TL;DR: My Biggest SMX Advanced 2026 Takeaways
- AI search is exposing weak SEO fundamentals.
- Visibility is a more important metric than rankings.
- Entity clarity matters more than keyword optimization.
- AI citations follow retrieval signals, not just rankings.
- Content freshness needs real updates, not just a new publish date.
- Site architecture and internal linking still matter.
- Many AI crawlers still struggle with JavaScript heavy websites.
- Brand mentions and authority signals influence AI visibility.
- GEO is becoming a process and operations challenge, not just an optimization challenge.
- The future belongs to brands that are easy for both humans and machines to understand.
The theme of SMX Advanced 2026 was pretty clear: AI search is not replacing SEO fundamentals as much as it is exposing which fundamentals were weak all along.
Structure matters. Entity clarity matters. Technical crawlability matters. Freshness matters, but fake freshness does not. Brand authority matters, but not just in a “get more links” way. And maybe most importantly, visibility is becoming bigger than rankings and clicks.
Here are my biggest takeaways.
Purna Virji: Stop Measuring AI Activity
AI Commercialization Strategist Purna Virji opened SMX Advanced with a keynote that felt like a reality check for the industry. The big message: most companies are measuring AI adoption instead of AI impact.
What Stood Out
- Hours saved is not a business outcome.
- More content generated is not a business outcome.
- More workflows automated is not a business outcome.
- Leadership ultimately cares about revenue, growth, efficiency, and competitive advantage.
My Takeaway
The SEO industry has become obsessed with speed. We can research faster, analyze data faster, and produce content faster. But if none of that leads to better visibility, stronger performance, or better business results, it’s hard to justify the investment.
The conversation is shifting from: “Are you using AI?” to “What is AI actually doing for the business?”
Eli Goodman: Visibility Is Replacing Traffic as the North Star
Eli Goodman’s “searchpocalypse” session put numbers behind a trend most SEOs have been feeling for a while: search behavior is changing faster than our reporting tools can keep up with.
What Stood Out
- Zero-click searches continue to grow.
- AI traffic isn’t replacing lost organic clicks.
- More user journeys happen without a website visit.
- Traditional analytics miss many of these interactions.
- AI search experiences are becoming increasingly commercialized.
My Takeaway
I don’t think the story is traffic loss. I think it’s measurement loss. Users can discover your brand, consume your content, trust your expertise, and make buying decisions without ever generating a click. That doesn’t mean traffic stops mattering. It means visibility now includes:
- Clicks
- Citations
- Mentions
- Brand recall
- Branded searches
- AI recommendations
That’s a much bigger measurement challenge than ranking reports were.
Dave Davies: Ranking Is Just the Beginning
Dave Davies presented “Predicting and Influencing AI Citations with Retrieval Signals.” He made the point that ranking is no longer the finish line.
In traditional SEO, the flow was basically:
- Crawl/Index > Rank > Click
In AI search, it is more like:
- Crawl/Index/Base Model Knowledge > Rank > Retrieve > Synthesize > Cite or Mention > Maybe Click
And that last step is becoming less guaranteed.
What Stood Out
Dave talked about retrieval behavior becoming more inspectable, which is important because AI citations have largely felt like a black box. While we still don’t have complete visibility into how AI systems choose sources, there are now several ways to see what information they retrieve, which queries they generate behind the scenes, and whether your content is being surfaced during that process.
He pointed to four visibility windows:
- DevTools: Can reveal retrieval signals such as fan-out search queries, cited URLs, snippets, and source metadata.
- APIs: OpenAI’s Responses API, DataForSEO, and SerpAPI, can provide a more scalable way to inspect retrieval behavior and generated search queries.
- AI Visibility Dashboards: Platforms such as Semrush help track brand and content visibility across AI search experiences.
- Citation Tracking Platforms: Tools like Scrunch focus specifically on monitoring citations, mentions, and source inclusion within AI-generated answers.
