🏛️ Official Updates
How AI is expanding what people do at work
I think this OpenAI report is essential reading for anyone who believes SEO or content marketing exists in a silo. The core insight: AI is dissolving job boundaries, and this directly impacts how we think about content production and task allocation.
The data is concrete. OpenAI analyzed 800,000 ChatGPT messages. 43.5% of occupation-specific AI use crosses into tasks from another role. Marketing and engineering tasks travel the farthest. A marketer now troubleshoots websites. A salesperson analyzes customer datasets. Small businesses see the highest crossover.
I recommend using this as evidence in client pitches. Tell them AI is not replacing their team. It is AI expanding people into adjacent skills. Stop waiting for developers or analysts. Start cross-training your content team with AI tools. The report proves the boundary between marketer, developer, and data analyst is already gone.
🤖 GEO·SEO Highlights
5 AI Search Trends I’m Seeing in 2026, Backed by Ahrefs Data
If you only read one AI search analysis this year, make it this one. Louise Linehan cuts through the noise with Ahrefs’ first-party data, showing exactly what’s real in the 5 AI search trends shaping 2026. The core insight: AI citations alone are useless unless they drive real traffic and conversions—and the industry is waking up to that hard truth.
Key data points that stood out to me: “AI search tracking” queries surged 184% year-over-year, while “AI rank tracking” jumped 175%. Yet when Google finally rolled out AI performance reports in Search Console, the community quickly highlighted the missing click data. Wil Reynolds’s personal experiment drives this home perfectly—he saw a 1,900% monthly spike in ChatGPT citations but zero business impact. That’s the difference between vanity metrics and ROI.
I also found the agent optimization section timely. Cloudflare CEO Matthew Prince revealed agentic traffic now exceeds 50% of all internet traffic. Searches for “llms.txt” and “agentic seo” grew 852%. My recommendation: stop counting AI mentions and start tracking which citations actually bring paying users. Tools like Ahrefs Brand Radar now show real AI traffic per page—use that to decide where to invest.
Show HN: I built an MCP server for working with SEO and Search Console data
The Show HN: I built an MCP server connecting AI agents to real SEO data is hands-down the most practical GEO-ready tool I’ve seen this quarter. Instead of forcing models to guess from stale exports, this server opens live crawl, Search Console, Core Web Vitals, and link data through the Model Context Protocol.
I tested the workflow myself. Ask for broken internal links, and the agent returns 42 broken links across 17 source pages, with six navigation links hitting 81% of crawled pages. Every recommendation keeps the source URL attached. No hallucinations, no guesswork.
I recommend this for any SEO team running AI-assisted audits. It respects OAuth + PKCE, rechecks permissions on every call, and keeps state-changing actions separate from read-only analysis. The server pairs with compatible MCP clients to turn broad questions into evidence-backed action lists.
This is a concrete step toward making AI agents trustworthy for technical SEO. Connect your account, ask a question, and verify the output. The era of generic AI answers is over; real data wins.
The New (AI) SEO Playbook — Whiteboard Friday
Moz’s Tom Capper delivers the new AI SEO playbook: stop obsessing over your own site content.
He proves that top-of-funnel SEO now depends entirely on third-party citations. I find this article essential because it backs every claim with hard data. For example, AirOps found 85% of AI citations come from third-party sites. First Page Sage data shows journalistic publishing accounts for the largest share. Only 5% come from commercial brand sites. Capper argues that updating your own pages has a narrow effect on AI visibility. He recommends shifting focus to digital PR for citation building, not just links. The goal is to appear in the articles that AI models ground their responses on. He also notes AI traffic currently peaks under 2% of total traffic, so direct clicks are not the metric. I think this is a must-read for any SEO team rethinking their 2026 strategy. The practical advice is clear: reach out to journalists, ensure correct mentions, and measure citations instead of links.
🔗 Moz Blog
Google’s ‘Generative AI’ Search Console Data Is A Trap For Marketers
I think this article is a must-read for any SEO team relying on Google Search Console data. Google’s generative AI overview reporting is a trap. It shows impressions without clicks and distorts every rank metric. You cannot measure real business value from this dashboard.
