SEO Recap covering July 22, 2026
Daily summary of what matters in SEO, GEO, AEO, and AI search generated with Claude Code (beware of hallucinations)
Generative Engine Optimization (GEO) & Answer Engine Optimization (AEO)
Semrush Study: Topical Authority Still Decides Who Owns AI Answers
Source: Search Engine Land
- The Semrush AI Visibility dataset covered 1,094 US categories, 5 prompts each, monthly snapshots January to June 2026 in ChatGPT, spanning 220,000+ domains and 600,000+ citations.
- Only 15.2% of categories had a clear "owner" in June 2026, and 89.3% of estimated AI-search demand sits in categories with no owner, meaning the window to claim a topic is currently open.
- Clear owners kept the top spot in 90.4% of month-over-month comparisons; leadership changes clustered in categories with narrow leads (median 1.3 points versus 2.9 for retained leaders).
- Brand mentions, not citations, track ownership: the most-cited domain was the most-mentioned only 20.8% of the time, and product/service landing pages were cited most while homepages appeared just 4% of the time.
Contentful's Dillon: Volume Content Fails as Zero-Click Search Grows
Source: Search Engine Journal
- Gabriel Dillon frames the argument around the stat that 60% of Google searches end without a click, so content must be tied to a business outcome and built for a specific reader.
- He runs four accountability questions on every piece before publishing: does it produce the expected outcome, who is it for, how do you identify those people, and how does the insight scale.
- Dillon argues detection of AI content is the wrong fight and teams should instead compete for the AI answer layer through GEO and AEO.
- He lays out three personalization signal tiers starting with new versus returning visitors, using signals ad campaigns and loyalty programs already generate.
Building an AI-Ready "Source of Truth" Around Decision Knowledge
Source: Search Engine Journal
- Bill argues AI recommendations are "confidence decisions" assembled from evidence, so brands compete to provide the highest-confidence information rather than just to be found.
- Most companies have product specs but lack "decision data" (whether a mattress sleeps cool, suits side sleepers, offers financing), leaving retailers and review sites more authoritative about products than the brands themselves.
- He cites a site-search example where one query drew 100,000+ requests about upgrading a pass; the FAQ answered "yes" but never linked the process, ending the journey instead of advancing a sale.
- The framework rests on four capabilities (knowledge completeness, connectivity, answer readiness, governance) and proposes a "VP of Answers" style role to own knowledge integrity.
Getting Local Businesses Cited When AI Overviews Answer Most Searches
Source: Search Engine Journal
- A Whitespark study found AI Overviews appear for 68% of local searches versus 39% for local packs; informational queries like "how long does an eye exam take near me" trigger AI Overviews 92% of the time and hybrid-intent queries 97%.
- Location pages should prioritize fact density: table-formatted NAP and hours, city-specific testimonials and FAQs, real project examples, and pricing ranges, since "price, cost, buy" queries trigger AI Overviews over 80% of the time.
- Off-page sources dominate citations: Omniscient Digital found owned content is only about 23% of citations in branded queries, and Averi.ai found Reddit (21%) and YouTube (18.8%) are the most-cited sources in Google AI Overviews.
- Kevin Indig's analysis of 1.2 million ChatGPT responses found 44.2% of citations come from the first 30% of page content, and Ahrefs found AI-cited content averages 25.7% fresher than traditionally ranked content.
"Synthetic Competition": AI Shapes Buyer Opinions Before Your Site
Source: MarTech
- The author describes competing not just with rival brands but with AI answers, comparison sites, Reddit threads, and analyst reports that shape decisions before a buyer reaches your website.
- The 2026 AI Attribution Report from Knocommerce and GR0 (millions of surveys across 6,000+ Shopify brands) found AI-attributed product discovery rose roughly twelvefold in 2025, with AI-driven discovery growing 25 to 30% month over month.
- SparkToro found 68% of Google searches now end without a click, reducing chances to shape first impressions on owned properties.
- The piece recommends investing in content only you can create (original research, proprietary data, first-hand experience) and tracking brand mentions in AI responses, share of voice, and third-party citations rather than traffic alone.
AI in Search / AI Overviews
Google Search Now Routing Queries to Gemini 3.5 Flash-Lite
Source: Search Engine Roundtable
- Google confirmed Gemini 3.5 Flash-Lite is rolling out in Search, initially cited for agentic search and document processing tasks.
- Google's Rajan Patel said it will be one of the models Search queries get routed to depending on the question, and is especially strong at understanding intent for conversational queries.
- The model delivers 350 output tokens per second and outperforms prior Flash-Lite generations on agentic and coding benchmarks like Terminal-Bench 2.1 (54% vs 31%).
- Robby Stein said the model offers stronger instruction following, and observers expect it to eventually power AI Overviews and AI Mode.
Technical SEO
Using "Query Deserves a Page" to Decide Site Architecture for Local SEO
Source: Search Engine Land
- The author introduces "Query Deserves a Page" (QDP), adapted from Query Deserves Freshness, to decide whether a query variation warrants its own page or just a section.
