What 20+ Search Leaders Make of Google I/O 2026 and the AI Search Shift
Google I/O 2026 turned AI into the default layer of Search rather than a feature sitting on top of it. The reactions came fast, and most of them landed on LinkedIn before they landed anywhere else. We read through the breakdowns, the newsletters, the videos, and the published analyses from the people the industry actually listens to, and pulled the signal into one place. More than twenty SEO and GEO leaders, what they agree on, where they split, and the throughline an enterprise team should act on.
The short version. Almost nobody is arguing the announcements are minor. The argument is about what to call the work that comes next, and whether the click economy has enough left in it to keep funding the old playbook. We covered the operator's response in our 90-day playbook on the I/O reset. This post is the industry's response.
What Google announced
| Announcement | What it does | Rollout |
|---|---|---|
| Gemini 3.5 Flash in AI Mode | New default model for AI Mode worldwide. AI Mode has passed one billion monthly users, with queries more than doubling every quarter. | Live, all AI Mode markets |
| Rebuilt search box | Biggest change in over 25 years. Expands for conversational prompts and accepts text, images, files, videos, and Chrome tabs as input. | Rolling out globally |
| AI Overviews and AI Mode fused | A follow-up question from an Overview flows into a conversation with AI Mode, carrying context across the turn. | Live, desktop and mobile |
| Search agents | Information agents monitor the web 24/7 without a new query, and agentic booking expands to local services. | AI Pro and Ultra, US, summer 2026 |
| Generative UI and mini apps | Agentic coding lets Search assemble custom dashboards, trackers, and mini apps on the fly. | US first, coming months |
| Personal Intelligence | AI Mode connects Gmail, Photos, and soon Calendar, so two users can see very different answers for the same query. | ~200 countries, 98 languages |
Search is AI search now, and the leaders heard it
Elizabeth Reid, VP of Search at Google, set the framing the whole industry reacted to. She told the keynote that AI is no longer a layer on Search, that "Google search is AI search through and through." That single line did more to move the conversation than any feature demo.
Source: https://blog.google/products-and-platforms/products/search/search-io-2026/
Marie Haynes, founder of Marie Haynes Consulting, published the most detailed practitioner breakdown the day after the keynote, and her read was that the impact splits the field rather than hitting everyone equally. Some sites lose, she argued, while the ones "on the cutting edge in creating agentic experiences, creating tools that agents can use" stand to gain. Her sharpest point was structural, that SEO has always optimized for the single moment a user types a query, while background agents force teams to optimize for continuous monitoring instead.
Barry Schwartz of Search Engine Roundtable gave the community its factual baseline, flagging that the model swap was immediate rather than a future promise, with Gemini 3.5 Flash powering AI Mode as the new default from day one alongside the agentic features.
Source: https://www.seroundtable.com/google-search-gemini-3-5-flash-agents-41346.html
The fault line, is this still SEO or a new discipline
Mike King, founder and CEO of iPullRank, drew the hardest line. Reacting to Google's framing that AI search optimization is simply still SEO, published the same week, he called the guidance "naive and self-serving" and warned the industry against accepting a role as the "janitors of the web." His argument is organizational as much as technical, that folding AI search back into SEO keeps teams underfunded for the skills the work actually demands, from retrieval theory to passage engineering to agent protocols.
Source: https://ipullrank.com/google-ai-search-guidance
Lily Ray, VP of SEO Strategy and Research at Amsive, holds the steadier counterweight, that the rise of AI search does not mean SEO is dead and that the same fundamentals still decide who wins. She has been consistently skeptical of most GEO and AEO hacks, including aggressive chunking and llms.txt, favouring original content, demonstrated expertise, and clean technical work.
Source: https://www.linkedin.com/pulse/rise-ai-search-does-mean-seo-dead-lily-ray-g2lkf
Garrett Sussman of iPullRank called out how noisy the discourse has become, arguing the AI search field is full of hot takes and that strategy has to be grounded in how retrieval systems actually select content rather than in whatever trends on the feed.
Source: https://www.linkedin.com/posts/garrettsussman_the-ai-search-seo-industry-is-filled-with-activity-7414315459493281792-qYPB
Matthew Proctor, CEO of Narrative Bent, offered the most pragmatic read on the SEO versus GEO fight, that whatever the discipline ends up being called, the work keeps landing on the people already responsible for search. The job-title debate matters less than who owns the outcome.
Source: https://ppc.land/mike-king-googles-ai-search-guide-serves-platform-over-the-open-web/)
Mordy Oberstein of SE Ranking frames the moment as old SEO giving way to new SEO and an era of refinement rather than a clean break, with the brand and entity signals that always mattered now carrying more weight, not less.
Source: https://www.linkedin.com/in/mordyoberstein/
Traffic stops being the scoreboard
Aleyda Solis, founder of Orainti and SEOFOMO, has been the clearest voice on measurement, urging teams to stop treating traffic as a reliable KPI and to move to a blend of brand and performance signals, AI visibility, sentiment, purchases, and revenue. She also documents how the SERP itself is being re-monetized, with paid results capturing a growing share of the clicks that remain.
Source: https://www.aleydasolis.com/en/search-engine-optimization/serp-shifts-ads-remonetized/
Rand Fishkin, co-founder of SparkToro, sees I/O as confirmation that his long-running zero-click thesis is now the baseline rather than the warning. The move he keeps pushing is to optimize for brand and for being the answer people remember, not for a click that increasingly never comes.
