Check the latest stories happening in digital marketing…
LLM optimisation in 2026: Tracking, visibility, and what’s next for AI discovery
https://searchengineland.com/llm-optimization-tracking-visibility-ai-discovery-463860
The discipline of optimising for large language models (LLMs) like ChatGPT, Gemini and Claude is moving from intuition to measurement. Key challenges include the fact that LLM queries do not publish volume data, responses vary widely and are shaped by hidden context. The recommended tracking model uses a consistent sample of high-intent queries to measure when your brand is cited or mentioned and to compare share-of-voice over time.
Optimising for LLM visibility blends traditional SEO fundamentals with new tactics: produce clear, factual content; build authoritative entity-based clusters; and earn mentions in the sources that LLMs trust (such as Wikipedia or other authoritative forums). Although LLM traffic is still small compared to search, early movers who adopt measurement and strategic alignment are better positioned for the AI-discovery era.
Anthropic Research Shows How LLMs Perceive Text
https://www.searchenginejournal.com/anthropic-research-shows-how-llms-perceive-text/559636
Anthropic’s research reveals that large language models (LLMs) don’t just generate text token by token but instead form internal “maps” that mirror how humans perceive and organise information. These models appear to develop layered abstractions of meaning that guide their predictions and outputs.
For anyone working with or analysing AI-driven systems, this means content must be designed not just for keywords but for structural clarity, coherent layering of ideas and strong semantic signals that align with how LLMs interpret context and meaning.
Regex for SEO: The simple language that powers AI and data analysis
https://searchengineland.com/regex-seo-ai-data-analysis-463933
Regular expressions (regex) are highlighted as a highly efficient tool for both SEO and AI-driven workflows. They can be used to filter and analyse large datasets—from keywords in Google Search Console to URLs extracted during a crawl—with a single line of pattern matching. The article explains how regex is also at the heart of how large language models (LLMs) and natural language processing (NLP) systems parse and interpret text.
Key use cases include extracting branded vs non-branded queries, cleaning messy query data, grouping URLs by pattern, and validating text inputs in tools like Google Analytics, Looker Studio, and crawl software. Learning the fundamentals of regex empowers marketers and analysts to create more precise filters and workflows, especially when combining human insight with AI assistance.
Google Q3 Report: AI Mode, AI Overviews Lift Total Search Usage
Google’s Q3 2025 report shows that AI-driven features such as AI Mode and AI Overviews are not reducing search activity but actually increasing it. Usage across Google Search continues to grow, particularly among younger audiences who are more comfortable using conversational AI to find information. This suggests users are conducting more searches, not fewer, as they explore the new, interactive formats offered by AI-powered results.
For SEO and content teams, this marks a significant change in how visibility is earned. Traditional rankings alone no longer capture the full picture of opportunity. Brands now need to think about how their content can appear within AI-driven discovery paths, such as citations in overviews or generative summaries, rather than focusing solely on page-one listings.
Chrome To Warn Users Before Loading HTTP Sites Starting Next Year
Starting in October 2026, Google Chrome will begin warning users before they access any website that still uses HTTP instead of HTTPS. The change will roll out with Chrome version 154, where the default “Always Use Secure Connections” setting will be enabled for all users. This means Chrome will attempt to upgrade every connection to HTTPS automatically and display a warning if the site does not support it.
The rollout begins in April 2026 for users enrolled in Enhanced Safe Browsing under Chrome 147, before expanding to everyone later in the year. The warnings will only apply to public websites, excluding internal networks and local addresses. Although 95–99 percent of all public web traffic already runs on HTTPS, Google says even the small remaining percentage poses security risks, as one unencrypted connection can expose users to potential attacks.
Google’s Advice On Canonicals: They’re Case Sensitive
https://www.searchenginejournal.com/googles-advice-on-canonicals-theyre-case-sensitive/559440
John Mueller from Google has emphasised that canonical URLs are case-sensitive when it comes to paths, filenames and query parameters, though not for the hostname/domain.
He pointed out that relying on hope that Google will automatically handle mismatches is a weak strategy.
If a URL appears as /Site/Topic/topic-title/ but the canonical tag points to /site/topic/topic-title/, this inconsistency could result in duplicate-content issues or signal confusion to crawlers.
Ensuring exact match in case between canonical tag and the live URL boosts clarity and helps crawling, indexing and ranking.
Also been reading…
Disney Google Sitelinks Blackhat SEO Hack Still Live
https://www.seroundtable.com/disney-google-sitelinks-hack-blackhat-seo-40349.html
Google Search Console adds Query groups
https://searchengineland.com/google-search-console-adds-query-groups-463820
The agentic web is here: Why NLWeb makes schema your greatest SEO asset
https://searchengineland.com/agentic-web-nlweb-schema-seo-asset-463778
Google AI Mode: What it is & how it impacts search
https://searchengineland.com/guides/google-ai-mode
Chrome To Warn Users Before Loading HTTP Sites Starting Next Year