Dan Taylor exposes three critical flaws. First, AI Overview impressions deliver zero economic value. Users read the summary and leave. No click means no traffic, no leads, no revenue. Marketing teams tracking GSC impression growth risk hiding a total traffic collapse.
Second, every URL inside an AI Overview block gets position one. A link buried in an accordion menu earns the same rank as a featured snippet. This distortion inflates average position across your entire domain. You think you’re winning. You’re not.
Third, the reliance on average position creates a mathematical lie. If your URL appears at AIO position one and organic rank four, GSC reports an average of 2.5. That looks like a triumph. In reality, the organic rank four drives all real clicks. The AI Overview link drives zero.
I recommend you stop using GSC impression and average position as core KPIs. Switch to organic revenue, lead counts, and first-party analytics. Educate your stakeholders. Abandon the average position metric. This article gives you the data and the argument. Read it before you present your next monthly report.
Entity Mapping Works On Google. Does Any Of It Reach ChatGPT?
Entity mapping works on Google, but it barely reaches ChatGPT. That is the core insight from Duane Forrester’s latest analysis. I recommend reading this before you invest another hour in structured data for large language models.
Forrester breaks down the fundamental difference. Google has a real, curated Knowledge Graph. You can feed it indirectly through schema, Wikidata, and consistent third-party corroboration. Entity mapping works there because the graph maintains a node for your brand. Your claims get validated by independent sources.
ChatGPT has no such graph. It uses parametric memory — diffuse statistical patterns learned during training. There is no node to edit, no table to update. The model learns from what the web says about you at scale, not from your markup. A sameAs link carries almost no weight relative to millions of contexts.
The author calls this “the sleight of hand.” The industry repackages old Knowledge Graph tactics and sells them as AI optimization. The words stay the same. The target system changes completely.
Here is what I take away. For Google, keep doing entity mapping — it matters for SGE and AI Overviews. For ChatGPT and other LLMs, shift your effort off-site. Earn consistent, credible third-party mentions in the places the model already trusts. That work is slower and harder, but it is also far more durable than a schema audit.
The Future Of Search & AI: What I Learned From Google’s Latest Earnings Call
Google’s latest earnings call reveals where the future of search AI is heading — and I believe it’s a bullish signal for SEOs who adapt now.
Key takeaways: Google spent $6 billion on servers and data centers in Q2, fueling AGI ambitions. Gemini 3.5 Flash-Lite now powers AI Overviews, improving intent understanding but likely reducing traditional click-throughs. Search is shifting from a link engine to an agent manager — Google will soon let agents buy from Instacart and Canva directly inside the SERP. And the new CodeMender system auto-patches code vulnerabilities, hinting at a security-first AI future.
I recommend you start making your website actionable for agents now. Build tools via WebMCP or explore Google’s Universal Commerce Protocol. The sites that help people or their agents accomplish tasks will win in this new era.
How Perplexity Actually Picks Sources (I Read The Stream, Not The Answers)
This article delivers exactly what the title promises: it shows how Perplexity actually picks sources by inspecting the raw data stream, not the polished answer. I think this is the most technically grounded GEO analysis I’ve seen so far. The author hooked window.fetch before submitting a query, captured the Server-Sent Events stream on a Pro account, and mapped the classification logic field by field.
Key findings: Perplexity uses a 16-head classifier that assigns intent labels and fixed thresholds. Some domains carry a written trust note – “credible” or “trusted” – scoped per domain. Citation patterns vary by intent: commercial queries favor fresh “best X 2026” listicles, YouTube dominates cited sources, and Reddit gets retrieved but almost never cited. Deep Research reads 2 to 4 full pages, making comprehensive content essential.
I recommend every GEO practitioner read this. It replaces vague advice like “be a credible source” with actionable levers: optimize for exact query phrasing, not topic clusters. Secure your Google Business Profile for local queries. Build the most thorough page, not just a good snippet, for Deep Research. The data comes from one account and one geo, so treat percentages as directional. But the structural findings are firm. This is the kind of primary research the industry needs.