Source: https://sparktoro.com/blog/
Kevin Indig, author of Growth Memo, distilled his AI Mode research into one reframe, that visibility is replacing traffic as the real currency. Users stay inside AI Mode, the clicks vanish, and being retrieved and cited becomes the measurable goal.
Source: https://www.growth-memo.com/p/state-of-ai-search-optimization-2026
Tom Critchlow of Raptive has been mapping the downstream economics, especially for publishers facing traffic and licensing pressure as AI search scales, and his call is for the industry to build the new playbook together rather than each team guessing in isolation.
Source: https://www.linkedin.com/posts/tomcritchlow_leveling-up-seo-for-the-ai-era-tom-critchlow-activity-7386028367185358849-Auba
Nikita Vlasyuk, CTO of DeepSEO, brought the hard numbers to the post-keynote discussion, pointing to AI Overviews past a billion users, more than half of Google searches ending without a click, and Gartner's forecast that traditional search volume keeps falling.
Source: https://ppc.land/mike-king-googles-ai-search-guide-serves-platform-over-the-open-web/
The voices saying do not overcorrect
Cyrus Shepard, founder of Zyppy SEO, is the useful corrective to the everything-has-changed narrative. His analysis finds that classic ranking strength remains one of the strongest predictors of being cited in AI answers, second only to whether the URL is accessible to crawlers. Abandoning rankings would be a mistake.
Source: https://zyppy.com/
Glenn Gabe of G-Squared Interactive keeps warning about what he calls Mt. AI, the pattern of sites that spike on thin AI content and then collapse, and his read on durable visibility is consistency, quality, and trust across surfaces rather than chasing the latest tactic.
Source: https://www.gsqi.com/marketing-blog/when-mt-ai-crumbles-chatgpt-follows/
Britney Muller, founder of Orange Labs, put it bluntly, that a lot of so-called AI optimization is the industry rediscovering SEO under a new name, and that real expertise, original work, and sound fundamentals still do the heavy lifting.
Source: https://www.linkedin.com/posts/britneymuller_hot-take-yall-are-just-rediscovering-seo-activity-7378848008609746945-3NR6
Wil Reynolds, founder and co-CEO of Seer Interactive, pushes back on visibility as the goal at all, arguing marketing was never just about being seen, and that the real progression runs from being seen to being believed to being chosen. He points to Seer's own numbers, revenue climbing even as organic traffic fell, as proof that who you reach and whether they trust you matters more than raw sessions.
Source: https://www.seerinteractive.com/insights/the-3-new-kpis-for-ai-search-how-to-measure-brand-performance-in-the-age-of-llms
Nathan Gotch, founder of Gotch SEO, speaks to the volatility most teams feel, that the way to win when the rules keep changing is to over-invest in brand, topical authority, and content genuinely worth citing.
Source: https://www.linkedin.com/pulse/seo-age-ai-mode-how-win-when-rules-keep-changing-nathan-gotch-cspwc
The technical read, optimize for retrieval and structure
Andrea Volpini, CEO and co-founder of WordLift, argues AI search runs on memory, structure, and connection, that structured data directly improves retrieval accuracy, and that search is becoming navigation rather than a list of answers. Entity and knowledge-graph work moves from nice-to-have to load-bearing.
Source:https://www.linkedin.com/in/volpini/
Dan Petrovic of DEJAN approaches the shift through machine learning and information retrieval, focusing on the actual passages and embeddings that systems select, with the practical implication that teams should test and engineer content at the passage level, not just the page level.
Source: https://www.linkedin.com/in/seoguy/
Gianluca Fiorelli of ILoveSEO raises a risk the keynote optimism skips over, that Personal Intelligence and hyper-personalization can create a search echo chamber where familiar brands and strong entities get reinforced while new entrants struggle to break in. Brand familiarity becomes a moat.
Source: https://www.iloveseo.net/ai-search-seo-strategic-framework-2026/
The view from the other side of the market
Jordi Ribas, Corporate VP of Search and AI at Microsoft, is worth including for the contrast he draws. Where Google's guidance plays down GEO, Ribas names generative engine optimization as a real emerging discipline and argues agents are drawn to structured, verifiable content. Several leaders, King among them, point to exactly this gap between the two platforms as the reason a Google-only strategy is now too narrow. We made the same case when Microsoft published the AI search guide Google could not write].
Source: https://ppc.land/mike-king-googles-ai-search-guide-serves-platform-over-the-open-web/
The throughline
Strip away the SEO-versus-GEO terminology fight and the leaders are saying the same thing in different words. The unit of value is shifting from a ranked link to a trusted, retrievable, well-structured brand that an AI system is willing to cite and act on. The optimists and the skeptics even agree on the mechanics, that rankings and technical health still matter as the floor, that structured data and entity clarity are now load-bearing, and that traffic is no longer a clean measure of whether any of it is working.
Where they split is on Google's framing. The "it is still SEO" camp reads the announcements as continuity. The King camp reads them as a platform protecting its own surface while the real work spreads across engines that do not share Google's index. We sit closer to the second camp, and we argued why Google's guidance is not the whole map before I/O confirmed the trajectory.
For an enterprise team, the move is not to pick a side in the naming debate. It is to re-baseline measurement around AI citation and branded demand, keep the ranking and technical fundamentals intact, engineer content for retrieval, and treat visibility across ChatGPT, Perplexity, Copilot, and Gemini as in-scope rather than a side project. The teams that do that early will still be visible when the click-based metrics stop making sense.
---
If you want this turned into a prioritized roadmap for your own domains, with the measurement stack rebuilt around AI citation and the retrieval work scoped against your category, get in touch. We run AI search audits for retainer and project clients.
