Why the winning brands in AI-powered search are the ones that never stop doing SEO right and why shortcuts actually backfire.
There’s a version of the current AI research moment that’s being talked about a lot in marketing circles. It goes like this: Google has changed, search has changed, and AI now answers questions directly, so the old rules don’t apply and you need a new strategy built around AI visibility, geolocation, AEO, citations, and mentions.
We understand why this narrative is so appealing. It’s fresh, it’s urgent, and it gives people something tangible to sell.
It is also wrong in most important respects.
What is actually happening in AI research is much more interesting and more demanding than the “everything has changed” story suggests. And if you’re a business trying to navigate it, understanding the difference between hype and reality could be the thing that protects your organic performance over the next 12 to 18 months.
What Google has already said about AI search and SEO
Google recently posted Formal guidance On how to improve generative AI features, including an overview of AI and AI mode. For anyone expecting a new set of rules, the guidance was a cold shower. The message, clearly stated, was this: SEO best practices are still relevant, because our generative AI features are rooted in our core search classification and quality systems.
In other words, AI Overviews and AI Mode are not two separate systems operating on separate signals. They’re built on the same index, the same quality ratings, and the same rating systems that define standard organic search results. The AI layer retrieves content from pages that already rank well, applies Retrieval Augmented Retrieval (RAG) technology to pull specific information from those pages, and generates responses based on that content.
What this means in practice is that the path to appearing in AI search responses goes straight through good SEO. Crawlable and indexable content. Strong technical foundations. Trusted and authentic pages proving real expertise. Reliable signals from third party. Everything that has always been important in organic search is still important because AI reads from the same source.
Google has also been clear about what doesn’t matter for AI search. You do not need to create special AI text files such as llms.txt. You don’t need to “break” your content into parts. You don’t need to rewrite everything in a specific AI style. You don’t need to follow unoriginal brand signals. Google emphasized that these tactics have no particular impact on the emergence of AI search.
What the data actually shows about AI content shortcuts
While Google was publishing its guidelines, SEO Consultant. Lily Rae was posting something every marketing team evaluating AI content tools should read carefully.
Over several months, it tracked more than 220 websites that had been publicly identified, either on their own or by AI content vendors, as users of AI content creation and scaling platforms. She wanted to know what happened after the case study headlines. The pattern that emerged across these 220-plus sites was consistent and stark. 54% of the sites we monitored lost 30% or more of their peak organic traffic. 39% lost 50% or more. 22% lost 75% or more. In many cases, the eventual traffic loss exceeds the peak gain, meaning sites are worse off than they were before they started scaling AI content.
The path was remarkably similar across industries: rapid growth in organic pages over six to twelve months, peak traffic three to six months after peak content, and then a sharp decline that typically erased most of the gains over the following year. Glenn Gabbie called this growth the “AI mountain”: steep growth, followed by an equally steep decline, once Google’s systems collect enough signals to understand what’s happening.
Most of the sites that declined were using a combination of eight content styles that Ray identified as high-risk: large-scale comparison pages, “what is X” glossary pages designed to quote AI, “best
What all eight have in common is that they are templates designed to influence rankings and AI citations, rather than content created because a real user really needs it. It can be detected as a pattern. When enough sites implement the same pattern at scale, Google’s systems become very good at identifying it and demoting it.
Ray also identified a possible, unconfirmed Google update in late January 2026, after which at least 40 sites she was monitoring saw organic traffic decline between 40% and 95%, most of it concentrated in blog subfolders where self-promotional lists and other AI-generated content were posted in large quantities. Some of these sites have seen the influence spread from the subfolder to the full scale.
Why this is important for AI search, not just traditional SEO
Here the two stories meet, and the dangers become clear.
AI search experiments, AI overviews, AI placement, and generative responses embedded in search engines are powered by RAG. They retrieve content from Google’s index and use it to generate responses. What is retrieved is what ranks. Ranking is determined by Google’s quality systems. Google’s quality systems are specifically tuned to detect and cut down the same types of low-quality, template-based content that Ray’s data shows actually breaks down in traditional search.
