The SEO You Built No Longer Works for Today's Search

Google I/O 2026 left little room for interpretation: AI-generated answers are now the default format in Google Search, not an ongoing experiment or a secondary tab. For any team that has spent the last few years building domain authority, crafting link architecture and tracking positions in the SERPs, the operating landscape has changed substantially.

TechCrunch's Equity podcast dedicated a recent episode to breaking down exactly this problem: most brands have virtually no visibility into how AI describes them to their customers. This is not simply about losing organic traffic; it is about the intermediary that now summarises, recommends and filters no longer leaving a trace in Google Search Console.

What Changed at Google I/O

The official presentation consolidated AI Overviews, the automated responses generated above any search results, as the default experience for a growing proportion of queries. Google does not provide an exact figure for what percentage of searches are affected, but traffic data from numerous publishers since late 2025 points to sustained drops in clicks even when rankings remain unchanged.

The mechanism is familiar: the model responds directly on the results page, the user gets what they need without clicking, and the site that informed that answer receives no visit. What is new is the scale and the official status of the change.

The Black Box Problem

What compounds the situation for brands and marketing teams is not just losing clicks; it is not knowing what the AI says about them. Traditional monitoring tools, rankings, impressions, click-through rates, do not capture how a model synthesises a product's reputation or a company's trustworthiness. A brand can hold the top organic result and still appear described inaccurately or simply absent from the generated response.

This opacity has immediate practical consequences:

• Uncontrolled reputation: if the model pulls outdated information or low-quality sources, the fix does not come from updating your own page.
• No clear attribution: AI Overviews cite sources, but which sources are cited and how much weight they carry remains opaque.
• Orphaned metrics: the organic search acquisition funnel becomes untraceable in the same way.

Who This Affects Most

Not all sectors are equally impacted. The most exposed are those where informational search was the primary entry channel: media outlets, e-commerce for generic categories, professional services (legal, healthcare, finance) and any business whose value proposition required the user to reach the website to understand it.

Those with more room to manoeuvre are those already operating with strong brand search, their own communities or direct channels. For them, the change stings less in the short term, though the trend affects them equally.

What the Sharpest Teams Are Doing

Some approaches gaining traction, though none yet represents a consolidated solution:

• Optimisation for LLMs: produce structured content with verifiable facts and clear citations that models can extract accurately. It is not classical SEO, but it shares some logic.
• Monitoring mentions in AI: emerging tools that track how different models respond to queries related to a brand. Still rudimentary, but the space is growing fast.
• Channel diversification: reduce dependence on organic search as a primary source and strengthen newsletters, communities and direct presence.
• Verifiable first-person content: interviews, proprietary data, original research; the kind of content models tend to cite because they cannot generate it themselves.

A Note of Perspective

It is worth avoiding catastrophism. SEO has spent decades adapting to algorithm changes, and every time Google adjusts its rules, predictions of total collapse emerge that do not fully materialise. What does seem true is that the adaptation cycle this time is deeper because it affects the logic of the link as a unit of value, not a ranking factor within a system that remains otherwise unchanged.

At ElephantPink, we have spent months watching content and marketing teams start asking questions that were once the exclusive territory of those working with LLMs. The fact that these conversations have reached mainstream reference podcasts like Equity suggests the topic has already moved beyond technical niches. Welcome to the problem.