SEO Still Matters. AI Is Expanding How Brands Get Discovered
Search is not disappearing. The environment around it is expanding.
A buyer may still type a query into Google and work through the results. But that same buyer can now ask ChatGPT to explain a category, use Perplexity to research suppliers, encounter an AI-generated summary in Google Search, or use Copilot or Gemini to synthesize information before visiting a company website.
These experiences do not eliminate traditional search. They add new ways for information to be found, interpreted, and used.
That changes the visibility problem.
For years, search strategy has largely focused on whether a company could earn a prominent position for the right query and turn that visibility into traffic. That still matters. Increasingly, however, a company's information may also become part of a synthesized answer, appear as a supporting source, influence how a category or brand is explained, or contribute to a decision that begins somewhere other than a conventional search-results page.
The terminology surrounding this shift is still unsettled. AIO, GEO, AEO, LLMO, AI search optimization, and AI visibility are all in circulation. The label matters less than the underlying change.
Search visibility is becoming a broader information-discovery challenge. Companies need to think about whether their expertise can be found, understood in the right context, verified, and represented accurately across a growing set of search and AI-mediated environments.
01
SEO Still Matters. Discovery Is Expanding.
The arrival of AI-generated answers has produced plenty of premature declarations about the end of SEO. Current guidance from the search platforms themselves points in a different direction.
Google's guidance for AI Overviews and AI Mode says that its existing SEO fundamentals remain relevant. Pages still need to be accessible, indexable, useful, and eligible to appear in Search. Google also says there are no additional technical requirements for its AI search features and no special schema or machine-readable AI file required to appear in them.
That is an important distinction.
Technical health, crawlability, strong internal linking, clear site architecture, useful content, authority, and accurate structured information have not suddenly become obsolete. They remain part of the infrastructure through which information is discovered on the web.
What has changed is what can happen after discovery.
A traditional search result typically presents a set of sources for the user to choose among. AI-powered search experiences can retrieve information from multiple sources, synthesize it, and present part of the answer before the user decides whether to click anything at all.
Even there, the evidence does not support a simplistic "AI replaces links" narrative. A 2026 Microsoft Research eye-tracking study found that people interacting with search pages containing generative AI results devoted significantly more attention to the AI-generated content while continuing to engage with traditional links in recognizable patterns.
The practical implication is not that search has become irrelevant. It is that discovery now has more surfaces.
A company may need to earn a ranking, a citation, a mention, a recommendation, or an accurate description depending on where and how the research is taking place.
02
AI Discovery Raises the Bar on Clarity and Authority.
When information is being synthesized, context becomes especially important.
Consider what must be reasonably clear for an outside system—or an unfamiliar buyer—to make sense of a company.
Who is the organization? What does it actually do? Which markets does it serve? What is a product versus a service versus a methodology? Who created the content? What experience supports the point being made? Are important claims documented? Are company facts consistent from one page to another? Does the organization demonstrate meaningful depth on the subjects it claims to know?
Those questions are bigger than keywords.
They point toward entity clarity: making the important people, organizations, products, services, locations, concepts, and relationships surrounding the business understandable rather than leaving them implicit or inconsistently defined.
The problem becomes tangible when a company describes itself differently depending on where someone encounters it. If METIS is clearly positioned in one place as a Growth Intelligence company for professional markets but elsewhere is loosely characterized as an agency, consultancy, or MarTech firm, those descriptions create different understandings of what the company actually is. Consistency does not guarantee how any particular AI system will represent a brand, but reducing that ambiguity makes the underlying information clearer for customers, search engines, and AI systems alike.
Structured data can contribute to that clarity. Google describes it as a way to provide explicit information about the meaning and classification of content on a page. But even correctly implemented structured data does not guarantee a particular search appearance, and Google's AI-search guidance says no special structured-data treatment is required for AI Overviews or AI Mode.
The same principle applies across the broader information environment. Consistent company descriptions, clearly defined products and services, credible author information, connected content, supporting sources, and accurate references all make it easier to establish who the company is and why its information matters.
OpenAI's current publisher guidance provides one practical example of how basic discoverability still matters in AI environments: public websites can appear in ChatGPT search, but publishers that want their content available for summaries and citations need to allow access to OAI-SearchBot. OpenAI also provides referral tagging so publishers can identify visits arriving from ChatGPT.