- Four tests determine a separate page: high search demand, distinct entities, low query similarity, and a repeatable pattern; missing these causes micro-cannibalization from near-duplicate pages.
- Query similarity is weighted by term importance, not string matching, citing Google's "Detecting query-specific duplicate documents" patent and BERT-based term weighting research.
- The piece recommends targeting the most important location-service pair from the homepage, since it carries the highest PageRank and crawl frequency.
The Case for Publishing Less: Crawl Budget, Indexing, and Pruning
Source: Search Engine Land
- HubSpot data links higher publishing frequency to more traffic, but the author argues every strategy hits diminishing returns where improving existing content beats adding more.
- More high-quality pages are being discovered but left unindexed as AI accelerates site creation; contributing factors include limited crawl budget, too many URLs, weak authority, low quality, and thin internal linking.
- QuickBooks deleted 2,000+ resource pages and saw traffic rise 20% with 70%+ more lead signups; a Userpilot case study cut nearly 25% of content and hit its highest-traffic month.
- The author recommends a Screaming Frog audit pulling 12 months of conversions, traffic, and keywords, then deciding per URL to keep, consolidate, refresh, or delete, and focusing on 1 to 4 strong articles per topic.
Google Rewrites Its "Optimize Your Crawl Budget" Documentation
Source: Search Engine Roundtable
- Google updated the crawl budget doc for clarity and consistent terminology, newly stating that every site starts with the same default, conservative crawl capacity limit that adjusts upward if demand grows and the site stays healthy.
- The doc clarifies that crawl capacity limit is shared across all Google crawlers, so heavy demand from one crawler can reduce capacity for others.
- Google reiterates practical levers: improve server response times and support HTTP 304 (Not Modified) caching so unchanged pages reuse the cached version and save server resources.
- Google says crawl demand for Googlebot varies by a site's size, update frequency, page quality, and relevance versus other sites.
Why Client-Side Rendering Makes Sites Invisible to AI Crawlers
Source: MarTech
- Contentsquare data shows AI-referred traffic to brand sites rose 632% in about 10 months; one client lost 40% of organic traffic to a page yet saw conversions rise as AI-referred visitors arrived more informed.
- AI crawlers for ChatGPT Search, Claude, and Perplexity fetch only raw HTML and do not run JavaScript, so client-side-rendered pages return an empty shell; Google's renderer is the exception, meaning a site can rank in Google yet be invisible to answer engines.
- The fix is server-side rendering plus "dual-mode media": transcripts, alt text, chapter headings, and structured metadata so machines can read video and image content.
- Forrester found 51% of software buyers now start research in an AI chatbot (up from 29%); recommended steps include auditing rendering with curl or a bot simulator, adding machine layers to assets, and baselining bot versus human conversion.
Organic Search & Algorithm Updates
Court Dismisses Google's DMCA Claims Against SerpApi in Scraping Suit
Source: Search Engine Land
- A federal court granted SerpApi's motion to dismiss Google's two DMCA claims on July 20, ruling that plain search results (URLs, snippets, factual index data) are public facts, not copyrighted works.
- The court permanently dismissed claims tied to non-copyrighted results but gave Google 21 days to amend the copyrighted-content portion if it can show copyright owners authorized its SearchGuard anti-scraping system.
- The judge did agree Google has standing and that SerpApi's IP rotation, fingerprint spoofing, and CAPTCHA solving amount to "circumvention" under the DMCA.
- SerpApi CEO Julien Khaleghy framed the ruling as a win for open access to public search data that rank-tracking and AI-visibility tools depend on.
Google Q2 2026: Search Ad Revenue Up 14.5%, Biggest Quarter Ever
Source: Search Engine Roundtable
- Alphabet revenue rose 24% to $119.8 billion; Google Search and other grew 17% to $63.3 billion, roughly in line with a $63.4 billion estimate.
- Sundar Pichai said AI Overviews and AI Mode have merged into one search experience, that AI Mode has surpassed 1 billion monthly active users, and that AI features are driving incremental query growth.
- Pichai said Google is now sending billions of clicks to websites weekly through AI features and cut the cost of AI Mode responses to its lowest level since launch.
- Google Cloud revenue accelerated to 82% growth at $24.8 billion, and the Gemini App reached 950 million monthly active users.
Judge Grants SerpApi's Motion to Dismiss in Google Scraping Case
Source: Search Engine Roundtable
- U.S. District Judge Yvonne Gonzalez Rogers dismissed Google's DMCA Section 1201 claims, ruling that URLs, snippets, and factual index data are publicly accessible facts, not copyright-protected works.
- Claims over uncopyrighted results were dismissed without leave to amend; claims over third-party content were dismissed with leave, giving Google 21 days to show copyright holders authorized its SearchGuard barrier.
- The judge affirmed Google has standing and that spoofing fingerprints, rotating IPs, and solving CAPTCHAs constitute circumvention, but noted circumvention is only illegal when the barrier guards authorized copyrighted works.
- Commentators called it a win for scrapers, though likely not financially material to Google.
Published on