He also pointed out 8 retrieval signals AEOs can influence:
- Keep Content Fresh: Update publish dates, refresh statistics, and add current examples when relevant.
- Entity Alignment: Use consistent brand naming across your site, maintain organization schema, author pages, and strong internal linking.
- Topical Coverage: Map query fan-outs and develop content that answers each specific user need.
- Page Structure: Lead with direct answers, use descriptive headings, keep paragraphs concise, and define key concepts clearly.
- Third Party Authority: Earn mentions from sources that AI systems already cite.
- Original Data: Publish original research, surveys, statistics, case studies, etc.
- Author/Source Credibility: Include author bios, credentials, bylines, and links to authoritative profiles.
- Technical Accessibility: Fast server response times (TTFB), clean HTML, no retrieval-blocking JavaScript, and accessible body text.
My Takeaway
One point Dave made about recency windows and content freshness that I think is really important: freshness can’t just mean changing the year in the title. Evergreen content doesn’t need to look like it was published five minutes ago (his example was something like “The Civil War: 2026 Updates,” which obviously makes no sense) But when freshness does matter, AI systems are looking for meaningful, substantive updates, not superficial refreshes.
I think simply updating the year, date, or timestamp without making any meaningful changes is still working in some cases today. But AI systems will get smarter (or that’s the plan?) and if your content isn’t as current, useful, or comprehensive as competing sources, changing the date alone won’t be enough for long. Freshness signals need to be backed by actual improvements. Maybe I’ll test the theory by titling this post “SMX 2027” and see what happens, lol. I rant some more about this in my Content Refresh Strategy for Modern SEO, AEO, and GEO blog.
Dawn Anderson: Architecture Still Wins
Dawn Anderson’s session, “RAG, Context Graphs, and Agents,” explored how AI search systems retrieve, understand, and synthesize information. It was one of the more technical presentations of the conference, but the practical takeaways for organic visibility were surprisingly familiar.
What Stood Out
One concept Dawn discussed was retrieval-augmented generation (RAG), which is the process many AI systems use to find information before generating an answer. Instead of relying solely on what the model already knows, these systems retrieve relevant content, evaluate it, and use it to build a response.
That makes content structure more important than ever.
She also talked about context graphs, which are essentially maps of entities and relationships. AI systems are not just looking for pages about a topic. They are trying to understand what entities are present, how they connect to one another, and whether those relationships are clear enough to be useful during retrieval.
My Takeaway
For organic visibility, that means your site architecture, headings, internal linking, lists, and entity relationships all help establish context. The clearer those connections are, the easier it becomes for machines to understand what belongs together.
While the underlying technology is becoming increasingly complex, many of the recommendations still point back to the same fundamentals marketers have followed for years:
- Create clear content
- Organize information logically
- Establish strong topical relationships
- Make key information easy to find
Will Scott: GEO is an Operations Problem
Will Scott focused less on theory and more on execution. One of his central points was that GEO is not being held back by strategy. Most marketers already understand the broad goals: improve entity clarity, create authoritative content, and increase visibility across the sources AI systems use. The real challenge is operationalizing that work consistently.
What Stood Out
He mentions that AI systems pull information from a fragmented ecosystem of sources, and the overlap between those sources can be surprisingly small. A site that performs well in Google may not automatically become visible in ChatGPT, Claude, Copilot, or other AI-driven experiences.
His framework centered around three concepts:
- Clarity – Make it obvious who you are, what you do, where you operate, and why you’re credible.
- Connections – Reinforce those signals across your content, entities, and supporting evidence.
- Ubiquity – Show up in the places AI systems are likely to retrieve information from.
One example he shared showed the difference between generic marketing copy and specific, verifiable information. AI systems need concrete details they can understand and trust. Locations served, certifications, memberships, reviews, expertise, and other supporting evidence give machines something tangible to reference.
My Takeaway
What I liked most about the session was the emphasis on process. Will repeatedly showed that finding a problem is only half the battle. Once you identify gaps in entity coverage, source visibility, or content quality, you still need a repeatable workflow to assign the work, make updates, review outputs, distribute content, and monitor results.
