Generative Engine Optimization: How to Show Up in AI Search

A client emailed me a screenshot last month. She'd asked ChatGPT to recommend companies in her category, and three of her competitors came back with neat little summaries. She wasn't in the list. Not wrong, not misranked. Just absent. Her first question was the one I get most these days: what is generative engine optimization (GEO), and how do I actually do it?

The short answer is that GEO isn't a separate discipline from SEO. There's a new set of surfaces, a new way people get answers, and a pile of new acronyms. Underneath all of it is the same job search has always been: make it easy for a system to find your stuff, understand who you are, and decide you're worth quoting. That job now includes AI assistants. It didn't stop including Google.

This piece is the map. I'll define the terms, walk through where AI answers actually come from, and lay out the strategic levers that move the needle. For the how-to on each lever, I'll point you to the deeper write-ups rather than cram everything into one wall of text. If you'd rather start by checking whether you have a problem at all, skip ahead to how to tell if your site is invisible to AI search. That post is the diagnostic. This one is the strategy that sits above it.

What is generative engine optimization (GEO)?

Generative engine optimization (GEO) is the practice of getting your content found, understood, and cited inside AI-generated search experiences. The surfaces include Google's AI Overviews and AI Mode, ChatGPT, Perplexity, Gemini, and Microsoft Copilot. The goal is to show up when one of those systems composes an answer for someone in your market. The term was coined in a paper from researchers at Princeton, Georgia Tech, the Allen Institute for AI, and IIT Delhi, first posted to arXiv in November 2023 and presented at the KDD 2024 conference. That paper defined the optimization target as visibility inside generative engine responses rather than a list of blue links. That's the clean academic definition. In June 2026, Google formalized the term in its AI Optimization Guide and made the key point: GEO is still SEO, because its AI features run on core Search ranking and quality systems.

You'll also see AEO, answer engine optimization. The two terms overlap so much that people use them interchangeably. AEO leans toward answering a question cleanly enough to be surfaced and reused. GEO leans toward showing up inside a generated response. In day-to-day work the distinction rarely matters.

Here's the part that cuts through the noise. In June 2026, Google's AI Optimization Guide said it directly: "'AEO' stands for 'answer engine optimization' and 'GEO' for 'generative engine optimization'... optimizing for generative AI search is optimizing for the search experience, and thus still SEO." The same doc adds that "the best practices for SEO continue to be relevant because our generative AI features on Google Search are rooted in our core Search ranking and quality systems."

That's the thesis of this whole piece, straight from the source. GEO is not a parallel discipline you bolt onto SEO. It's SEO, with the surfaces expanded to include AI answers and the measurement expanded to match. When I talk about what I do, I don't split search visibility from AI visibility. They're one job. The levers overlap almost entirely.

Why does showing up in AI search matter now?

Showing up in AI search matters now because a growing share of search ends inside an answer instead of on your website. AI is no longer a sidebar to search behavior. It's increasingly the front door. Start with how often Google itself injects an AI answer. The Semrush AI Overviews study, published December 2025 and built on more than 10 million keywords, found AI Overviews appeared in 6.49% of searches in January 2025, peaked at 24.61% in July 2025, and settled to 15.69% by November 2025. Prevalence is volatile and varies by tracker, so read those as snapshots, not a fixed rate. The direction is the point: a meaningful slice of Google results now opens with a generated answer. Google's AI Overviews are available in over 200 countries and more than 40 languages as of May 2025.

And it's spreading into commercial territory. The same Semrush study found commercial queries triggering AI Overviews grew from 8.15% in January 2025 to 18.57% in October 2025. Navigational queries jumped from 0.84% to 10.33% over a similar stretch. People searching with intent to buy, or to find a specific brand, are increasingly meeting an AI answer first.

