AI Shakes Up Search and SEO

While the impact of AI on many professions is still hard to assess, the effect it’s had on search results is becoming crystal clear.

The AI-generated summary at the top of the results page is now universal among search engines and many searches stop there. New research from Bain & Company found that 60% of searches now end without the user clicking on a link in the list of results; additionally, 80% of users “rely on AI summaries at least 40% of the time.” The effect on traffic to news sites is clear.

But the business models of search engines themselves are also being impacted as searchers move away from traditional search and go straight to AI to get answers. For the first time in a couple of decades, Google searches are falling and the news resulted last month in a 7% drop in the search giant's stock price.

For its part, Google has pushed back against the claim that search volumes are dropping. Nevertheless, according to Eddy Cue, Apple’s SVP of services, users of the browser Safari (where Google pays a small fortune to be the default search engine) are performing fewer Google searches.

The end of SEO as we know it?

While this is certainly scary news for publishers and search giants, it has big implications for SEO and search strategies across the spectrum. Marketers who have spent years fine-tuning their SEO strategies to ensure their brands surface at the top of Google search results find that it is no longer enough. Everyone needs to adjust.

Measuring the difference: Google vs ChatGPT

But how valid an alternative are the AI chatbots to traditional search engines?

To find out, the Tepper School of Business at Carnegie Mellon University conducted an experiment to see whether AI queries can really displace search.

Students split up into two groups and planned a trip. One group used traditional search while the other used GenAI. “The findings were striking: GenAI users reported a 17% higher satisfaction rate and achieved their goals faster. Specifically, 88% of GenAI users found what they needed on the first try, compared to 79% of Google users.” In another test, “Participants using gen AI search completed their tasks in less than half the time. Despite Google’s attempts to integrate AI snippets, most users ignored these features or found them ineffective.”

The research ultimately showed that users prefer GenAI for more complex tasks, but that traditional search still works best for simple queries, like finding a nearby restaurant or a quick answer.

The study's author concluded: "These insights suggest that GenAI is unlikely to completely replace traditional search engines. Instead, the future of search may be segmented, with each platform serving different needs."

AI search - the missing monetization

Another question remains: How will GenAI monetize itself? Google’s ad placement model won’t work when there’s no need to click through to find an answer to a question. At the same time, AI-driven searches are computationally more costly than standard search.

Students taking part in the Tepper study were asked to suggest possible monetization models, but the various possibilities (" sponsored follow-up questions, embedded content, or subscription models.”) all had serious drawbacks: Is anyone willing to pay for GenAI when free search engines still exist? And will users be put off by sponsored content that disrupts the experience? Without answers to these questions, it’s not clear whether GenAI will be able to sustain itself long enough to really give traditional search a run for its money.

What does this all mean for brands?

Even if that’s not the case, as click-throughs from search results drop, digital marketers have to completely rethink their strategies. Spending time and energy on SEO may feel futile at this point, and to some degree, it might be. However, some of what is going on behind the scenes at Google may convince digital marketers to invest in great content.

Quality is king, again

As Search Engine Land reports, Google’s servers are getting overwhelmed, especially with the increase of AI-created content, leading the search giant to spend fewer resources on crawling “websites that have higher content velocity to their trust rating.” This means that quality, in-depth content is more important than ever, and a specific kind of content will win the day:

“Today, the best way to get your content seen by Google is through content written with experience and opinion. AI cannot write based on experience.” In other words, the glut of AI-generated content is leading search engines like Google to ignore that content and focus on high-quality content created by individuals.

Optimizing your content for AI search

Interestingly, one of the main takeaways may be not to rely too heavily on GenAI to create your content. Starting from a base of “good content,” there are a few other best practices to keep in mind. Bain & Company outlined three strategies for marketers:

• Optimize for AI crawlability. Adapt content for semantic search, emphasizing high-intent, long-tail terms.
• Diversify content formats. Go beyond text. Use video and interactive formats to boost visibility in generative AI search.
• Redefine metrics. Shift from click-focused metrics to measuring search impressions and AI reach. Optimize for influence over direct conversions.
  

“Putting your audience first takes on a new meaning. Dive into search intent data to find opportunities to answer the same questions they are asking AI. ”

 C. David Minifie, Chief Marketing Officer, Terakeet, has more advice on Forbes. Putting your audience first — always a good practice — takes on a new meaning. Dive into search intent data to find opportunities to answer the same questions they are asking AI. Optimizing that content for conversational search will help make it more AI-friendly.

Can you control GenAI search?

Can marketers expect the same level of control of search results that they are used to with well-tuned SEO operations? David Minifie is optimistic: "We found that across the three search tools, GenAI results have a controllability of between 46% and 52% for individual brands and between 37% and 48% controllability for brand products and services brands."

Finally, it’s important not to abandon SEO entirely. Minifie says that he saw a correlation between search engine rankings and the likelihood that it will surface in AI Overview results.

For the time being, marketers are probably better off optimizing for both search modes.