Welcome to our latest SEO Trends & Insights. Each month, our specialist SEO team analyses the shifting search landscape, algorithm updates, and emerging behaviours to bring you the insights that matter most. Stay ahead with expert commentary, actionable recommendations, and the trends shaping performance right now.
Google: llms.txt is not an SEO lever – Search ignores it, so it won’t move rankings
A new announcement by Google clarifies that adding an llms.txt file will neither help nor harm Google Search rankings because Google Search ignores it. The guidance also implies you don’t need special AI-targeted files like llms.txt to appear in Search (including generative AI experiences), so implementing it should be viewed as optional and primarily relevant to non-Google tools or agents rather than a Search optimisation tactic.
Why we care: For anyone weighing llms.txt as part of an AI SEO roadmap, this is a useful de-prioritisation signal: it should not distract from fundamentals that do influence visibility (crawlability, indexation, structured content, authority, and technical hygiene). If llms.txt is adopted for broader AI ecosystem reasons, it should be treated as a separate workstream with clear objectives and measuremen, without expecting Google ranking impact.
AI Overviews may cite your “best of” content but still recommend competitors most of the time
A recent analysis by Lily Ray of 100 B2B “best [category] software” queries found a clear citation–recommendation gap in Google AI Overviews: self-promotional “best of” listicles can be cited as sources, yet the publishing brand is often not recommended. In the dataset, AI Overviews cited self-promotional listicles while excluding the publishing brand from recommendations in about 69% of cases meaning your page may supply the evidence, but competitors get the endorsement.
Why we care: For UK and EU strategies, this reinforces that “being cited” in AI answers is not the same as “being chosen.” If your listicle names competitors, you may be helping Google confidently recommend them while you only receive a citation. Measurement and content strategy should therefore separate citation visibility from recommendation outcomes, and evaluate “best of” content by whether it drives recommendation placement (and downstream demand), not just source attribution.
Early‑2026 data suggests most Google searches don’t send traffic out and only ~23% of queries lead to open‑web clicks
Recent clickstream-based reporting on U.S. Google Search (SparkToro analysis using Similarweb data, January–April 2026) suggests that most searches now end without a site visit. One view puts zero-click at 68% over that period, while another frames the same shift as only ~23% of queries sending a click to the open web. 39% of searches end with no further action, 29% lead to a new query, and 32% produce a click; of the clicks that do happen, 66% go to the open web, 27% to Alphabet properties, and 6% to paid ads.
Why we care: The takeaway is to expect “visibility without visits” to increase, even when rankings hold, and to evolve reporting beyond sessions to include SERP/feature presence and brand demand. Strategy should focus on winning on‑SERP/assistant-led placements where possible, while also designing content and journeys that create a clear reason to click when a site visit is needed for conversion.
Google’s May 2026 core update boosted Reddit across niches, while ChatGPT recommendations are linked to higher‑intent brand visits
A recent analysis highlighted two converging discovery shifts that matter for UK and EU clients: forum-heavy SERPs in Google and recommendation-led influence from AI assistants. After Google’s May 2026 core update, SE Ranking’s study of ~100,000 keywords across 20 niches found Reddit increased its top‑3 presence in every niche tracked, reaching 10% of all top‑3 positions, with #1 rankings up 54%. Separately, a Similarweb study reported that users were on average 2.5× more likely to visit an AI‑recommended brand’s website than a direct competitor within seven days, and those AI‑influenced visits were more engaged (12 pages/11.8 minutes vs 6.5 pages/5.6 minutes), with many AI‑influenced visits ultimately arriving via search rather than an “AI” referrer.
Why we care: Together, these findings reinforce that winning visibility is increasingly about being present where users seek opinions and recommendations – whether that’s UGC-heavy platforms surfacing in Google or assistant-driven brand shortlists shaping downstream demand. For UK/EU strategy, that means broadening beyond classic “rank → click” thinking: strengthen credible third‑party/community presence (to capture SERP real estate as forums rise), and invest in brand narratives and proof points that make it easier for AI tools to recommend you (because recommendations can translate into higher-quality visits). Measurement should also evolve to track influence signals (recommendation exposure, community visibility, branded uplift) alongside traditional SEO KPIs.
Want to understand what these changes actually mean for your organic performance in 2026? Now’s the moment to review how your content, structure and measurement stack hold up in AI-driven search.
Get in touch: hello@threepipereply.com