The behavior shift shows up in clicks. A SparkToro analysis of Similarweb data found 68.01% of U.S. Google searches ended without a click to any website in the first four months of 2026, up from 60.45% in 2024. For every 1,000 U.S. Google searches, only 276 clicks now reach the open web. If your strategy depends entirely on the click, that's a shrinking base.

Then there's the traffic happening off Google entirely. ChatGPT reached 900 million weekly active users as of February 2026 per OpenAI's own reporting. Not all of those are commercial searches, so don't read it as 900 million buyers. But Cloudflare network data reported in Q1 2026 put AI search visits at 27.4 billion, up 42.8% year over year from 15.6 billion. People are asking AI assistants the questions they used to type into a search box.

Here's the honest counterweight, because I'd rather you plan with real numbers. Conductor's 2026 AEO/GEO Benchmarks Report, published April 2026 and built on 3.3 billion sessions across 1,215 enterprise domains, found AI referral traffic was only 1.08% of total web visits. That looks small. But it understates the value, because AI-cited content often drives a brand impression without a click at all. The point of showing up in an AI answer frequently isn't the referral. It's being named as the answer.

Where does AI search pull its answers from?

Every AI search surface has to do the same two things: retrieve candidate sources and decide which ones to cite. Most of them lean on a technique called retrieval-augmented generation, where the system pulls relevant pages and grounds its answer in them rather than relying purely on what the model memorized. The catch is that each platform retrieves from a different place and gates access with a different crawler.

Before the surface-by-surface breakdown, the single most important fact: for Google's AI features, there is no separate AI index. The gate is the same one it has always been. Per Google's AI Features documentation, updated December 2025, "to be eligible to be shown as a supporting link in AI Overviews or AI Mode, a page must be indexed and eligible to be shown in Google Search with a snippet." There is, in Google's words, nothing exotic required. If you're indexed and snippet-eligible, you're in the running.

AI surface	Where it retrieves from	The access gate
Google AI Overviews & AI Mode	Google's own index, via retrieval plus a "query fan-out" that issues multiple related sub-searches	Indexed and snippet-eligible in Google; Googlebot allowed
ChatGPT Search	Bing's index plus OpenAI's own crawl for freshness	OAI-SearchBot allowed in robots.txt; content in static HTML (no JavaScript rendering)
Perplexity	Real-time web search across multiple APIs plus its own large index	PerplexityBot allowed in robots.txt
Gemini	Google's index, grounded in real-time Search and Knowledge Graph entity resolution	Indexed in Google; Googlebot allowed
Microsoft Copilot	Bing's index, via retrieval-augmented generation	Standard Bing indexing; Bingbot allowed

A few things are worth pulling out of that table. Google's AI features and Gemini both run on Google's index, so the work that earns you standard Google visibility is the same work that makes you eligible for both. Both AI Overviews and AI Mode use query fan-out, issuing multiple related searches across subtopics to build a response, which is why chasing a single exact-match phrase rarely works for AI visibility.

ChatGPT and Copilot both lean on Bing's index. One important distinction on the ChatGPT side, from OpenAI's crawler documentation: OAI-SearchBot is the live search retrieval crawler, while GPTBot is the training-data crawler. Blocking GPTBot does not block ChatGPT Search. Blocking OAI-SearchBot does. People mix those up constantly and accidentally wall themselves off. Perplexity has its own gate too; Perplexity's documentation says allowing PerplexityBot is recommended to ensure your site appears in its search results.

What are the strategic levers to show up in AI search?

Here's the part that makes the landscape navigable. Five surfaces, several crawlers, plenty of acronyms, but the underlying logic is simple. To show up in AI answers you need to clear five levers: be findable (indexed and crawlable, with AI crawl bots allowed in robots.txt), be recognizable (a clear entity AI systems can attribute claims to with confidence), be worth citing (original, experience-based content that answers real questions), be present off-site (earned mentions across the open web, where LLMs build brand associations), and be current (fresh content, especially for ChatGPT and Perplexity). Google's June 2026 AI Optimization Guide confirms the core of that list: the gate for its AI features is simply being indexed and snippet-eligible in Search. Below is each lever, what it actually means, and where to go for the deep dive.