The biggest takeaway for me was that GEO is becoming less about individual optimizations and more about building systems. The organizations that win won’t necessarily be the ones with the best prompts. They’ll be the ones with repeatable processes for creating, validating, distributing, and monitoring information across the web.
Sam Torres: If JavaScript Hides It, AI May Never See It
Sam Torres of Pipedrive delivered one of the most practical technical organic visibility sessions at SMX. Her main message was simple: many websites are built for browsers and users, but not necessarily for AI crawlers. 69% of AI crawlers reportedly cannot execute JavaScript. As a result, a page can rank well in Google while appearing nearly empty to AI systems.
What Stood Out
- Many AI crawlers still can’t render JavaScript.
- A page can rank well in Google and still appear empty to AI systems.
- Critical SEO elements should exist in server-rendered HTML.
- Rendering is a discoverability decision.
Quick Audit Checklist
- View page source.
- Confirm key content exists in raw HTML.
- Disable JavaScript and reload.
- Test crawler responses with cURL.
- Review AI crawler activity in server logs.
My Takeaway
If your content only exists after JavaScript loads, there’s a decent chance some AI systems don’t know it exists at all. That’s becoming a much bigger problem than most marketers realize.
Tom Capper: Rankings Are Losing Context
Surpriiiiseee…. Rankings are becoming a less important visibility metric.
What Stood Out
Tom Capper presented data from 46,000 keywords showing that rankings no longer tell the whole story, and shared some eye opening statistics:
- The median #1 organic desktop result now appears roughly 635 pixels down the page (a typical laptop browser viewport is around 800-900 pixels tall).
- The median #2 desktop result sits below the fold.
- On mobile, even the median #1 result is frequently below the fold.
This is because search results are increasingly crowded with AI Overviews, ads, shopping results, knowledge panels, People Also Ask boxes, and other SERP features.
My Takeaway
The future metric is not ranking, it’s visibility. Two sites can both rank #1, but one may occupy significantly more screen real estate through rich results, images, FAQs, review stars, product features, or AI Overview inclusion. Not all rankings are created equal.
- Track visibility, not just rank position. Consider how much screen real estate your brand occupies and whether you’re appearing in SERP features, not just where you rank organically.
- Prioritize rich results. The data showed that review stars, FAQs, images, videos, and other enhanced listings significantly increase the size and visibility of organic results.
- Analyze SERPs by intent and vertical. The features pushing down organic listings vary widely by industry and query type, so opportunities differ across markets.
- Pay attention to impressions, not just clicks. Visibility itself creates value, much like advertising. Users may see your brand, remember it, and convert later even if they never click during that search session.
- Invest in brand building. Brand authority is becoming increasingly important for both traditional search visibility and AI-generated recommendations.
Grant Simmons: Earn the Entity
Entity optimization showed up in almost every session. Grant Simmons gave a clear framework.
What Stood Out
Simmons organized his entity optimization recommendations into four practical areas:
- Build an entity map that identifies your core entities (brand, products, services), supporting entities (people, methodologies, certifications), and contextual entities (industry, geography, categories).
- Strengthen on-page signals by clearly naming entities, providing context, backing up claims with evidence, and maintaining consistency across pages.
- Use schema to define entities through key properties like @type, @id, name, sameAs, about, and mentions.
- Structure internal links like a knowledge graph, using hub-and-spoke architectures, entity-focused anchor text, and strong topic clusters.
Simmons also suggested creating an entity inventory that maps 5–10 core entities to dedicated pages, then identifying:
- Content gaps where entities are mentioned but never fully explained.
- Orphan pages that aren’t connected through internal links.
- Supporting entities that strengthen credibility and context.
For content optimization, he recommended a simple four-part checklist:
- Language: Use the entity name in the H1 and early in the page.
- Context: Mention related entities and concepts.