The durable objective is not to reverse-engineer every proprietary retrieval or citation system. It is to make the business and its expertise difficult to misunderstand.
03
Generic Content Is Getting Cheaper. Original Expertise Is Becoming More Valuable.
This may be the most consequential content shift created by generative AI.
Producing a competent overview of a common subject is now extraordinarily easy. A company can generate ten articles explaining industry trends, five definitions of the same concept, and another round of generic "best practices" before lunch.
So can everyone else.
As the supply of competent but interchangeable content grows, simply publishing more information becomes a weaker source of differentiation.
The scarce asset moves toward what cannot be generated from a generic prompt without access to the underlying knowledge: original research, first-party data, proprietary methods, specific operating experience, detailed examples, useful comparisons, credible evidence, and a point of view that reflects real understanding of the market.
Google's own people-first content guidance is notable here. Its quality questions ask whether content contains original information, reporting, research, or analysis; whether it adds insight beyond the obvious; whether expertise is demonstrable; and whether material drawing on other sources adds substantial value rather than merely rewriting them.
Those principles predate many of the current arguments about AI visibility, but they become more important in an environment flooded with inexpensive derivative content.
Imagine two manufacturers publishing about contractor ecommerce adoption.
One produces a familiar article explaining that digital tools are becoming important and contractors value convenience.
The other analyzes its actual pro network and shows that ecommerce adoption differs materially by business size, product category, purchasing pattern, and branch relationship. It documents where contractors stall in registration, what field teams are observing, how assisted onboarding changes behavior, and what the company still does not know.
Both pages can be technically optimized.
Only one contributes knowledge that did not already exist in hundreds of other places.
That is the larger content opportunity in an AI-rich environment. Companies do not need to become content factories. They need to become better publishers of what they uniquely know.
The website, in that sense, becomes more than a collection of campaign pages and SEO articles. It can become an owned repository of market knowledge: the organization's expertise, evidence, methods, definitions, examples, and evolving point of view made accessible to customers, search engines, and AI-powered discovery systems.
04
Structure Helps Machines. Substance Gives Them Something Worth Using.
Strong thinking still needs to be legible.
Clear page titles help establish what a page is about. Logical headers make complex information easier to navigate. Descriptive URLs, internal links, and well-maintained site architecture help connect related ideas. Appropriate structured data can provide explicit information about entities and page content. Sources and citations make claims easier to verify. Author pages and organizational information can establish who is behind the material. FAQs can be useful when they answer questions customers genuinely ask rather than existing simply to create more indexed text.
Microsoft's 2026 guidance around its AI Performance tools similarly recommends strengthening depth and expertise, improving clarity and structure, supporting claims with evidence, maintaining current information, and reducing ambiguity across the way entities and concepts are represented.
None of those practices should be mistaken for a secret AI-content format.
Google explicitly says that publishers do not need to create new AI-specific files or special schema to participate in AI Overviews and AI Mode. Its broader guidance continues to emphasize helpful, reliable content created for people rather than content engineered primarily to manipulate search systems.
That distinction should shape content strategy.
A detailed industry analysis should not be reduced to a shallow listicle because a marketer believes an LLM will find bullet points easier to parse. A nuanced argument should not be broken into dozens of artificial question-and-answer blocks because someone has declared that to be the latest AI-search hack.
Clarity is useful because it helps ideas travel.
Structure is the transmission layer. It cannot manufacture expertise that is not there.
The better question is whether the page contains something specific and credible enough that a customer would learn from it, an industry peer might reference it, and a search or AI system has enough context to represent it accurately.
05
Search Visibility Can No Longer Be Measured Only by the Click.
Traffic is still important.
So are rankings, impressions, click-through rates, organic conversions, and branded search. The emergence of AI discovery is not a reason to discard the measures that already show whether search is creating business value.
It is a reason to widen the measurement perimeter.
In 2026, Google began introducing dedicated Generative AI performance reporting in Search Console for a subset of sites, separating impressions generated through features such as AI Overviews and AI Mode. Microsoft now reports AI citation activity in Bing Webmaster Tools across Copilot, AI-generated Bing experiences, and supported partner environments, including cited URLs and sampled grounding queries.
