuable as is commonly assumed.
If you operate a business that isn’t naturally suited for mention in the AI Overview, don’t sweat it. There are plenty of other ways to get found online, and BBM can help you with that.
If you are able to get into AI Overview, however, it can bring some pretty solid benefits, even if for now they are mostly on the visibility side. For example, a recent study of searches for business to business (B2B) products found that about 90% of users clicked through to the citations in the AI Overviews. Moreover, a separate study from the UK suggested that for at least some searches and businesses, AI Overviews have actually funneled more traffic to relevant sites, not less. The difference seems to be one of quality over quantity, both in terms of the answer delivered and the traffic going to the cited page.
How Do I Optimize for Answer Engines?
Again, there is a lot of hype and buzz about “answer engines” and “answer engine optimization,” so this, from the mouth of Google itself, bears repeating:
The content signals used to create Google’s AI Overview are the same signalsthat Google uses to rank content in organic search results.
What this means is that if you write clear, helpful, well-organized content that addresses specific questions about your product or service—congratulations! You are doing both SEO and AEO, and you are still giving yourself the best chance to rank high in organic results as well as appear in the AI Overview.
Will AEO Be More Important in the Future?
This is hard to say, but given that users aren’t likely to stop searching for other things besides direct answers, I would venture that AEO won’t become separate from or outpace SEO in importance anytime soon. And no, I’m not an excessively tech-savvy person. I don’t understand the true ins and outs of AI or LLMs, have a degree in computer science, or have direct insight into where the technology is going. I would add that few who are commenting so vigorously on this subject actually possess these credentials, either.
However, as someone who has studied, written about, and interacted with people in various careers and life situations, especially marketing, what I can say is that I know people. And people are never going to want to get their information in just one way, and businesses are always going to be coming up with fresh ideas and innovations to get them the information they want in ways they ask for, or sell them products and services that solve a problem they have.
Just like these exchanges themselves, the spaces and platforms where they occur do not die or vanish—they evolve, and though it might happen faster than some of us would like or can fully comprehend, it’s only going to keep happening, and all we can do is keep an open mind so we can evolve accordingly. Beyond Blue Media is committed to helping you cut through the noise of today’s marketing landscape and do what’s best for your business. Go beyond average marketing; contact us today to see what we can do for you!