Lever	What it does	Go deeper
Be findable	Get indexed, snippet-eligible, and crawlable, with AI crawlers allowed in robots.txt	The invisibility diagnostic
Be recognizable	Make your brand a clear, consistent entity so AI systems can attribute claims to you with confidence	Entity confusion; entitymap; schema
Be worth citing	Publish clear, original, experience-based content that answers real questions	Google's content guidance
Be present off-site	Build earned mentions across the wider web, where LLMs learn brand-to-topic associations	LinkedIn AI citations
Be current	Keep key pages fresh, which several AI platforms reward (Google AI Overviews is the exception)	Covered below

Be findable: indexed, snippet-eligible, crawlable

This is the floor, and it's where most invisibility actually comes from. If Google can't index a page, it can't appear in AI Overviews. If you've blocked OAI-SearchBot, you won't show up in ChatGPT search answers. Most of the "we're invisible in AI" problems I see trace back to something boring at this layer: a page that never got indexed, a crawler quietly blocked in robots.txt, important content trapped behind JavaScript that ChatGPT's crawler doesn't render.

I'm not going to run the full checklist here, because there's a post that already does. If you want to actually check your situation step by step, the invisibility diagnostic covers indexation, the robots.txt audit, and what to do when AI engines can't see you. That post is the "how do I check and fix it." This section just tells you it's the first lever to pull.

Be recognizable: entity clarity

AI systems resolve who you are before they decide whether to cite you. Gemini, for instance, leans on Google's Knowledge Graph to understand entities before retrieving candidate sources. If your brand isn't a clear, consistent entity across your own site and the broader web, the system has lower confidence attributing a claim to you, and lower confidence means fewer citations.

This shows up as two different problems. One is AI getting your business flat wrong, mixing you up with another company or stating something outdated. That's entity confusion in AI search, and it has its own post. The other is the structural plumbing that defines your entity in the first place: a consistent entity definition across sources, and clean structured data that doesn't contradict itself. There's a proposed standard for declaring entity relationships called entitymap, and there's the practical reality that adding schema shouldn't be a dev project.

One honest caveat, because the schema-for-AI claim gets oversold. Google's June 2026 guide states plainly that "structured data isn't required for generative AI search, and there's no special schema.org markup you need to add." Schema still earns rich results and helps machines parse your pages, so it's worth doing. Just don't believe anyone who tells you adding Organization schema is a guaranteed switch that turns on AI Overviews. It's an entity-clarity signal, not a magic gate.

Be worth citing: clear, original content

This is the lever Google is most explicit about. Its guide tells you to "create the content yourself based on what you know about the topic" and to avoid recycling "what others on the internet have already said, or could easily be produced by a generative AI model." It asks for "non-commodity content that your readers will find helpful and reliable." Structure matters too: people, in Google's words, "appreciate it when web pages are organized by paragraphs and sections, along with headings that provide a clear structure."

What this lever does not require is the stuff vendors love to sell. Google's guide says you don't need to create AI text files, you don't need special markup, and you don't need to write "in a specific way just for generative AI search." That includes llms.txt files, which Google Search itself doesn't use. The honest version of this lever is unglamorous: write genuinely useful things only you could write, and structure them so a machine can extract the point cleanly.

Be present off-site: earned mentions across the web

This is the lever that reframes how most people think about AI visibility, so I'll spend a beat on it. In traditional SEO, your website is the center of gravity. In AI search, it isn't.

Ahrefs studied 75,000 brands and published the correlations in December 2025. Web brand mentions showed the strongest correlation with AI Overview brand visibility at 0.664 Spearman, nearly three times the correlation of backlinks at 0.218. YouTube mentions had the single highest correlation of any factor measured, 0.737, across ChatGPT, AI Mode, and AI Overviews. The top three correlated factors were all off-site. That's correlation, not causation, and the study skewed toward established brands. But the direction is hard to ignore: where your brand gets mentioned across the open web matters more than what you publish on your own domain.