- Proof: Include citations, dates, credentials, outcomes, and other verifiable details.
- Consistency: Use the same names, descriptions, and references across content, metadata, and schema.
One point I appreciated was his argument that schema still matters even if LLMs don’t directly consume it. Schema helps define entities consistently across search engines, knowledge graphs, and retrieval systems, creating a clearer understanding of who your brand is and how it relates to other entities online.
My Takeaway
The session ended with a simple takeaway: if you want better visibility in both traditional search and AI search, focus less on keywords and more on helping machines understand exactly who you are, what you do, and where you fit within your industry’s knowledge graph. As he put it, “Earn the entity. Everything else follows.”
Kyle Risley: Agentic Commerce Looks a Lot Like Good Ecommerce SEO
Kyle Risley of Shopify talked about a future where AI agents help users buy products. Lots of the recommendations sounded like ecommerce SEO best practices.
What Stood Out
His framework for ecommerce AI visibility came down to three pillars:
- Accessibility: Can AI agents access your content?
- Comprehension: Can AI agents understand your products?
- Authority: Do AI systems trust your brand?
Some of the most actionable recommendations included:
- Audit your robots.txt and CDN settings to ensure AI crawlers can access your content.
- Verify that critical product page content can be read without JavaScript.
- Check third-party apps that may be rendering reviews, FAQs, or other important content client-side.
- Test pages using tools like Onely’s “What Would JavaScript Do?” tool or simply disable JavaScript in your browser and see what disappears.
On the ecommerce side, Kyle repeatedly emphasized product data quality. AI systems rely heavily on structured product information, feeds, and page content.
My Takeaways
- Keep product feeds live, synced, and error free.
- Fix Google Merchant Center issues.
- Ensure product attributes, GTINs, brand information, shipping policies, and return policies are complete.
- Include detailed specs, dimensions, features, use cases, FAQs, ratings, and reviews directly on product pages.
- Implement Product Schema markup correctly.
His final point was that AI commerce readiness isn’t really a separate discipline from good ecommerce SEO. The same fundamentals still apply: make your content crawlable, make your products easy to understand, and build enough authority that AI systems feel confident recommending you. The difference is that those optimizations may soon influence not just who gets recommended, but who gets purchased.
So What Should GEOs Actually Do?
After two days of sessions, AI demos, GEO frameworks, and enough acronyms to make my head hurt, my action list is surprisingly simple:
Technical SEO
- Verify important content exists in raw HTML.
- Test key pages with JavaScript disabled.
- Review server logs for AI crawler access.
Content & Authority
- Refresh content with meaningful updates, not just date changes.
- Publish more original research, data, and firsthand insights.
- Strengthen author credibility and expertise signals.
- Fill topical gaps around your most important entities.
Entity Optimization
- Create an entity map for your brand, products, services, and people.
- Improve internal linking between related entities and topics.
- Audit schema implementation and entity consistency.
- Make it painfully obvious who you are and what you do.
Visibility & Measurement
- Stop reporting only on rankings and clicks.
- Track AI citations, mentions, impressions, and branded searches.
- Monitor visibility across Google, ChatGPT, Gemini, Perplexity, and other discovery platforms.
- Pay attention to where your brand appears, not just where users click.
Process & Operations
- Build repeatable workflows instead of chasing one-off AI hacks.
- Establish review processes for AI-generated outputs.
- Create systems for updating, validating, and distributing information.
- Focus on sustainable advantages instead of temporary shortcuts.
If there was one lesson that kept showing up throughout SMX Advanced, it was this: the future of SEO still belongs to organizations with strong fundamentals.
Contact Arc Intermedia for Help
At Arc Intermedia, we’ve been helping brands improve visibility across the entire discovery ecosystem long before GEO, AEO, and SEO became separate conversations. If you’re evaluating how your brand appears across Google, ChatGPT, Gemini, Perplexity, and other search experiences, learn more about our Organic Search Visibility Services or contact our team for a free consultation.