This means that the shorthand that many companies are being sold at the moment “scaling AI content to win AI citations” is a completely counterproductive approach. Sites that produce broadly boilerplate and formulaic content to pick up AI signals are doing the thing that will likely result in them being demoted from the index from which AI research is read.
In other words: bad SEO is bad geographic location. The signals that cause a website to lose visibility in traditional search are the same signals that cause it to lose visibility in AI-generated responses. They are not separate problems with separate solutions.
Winning brands in AI research
Ray’s analysis included an observation that hasn’t gotten as much attention as the decline data, but is arguably the most important finding: The brands that continue to grow across its data set, broadly, are those whose content doesn’t match the eight risky templates. This is not a coincidence. It reflects something that Google has been consistent in for years, and that its new AI search guidelines clearly confirm: The signal that a page is worth showing is whether real users will find it truly useful, authentic, and trustworthy. This signal does not change because the delivery mechanism has moved from a blue link to an AI-generated paragraph.
Brands that win in AI research tend to share certain characteristics. Their content shows real expertise, the kind that comes from people who actually know the topic, not from a claim that summarizes what’s already on the first page of results. Their pages contain authentic information: first-hand experience, proprietary data, unique perspectives, and specific examples. Their technical foundations are strong enough that Google can find and index everything they post. Their authority is supported by real signals from a third party; Coverage, citations, and links earned because they said or did something worth talking about.
These are the basics of SEO. It has always been the basics of SEO. The AI layer has increased the stakes of doing it well, because the content that appears in AI responses must be clearly relevant and trustworthy, and because the gap between the sites that do it and the sites that try to shortchange it is becoming increasingly visible in the data.
What this means for how we think about AI content tools
We want to be clear: AI content tools are not the problem, the problem is how you use them.
There are really valuable applications for AI in content workflows. Research and synthesis. Create a brief. Compile ownership data and present it clearly. Identify content gaps. Support writers who are subject matter experts but not natural writers. Accelerate the production of content that is still subject to expert moderation, fact-checking and editorial review at every stage.
What is risky, and what the data increasingly confirms is risky, is using AI to mass publish pages without quality controls. When the goal becomes the number of pages rather than the usefulness of each page, the content produced tends to look like the templates outlined by Ray. They are initially categorized because they are relevant. And it loses those rankings when Google’s systems collect enough signals to understand that it’s not really useful. Ray’s diagnostic questions are worth keeping on hand when you’re evaluating a content program:
- Does this page exist because a real user really needs it, or because a search engine or LLM might cite it?
- Can a competitor produce a near-identical copy of it tomorrow using the same router?
- Is there anything on this page; First-hand experience, proprietary data, and real point of view, not already in the top 10 results for this query?
If the honest answer to this last question is no, then the page probably isn’t newsworthy.
Practical result
AI research is real, and the changes it brings to how people discover, compare and evaluate brands are hugely significant. We do not suggest otherwise.
But the response to these changes should not be to separate AI search optimization from SEO quality and treat them as different disciplines requiring different approaches. This disconnect is precisely the error that produced the “AI Mountain” trails that Ray documents.
The correct response is to treat AI search as a reason for SEO, not differently. This means strengthening the technical foundations so that everything is crawlable and indexable. It means improving the usefulness and originality of your content so that what you post actually contains something that someone couldn’t find five seconds ago on a competing site. This means earning trusted third-party mentions through coverage, links, and brand mentions that are the product of real reputation, not artificial effects. This means being thoughtful about what AI systems can learn and infer about your brand from everything that’s out there about you on the open web.
The brands that will win in AI search, the ones that will be cited, recommended and trusted by AI systems over the next few years, will be the ones that produce truly quote-worthy content.
This has always been the case. The risks are higher now.
If you’re reviewing your content strategy in light of these changes and want a clear view on where the risks and opportunities lie, we’d love to take a look. Connect with the Koozai team.