Which surfaces? Semrush analyzed 100 million-plus citations in a study published November 2025 and found a handful of domains dominate: YouTube, Reddit, Wikipedia, and LinkedIn rank among the most-cited across the major platforms. A separate Semrush LinkedIn study, published March 2026 and built on 325,000 prompts, found LinkedIn was the second most-cited domain overall, referenced in about 11% of AI responses on average. Notably, 95% of cited LinkedIn posts were original content, and the median cited post had just 15 to 25 reactions. Reach isn't the lever. Original, consistent posting is.

The how-to for the off-site lever, at least on the LinkedIn side, lives in how LinkedIn drives AI citations for B2B. The strategic takeaway here is simpler than the tactics: LLMs can only learn that your brand is associated with a topic if content connecting the two actually exists somewhere they can read it. Build that presence and you build citability. One honest note from Google's own guide, though: "Seeking inauthentic 'mentions' across the web isn't as helpful as it might seem." Earned and real, not spammed and fake.

Be current: freshness, with one big exception

AI assistants tend to favor fresher content, but not uniformly, and the nuance matters enough that I won't let you take it as a blanket rule. Ahrefs analyzed nearly 17 million citations and published the findings in July 2025: AI assistants cited content averaging 1,064 days old versus 1,432 days for organic results, about 25.7% fresher. ChatGPT showed the strongest pull toward new content, citing pages averaging 458 days newer than organic.

The exception is the one most "keep it fresh" advice skips. Google AI Overviews showed essentially no freshness premium. It cited content nearly as old as standard organic results. So for Perplexity and ChatGPT visibility, keeping key pages updated genuinely helps. For Google AI Overviews specifically, fresh content helps your standard SEO but there's no documented AI freshness bonus. Plan your refresh cadence with that split in mind rather than chasing freshness everywhere.

How do you know if it's working, and where to start?

Start by finding out whether you have a problem, because the levers above are wasted effort if you're already showing up fine, or if your real issue is something specific like a blocked crawler. The honest first move is diagnosis, not a sweeping rollout. The cheapest check costs nothing: open ChatGPT, Perplexity, Gemini, and a few Google searches that trigger AI Overviews, and run the prompts your buyers would actually use. See whether you appear, and whether what's said about you is correct. Absent is one problem. Wrong is a different one, and it routes to the entity confusion post. Then confirm the technical gate: your pages need to be indexed and snippet-eligible in Google, and AI crawlers need to be allowed in robots.txt. For the full structured walk-through of what to check and in what order, the invisibility diagnostic is the companion piece to this map.

On the measurement side, there's a real new tool worth knowing about, with a real limit. Google announced a Generative AI performance report in Search Console in June 2026. It surfaces impressions from AI Overviews and AI Mode. The honest caveat, in Google's own words, is that "not all properties have access to the report, as we're rolling out over time." So it's promising, but you may not have it yet, and even when you do, it covers Google's surfaces only, not ChatGPT or Perplexity. Measuring AI visibility well still means combining a few imperfect signals, not reading one clean dashboard.

One thing I'll be straight about: this field moves fast, and any specific number here has a shelf life. The mechanics are stable. Indexed and snippet-eligible has been the Google gate for a long time and probably stays that way. The prevalence percentages and platform rankings shift month to month, which is exactly why I've date-stamped them. If you're reading this a year from now, trust the levers and re-check the figures.

If you're trying to figure out where your search and AI visibility actually stands, and what to do first, that's the kind of question an engagement is built to answer. It's one discipline, not a menu of separate AI add-ons.

GEO isn't a new discipline you bolt onto SEO. It's SEO, aimed at every surface where people now get answers.