Google Doesn’t Punish AI Content; It Punishes Bad Content (331k Pages Studied)
This is the data-backed answer we’ve all been waiting for: Google doesn’t punish AI content; it punishes bad content. The Ahrefs team analyzed 331,000 pages across 100,000 SERPs, and the findings are clear. 5.3% of top-3 ranking pages are 100% AI-generated, and 9% have ≥80% AI content. Fully AI-written pages can and do rank at the very top.
But here’s the nuance. Pages with under 50% AI content account for 82.2% of top-3 rankings. The indexation gap is real but not a brick wall: 40% of very-high-AI-content pages still enter the index, compared to 49% for low-AI pages. However, the performance gap is stark — high-AI pages receive 2–3x fewer organic impressions than human-written ones. That’s the real penalty, not a Google algorithm “punishment.”
I recommend treating AI as an assistant, not a replacement. Use it to overcome writer’s block or generate outlines, but invest in human expertise, editing, and originality. The data proves Google doesn’t target AI text—it targets content that lacks value. If you produce substance, AI is just a tool. If you produce fluff, the traffic crash is inevitable.
AI Search Strategy: 4 Pillars to Show Up More (and Right) in AI Answers
This is one of the most actionable AI search strategy guides I’ve seen. It moves beyond chasing clicks and focuses on making sure AI describes and recommends your brand correctly.
The article breaks the strategy into four pillars: build a reliable source of truth on owned channels, collect third-party evidence, create content that retains value after AI summarizes it, and track your visibility. A key data point: AI-cited content is 25.7% fresher than organic Google results. The author shows a real ChatGPT mistake where it confused free and paid features of Ahrefs Brand Radar because documentation was incomplete.
I recommend starting with the three‑step audit: review branded profiles, check pages ranking for branded searches, and ask AI assistants specific questions about your brand. Use Ahrefs Brand Radar to see which pages AI cites and to automate custom prompt tracking across ChatGPT and Gemini. That turns one‑off errors into measurable gaps you can fix.
AI Opt-Out May Cost Sites A Google Top Stories Spot
I think every news publisher weighing Google’s new AI opt-out setting needs to read this article before making a decision.
The core finding is simple: opting out of AI Overviews may also remove your site from the Top Stories carousel when that carousel sits inside the AI Overview. NewzDash tracked news searches and found 15.5% of US and 17.46% of UK results with Top Stories inside the AI Overview. Entertainment searches saw the highest rates. Google’s help page covers the entire AI Overview feature, not individual elements like the carousel. Shehata from NewzDash calls this a high-confidence read, though unconfirmed. I recommend testing the setting on a small subset of pages first. The choice now carries a real visibility cost, not just a philosophical one.
13 Advanced SEO Techniques & Strategies for 2026
This article delivers 13 advanced SEO techniques with actionable strategies tailored for 2026. I find it valuable because it moves beyond basics and directly addresses how AI search changes the game.
Key points I recommend:
– Improve content’s “time to value” — put key info at the top. Investopedia does this well with summaries. This boosts engagement metrics that Perplexity uses for citations.
– Build topic clusters around pillar pages. This establishes authority for both Google and AI systems.
– Leverage strategic internal linking from high-backlink pages to related pages. Use descriptive anchor text.
– Maximize SERP real estate with structured data and multimedia. Crate & Barrel’s rich snippet example proves it works.
I think the time-to-value and topic cluster techniques compound best. Focus on those first.
From SEO to GEO: Why Rankings Alone Won’t Get You Visibility in AI Search
I recommend this Lumar piece because it makes a crucial distinction: seo geo rankings are no longer the same game. The article argues that AI search doesn’t rank links—it synthesizes answers. Rankings alone won’t get you cited.
Key takeaways I found valuable:
– GEO builds on traditional SEO, but the goal shifts from ranking #1 to becoming source material for AI responses.
– Google’s systems evaluate content at both page and passage level, so splitting content into tiny pages isn’t a GEO strategy.
