Google recently published an official guide on optimizing for generative features in search. The industry’s reaction was immediate and polarized. The reason: Google came out and said, plainly, that AEO (Answer Engine Optimization) and GEO (Generative Engine Optimization) are not new disciplines. They are, in Google’s words, just SEO.
That single statement set off a wave of conversation on LinkedIn and beyond. On the latest episode of SEOTalk Spaces, host Malhar Barai and co-host Parth Suba brought together practitioners Krinal, and Shweta to dig into what Google actually said, what it conspicuously left out, and whether the SEO industry overcomplicated AI search or simply responded rationally to a very real shift.
Here is a distilled version of that conversation.
What Google Actually Said (and What It Didn’t)
Google’s guide defines AEO as Answer Engine Optimization and GEO as Generative Engine Optimization, both framed as approaches to improving visibility in AI-powered search. The core position: optimizing for generative AI search is the same as optimizing for classic search. Same fundamentals. Same game.
But as Parth pointed out, Google was careful about what it chose not to say. It did not say nothing has changed technically. In the same guide, Google introduced concepts like RAG (Retrieval Augmented Generation) and “query fan out,” a process where AI models generate a set of concurrent related searches to answer a single query. The machinery is new. The optimization principles, per Google, are not.
Malhar’s read: “They are not ready with their building blocks, and hence they still don’t want to come out and say here are some new principles you need to play by.”
The diplomatic framing may be strategic rather than fully informational.
Did SEOs Overcomplicate This, or Were We Just Responding to Real Hype?
The honest answer from the panel: a bit of both.
Krinal observed that what has changed most is the audience paying attention. Investors, CX professionals, C-suite executives who never gave SEO a second look are now deep in conversations about AI search. Perplexity is running ads targeting SEOs. Brands are spending millions on AEO messaging on physical billboards in the Bay Area. The hype is real, and it is coming from outside the industry.
At the same time, that expanded attention has created a genuine opportunity. Krinal argued that AEO is less a new technical discipline and more a window for SEOs to elevate their seat at the table.
“AEO is this window of opportunity for us as SEOs to get into that seat. Whether we take it or not, how much we are able to embrace it, is all up to us.” say Krinal
The reframe: if you walk into a CMO’s office saying “I do SEO,” you get a certain kind of budget and a certain kind of resourcing. If you walk in saying “I lead our AEO and GEO efforts,” the conversation changes entirely, even if the underlying work has meaningful overlap.
What Has Actually Changed in Practice
The panel pushed past the positioning debate to get specific about what practitioners are actually doing differently.
Brand and topical authority matter more now.
Malhar noted that AI search has moved beyond domain authority as the primary signal. Brand authority and topical authority both carry more weight in how LLMs surface and cite content.
The content distribution map has expanded.
Parth and Krinal both pointed to the sources LLMs draw from: Reddit, Quora, Medium, YouTube, G2 reviews, Wikipedia, affiliate sites, editorial placements on high-authority domains. Being present and credible across those surfaces is no longer optional for brands that want AI visibility.
Measurement has shifted from analytics to share of voice tools.
Parth described moving away from a near-exclusive reliance on Google Analytics and Search Console toward platforms that measure LLM share of voice. The data inputs for strategy have changed even where the strategic goals remain similar.
Query fan out demands holistic content.
Shweta offered a clear practical example. A user searching for running shoes in an LLM is not typing “running shoes.” They are describing their context: training for a marathon, specific terrain, specific goals. Content written around a single keyword will not surface there. Content written around the full problem space, the benefits, the trade-offs, the use cases, has a better shot. “It’s not about stuffing keywords,” she said. “It’s about writing for user intent at a much deeper level.”
Cross-platform content consistency is a ranking signal.
Shweta also flagged something she had validated through testing: consistent messaging across YouTube, Instagram, and web properties tends to perform better in LLM results than fragmented or platform-siloed content.
The Enterprise Dimension: Testing Over Belief
One of the sharper contributions to the discussion came from the enterprise SEO perspective. The core insight: enterprises, because they have the budget, are less concerned with debating Google’s official guidance and more focused on running experiments.
One experiment that produced measurable results: adding buttons that allow users to summarize blog content via ChatGPT. The result was a 4x increase in LLM crawler activity, along with measurable lifts in citations and brand visibility. The test became a permanent change.
The bigger challenge at the enterprise level is attribution. If AI search drives someone to recognize your brand, but they then execute a branded Google search and convert through a paid campaign, the SEO effort that built that AI visibility gets no credit. That attribution gap is one of the defining unsolved problems in the space.
There is also a structural gap that the enterprise view surfaced as the clearest difference between SEO and GEO: agents. Bots crawl and leave. Agents come to your website and take actions. They can book flights, complete forms, initiate purchases. The question for GEO is not just whether your content surfaces in an AI response but whether your website is architected for agents to actually do something when they arrive. That is a capability SEO never needed to build for.
What Should New Practitioners Do?
The question of where to start came up repeatedly, and the panel’s collective answer can be summarized in a few principles:
Build the fundamentals first.
Backlinks, on-page SEO, solid site architecture: these are not outdated. They are the foundation GEO is built on. Krinal was explicit: do not jump straight into Claude Code or AI-assisted pipelines before you understand how to read Search Console data, how to interrogate GA4, how to integrate both data sets and understand what they mean.
Learn how LLMs actually retrieve and weight content.
Understanding RAG in plain terms matters: LLMs pull from external sources at query time to improve freshness and accuracy. Your content needs to be the kind of content worth pulling.
Experiment relentlessly, and do not trust anyone who says they have cracked it.
Shweta was direct: “This is a time when everyone is experimenting. If someone says they have cracked it, something is being sold.” The field is moving fast enough that even well-resourced practitioners are running tests, not executing settled playbooks.
Think about agentic search, not just human search.
The next wave is not just AI summarizing search results for humans. It is AI agents executing research and making decisions autonomously, with humans entering the loop later. Preparing for that means thinking about how an agent reasons through context, not just how a human phrases a query.
On llms.txt and Google’s Specific Guidance
A pointed question in the session: Google explicitly said not to use llms.txt and not to chase inauthentic mentions. Should practitioners follow that?
The panel’s consensus was measured skepticism. Krinal’s take: “Never blindly follow what Google says. Always experiment on the campaign you are working on.” The llms.txt implementation takes an hour or two. Run it. See if it moves anything in your context. Google’s guidance is a starting point, not a ceiling.
The broader dynamic has not changed: it is still a Tom and Jerry relationship between SEO practitioners and Google’s search engineers. The recent episode of a planned one-on-one with Google engineers at Google I/O being cancelled at the last minute was a fitting metaphor for how that dynamic tends to play out.
The Debate Will Continue
The session closed on a note of realistic expectation-setting. The subdomain vs. subdirectory debate has been running for two decades and shows no signs of resolution. The AEO vs. SEO debate is unlikely to settle on a shorter timeline.
What is clear is that the fundamentals of earning visibility, building trust, and creating content that genuinely serves user needs have not changed. What has changed is the infrastructure through which that visibility is delivered, the signals that influence it, and the scale at which practitioners need to think about brand presence across platforms they have never had to optimize for before.
Whether you call it AEO, GEO, or SEO with a wider aperture, the practitioners who will navigate this era well are the ones running experiments, building cross-platform authority, and keeping their eye on the revenue conversation rather than the terminology one.
SEOTalk Spaces is a weekly community conversation hosted by Malhar Barai and Parth Suba. Episode 131 featured contributions from Krinal, Shweta, and community members from the SEOTalk audience.