OpenAI, meanwhile, enables publishers to identify referral traffic from ChatGPT through tagged outbound links.
Taken together, these developments signal something more important than any individual platform feature: visibility is beginning to include what happens inside synthesized-answer environments as well as what happens on a traditional search-results page.
A broader search scorecard may therefore include traditional organic performance alongside emerging signals such as AI citations or references, referral traffic from conversational platforms, brand mentions, and whether important AI systems represent the company accurately.
Those newer measures are still evolving. Citation frequency does not necessarily equal authority or business impact. "Share of AI visibility" can depend heavily on which prompts and platforms are being tracked. Increases in branded search after AI exposure may be meaningful without proving causation.
There is not yet one accepted framework for measuring AI visibility.
For now, the strategic move is straightforward: keep measuring rankings, traffic, and conversions while beginning to observe the additional places where discovery and influence may now occur.
The METIS Perspective
Become a Source Worth Finding—and Worth Referencing.
The deeper change in search is ultimately an information strategy problem.
A company's website has traditionally had several jobs: explain the business, generate demand, support campaigns, rank for relevant searches, and move visitors toward action.
It increasingly has another role.
It is one of the primary places where the organization can establish an authoritative, connected body of knowledge about who it is, what it knows, and what it believes.
That means individual content pieces should not exist only to capture isolated keywords. Over time, they should contribute to a clearer picture of the company's expertise: consistent definitions, connected subjects, credible examples, original insights, supporting evidence, and a recognizable point of view.
For a company with meaningful market expertise, this is intellectual property made discoverable.
The objective is not to turn the website into an artifact engineered for an LLM.
It is to make the organization's knowledge clear enough to find, specific enough to understand, credible enough to trust, connected enough to establish context, and useful enough that someone—or some system helping that person research—has a reason to reference it.
That is good search strategy. It is also good publishing.
SEO remains part of the foundation because information that cannot be crawled, indexed, navigated, or understood has a discoverability problem before AI enters the conversation.
AI discovery extends the challenge. Companies now have to consider whether the same body of information can survive synthesis: whether the brand is represented accurately, whether its expertise is distinguishable from generic commentary, and whether its strongest ideas have enough evidence and context to stand on their own.
The companies best positioned for this environment will not be the ones that master the newest optimization acronym.
They will be the ones building something more durable: a credible, maintained body of knowledge that people can find, search engines can index, AI systems can interpret, and the market has reason to trust.
The Signal
Better growth decisions start with better intelligence.
Catch the Signal for practical thinking on Growth Intelligence, professional markets, connected data, AI, go-to-market strategy, and what is changing next.
Catch the Signal. Stay ahead of the noise.
Industry Research
The article is grounded in current platform guidance and 2026 research.
Google Search Central — AI Features and Your Website. Google's current guidance states that established SEO fundamentals remain relevant to AI Overviews and AI Mode, that there are no additional technical requirements to appear, and that publishers do not need special AI files or AI-specific schema. Google: AI Features and Your Website
Google Search Central — Creating Helpful, Reliable, People-First Content. Google's content-quality guidance emphasizes original information, research and analysis; demonstrable expertise; clear sourcing; and substantive value beyond rewriting existing material. Google: Creating Helpful, Reliable, People-First Content
Google Search Central — Generative AI Performance Reports, June 3, 2026. Google introduced dedicated Search Console reporting for impressions in AI Overviews and AI Mode, initially for a subset of website owners. Google: Generative AI Performance Reports
Microsoft Bing — AI Performance in Bing Webmaster Tools, February 10, 2026. Microsoft's reporting includes AI citations, cited pages and sampled grounding queries across Copilot, Bing AI answers and supported integrations. Its accompanying guidance also discusses content depth, evidence, clarity and freshness. Microsoft: AI Performance in Bing Webmaster Tools
OpenAI — Publishers and Developers FAQ. OpenAI explains how public websites can be made available to ChatGPT search through OAI-SearchBot and how publishers can identify referral traffic from ChatGPT. OpenAI: Publishers and Developers FAQ
Microsoft Research — 2026 SIGIR eye-tracking study. Researchers found that generative results receive significant user attention while traditional links continue to be used, supporting a more nuanced view of AI answers and conventional search results coexisting. Microsoft Research: Are AI Overviews Changing Search Behavior?