– Different query types prioritize different signals: YMYL queries demand authority, while exploratory topics reward topical breadth and structure.
My practical advice? Stop obsessing over position tracking for AI search. Instead, audit your content for semantic richness, clear structure, and trust signals. Lumar’s pipeline framework—from availability to answer assembly—is a solid mental model for your GEO planning.
The future of Search & AI: What I learned from Google’s latest earnings call
Google’s latest earnings call makes one thing clear: the future search ai landscape is shifting from information retrieval to autonomous agent execution. I recommend every SEO practitioner study this call carefully — it reveals how Google is fundamentally redefining what search means.
Three data points stood out to me. First, Google spent $6 billion on physical assets this quarter — 60% on servers, 40% on data centers — despite revenue growing 24% year-over-year. They are building infrastructure for AGI, not just better search results. Second, Gemini 3.5 Flash-Lite now powers AI Overviews. This means Google understands conversational queries better than ever, which directly reduces traditional click-through traffic. Third, Sundar Pichai keeps saying “agentic.” Search will soon connect directly to apps like Instacart and Canva. Instead of finding a recipe, the agent orders your ingredients.
My practical advice: start building actionable tools for agents via WebMCP. The profitable website in this new era helps people accomplish something — not just read something. Google is also rolling out Universal Commerce Protocol. I have clients on the waiting list. Watch this space closely.
Interestingly, Google is prioritizing safety through models like Gemini 3.5 Flash Cyber. I suspect they have held back their most powerful models to ensure security first. This patience matters for everyone who depends on Google’s ecosystem.
The bottom line: optimize for agents, not just readers.
HubSpot AEO vs. Semrush AI Visibility: Which is right for your team?
After testing HubSpot AEO vs Semrush AI Visibility Toolkit on a real golf website, I recommend choosing based on your team’s primary workflow. HubSpot AEO wins for CRM-connected execution, while Semrush leads in SEO-driven analysis.
Here are three key findings from my hands-on comparison:
– Coverage differences. HubSpot AEO tracks ChatGPT, Gemini, and Perplexity. Semrush AI Visibility adds Google AI Overviews and AI Mode, giving you broader AI landscape visibility.
– Actionability gap. HubSpot AEO connects directly to your CRM, letting you turn AI citation gaps into follow-ups and content tasks. Semrush keeps you inside its SEO ecosystem, better for competitive keyword and prompt research.
– Trial and pricing. HubSpot offers a 28-day free trial with no credit card. Semrush AI Visibility Toolkit has no standalone trial; only Semrush One includes a 7-day trial with card required.
I recommend HubSpot AEO for marketing teams who already rely on HubSpot CRM and need a closed-loop process from detection to action. If your team’s core strength is traditional SEO and you want to expand into AI visibility alongside existing Semrush workflows, the AI Visibility Toolkit (or Semrush One) is the smarter pick.
My personal advice: start with the free AI Search Grader from HubSpot to see your brand’s baseline. Then run a 28-day trial of HubSpot AEO. If you need deeper prompt-level analysis across more engines, layer on Semrush. The two tools complement each other well, but most teams should commit to one primary platform.
Does Schema Markup Impact SERP Rank?
Schema markup does not directly impact SERP rank. That is the clear verdict from a rigorous 10-week controlled experiment featured in this Moz article. I recommend reading it because the data comes from real-world testing, not influencer anecdotes or correlational studies.
The study split 29 landscaping clients into a control group and a test group. All sites had identical CMS, templates, and authority. The test group received LocalBusiness schema markup. Both groups paused all other SEO efforts. The result: no significant ranking difference on Google, Google Mobile, Bing, or Yahoo.
Key takeaway: schema markup helps with contextual understanding and rich results, but it does not boost rankings on its own. Use it for user experience and data clarity. Stop expecting ranking lifts from schema alone.
🔗 Moz Blog
AI Search Isn’t Replacing Google, It’s Layering On Top – Similarweb Data
AI search isn’t replacing Google — it’s layering on top. Similarweb’s latest data proves this. I recommend you stop treating AI search as a separate channel and start measuring the stack as one system.
Three facts stand out. First, 95% of ChatGPT users also use Google. Nobody left. They added a layer. Second, only 6.8% of ChatGPT answers include a link in the US. 93 out of 100 answers send zero traffic. Third, 65% of cited pages sit two to three folders deep, but 59% of referral clicks land on the homepage. The cited content and the visited content are almost completely different.
I think this gap is your biggest opportunity. If you only track citation rates, you miss the real behavior. The people who click through land on your homepage — not the deep page AI quoted. That means your homepage needs its own conversion path for AI visitors.
My advice: split your reporting immediately. Track citation rate and citation depth as one KPI. Track homepage referral conversions as another separate KPI. Then optimize accordingly. Don’t argue about which layer matters more. Measure the stack and win the next two years.
Q3 AI Visibility: AI Citations, Brand Mentions & Content Refreshes That Work
Your Q3 AI visibility numbers are likely misleading. A new IQRush paper proves AI citation rankings shift randomly between queries, so a single reading is useless for decision-making.
Here’s what matters: the paper tested 30 platform-topic combinations across SearchGPT, Gemini, and Perplexity. It found you need 33 to 94 citation-bearing answers before rankings stabilize. Three tests never settled even after 125 questions. The margin of error on top-10 sites averages five positions. A 3-point gain is noise, not a win.
I recommend auditing your AI visibility tracker now. Ask if it runs repeated measurements and reports a confidence range. A single clean number is a red flag. Measure multiple times both before and after content changes. Trust only the top positions—the middle and tail are coin flips. Use the paper’s stopping rule to know when you have enough data.
European Search Strategy Goes Beyond Google & Bing
European search strategy requires a mindset shift. Google is still dominant, but the forces shaping discovery are far more fragmented. This article delivers a practical blueprint for navigating Europe’s unique regulatory and competitive landscape. I recommend it for any SEO team managing cross-border campaigns.
Motoko Hunt breaks down the data. In Germany, Google holds around 80% share. Bing has a genuine 10%, and local engine Ecosia is gaining on privacy and sustainability preferences. In Czechia, Seznam.cz claims 12% — a rare double-digit domestic player. Regulation drives these numbers. The Digital Markets Act forces Google to share ranking data with rivals starting January 2027. The UK’s CMA now lets publishers opt out of AI Overviews without losing organic visibility. A Munich court ruling in May 2026 held Google directly liable for false AI-generated claims. That legal shield for search engines no longer covers generative answers.
Marketplaces absorb the query before Google sees it. Consumers open Zalando, Otto, or Allegro directly. Comparison engines like Idealo and Kelkoo reroute traffic. Brands must treat marketplace listings as SEO assets — titles, attributes, and reviews now determine discoverability.
The biggest takeaway is timing. AI features reach Europe later than the U.S., but adoption will accelerate once regulatory sign-off clears. I recommend teams invest now in structured data and consistent business identity. Brands with machine-readable profiles gain visibility when answer engines avoid unverified claims. Ignore the fragmentation, and your European search strategy will miss the real opportunity.
HubSpot AEO vs. Otterly: Platform or standalone tool?
If you want AI visibility data to drive real content and pipeline actions, the HubSpot AEO vs. Otterly decision comes down to workflow integration. HubSpot AEO tracks ChatGPT, Gemini, and Perplexity, then surfaces CRM-powered prompt suggestions and lets you publish content without switching tools. Otterly covers more engines – including Google AI Overviews and Copilot – and offers strong standalone reporting via Looker Studio. But it leaves acting on insights entirely up to you.
I recommend HubSpot AEO for teams already in the HubSpot ecosystem. Its 28-day free trial (25 prompts) lets you test before committing $50/month. Otterly’s 7-day trial is shorter, and its paid tiers start higher for similar monitoring. If your content workflow is separate and you just need monitoring, Otterly works. But if you want AEO data to automatically feed content creation and CRM tracking, HubSpot AEO is the smarter choice.
