Latest organic search news – June 26

News | 1st June 2026

Matthew Finch

Ella Grappy

We’ve compiled the essential updates from Google and the broader world of search from the last month – keeping you up to date with everything you need to know.

TL;DR

YouTube is the strongest signal for AI brand visibility. New Ahrefs research across 75,000 brands found YouTube mentions correlate more strongly with AI visibility than any other metric – beating backlinks, domain authority, and content volume.
“Agent-ready” is becoming a real conversation in SEO. Two technically credible practitioners have published detailed write-ups on what it means to prepare a website for AI agents. The framing is evolving fast.
The infrastructure layer is separating from the citation layer. There’s a meaningful distinction between what makes a site readable by agents and what makes a site cited by AI search. They’re related but not the same.
Away Resorts ranks 3rd in the UK staycations category. An independent leaderboard from Rise at Seven and Ahrefs puts a Distinctly client above Haven, Hoseasons, and Center Parcs – brands with significantly higher consumer recognition.
Google is increasingly rewarding brands people already know and look for. But the data shows it’s still possible to outrank more established players by building search infrastructure ahead of brand scale.
Google has introduced Preferred Source labels and a Highly Cited badge. Both have direct implications for digital PR strategy.
Google has warned against inauthentic brand mentions. And explicitly compared the practice to paid links.

Ahrefs AI Search Benchmark Report: YouTube is the strongest signal for AI visibility

The Ahrefs Q1 2026 AI Search Benchmark Report is one of the more substantial pieces of AI visibility research published so far this year. It covers 13 studies drawn from 146 million search result pages and 730,000 AI responses across ChatGPT, Google AI Overviews, and Google AI Mode.

YouTube mentions are the strongest correlated signal for AI brand visibility. Across a study of 75,000 brands, mentions of a brand in YouTube video titles, transcripts, and descriptions showed a Spearman correlation of approximately 0.737 with AI visibility – consistent across all three platforms. That beat branded web mentions (which previously topped Ahrefs’ rankings at 0.66 to 0.71), and it beat every traditional SEO metric by a significant margin.

The mechanism makes sense when you consider training data. Both Google and OpenAI have used YouTube transcripts extensively in model training. Google AI Mode, which crossed one billion monthly users this year, cites YouTube more than any other domain. The platform is baked into both the input and output of the AI systems that now shape how brands get discovered.

A few other findings from the report worth noting:

Content volume has almost no relationship with AI visibility. The number of pages on a site showed a correlation of approximately 0.194. This directly undercuts advice circulating in parts of the SEO industry to generate large volumes of content programmatically to improve AI citation rates. It doesn’t work.
Link volume showed only weak correlation. Domain rating sat between 0.266 and 0.326 depending on platform. Traditional link-building is not the AI visibility lever many have assumed.
The three platforms behave differently. Google AI Mode showed the strongest correlation with branded search volume and branded anchors, making it the hardest platform for emerging brands to break into. ChatGPT showed the weakest correlations with authority signals, making it the most accessible entry point for newer brands. AI Overviews showed the highest correlation with domain rating among the three, though the difference was modest.
Despite those differences, the same brands tend to win across all three. Output overlap between platforms correlated at 0.779 – the paths differ but the destinations are largely the same.
AI Overviews have reduced clicks to top-ranking content by 58%, up from 34.5% eight months prior, based on GSC data across 300,000 keywords comparing December 2023 with December 2025.
Google sends 190x more traffic than ChatGPT, based on Ahrefs’ Web Analytics data across 76,000 websites.

We’ve been pushing clients towards YouTube for the last 18 months – primarily because it’s where your customers already are, and being present there is table stakes for any brand that wants to show up across the modern search landscape. This data adds a second layer of importance to that argument. It’s not just about reaching people on the platform itself – it’s that YouTube presence is now one of the strongest signals feeding whether AI recommends your brand at all. Two good reasons to be there rather than one.

If you want a steer on which channels are worth prioritising for AI visibility right now, get in touch.

Is “agent-ready” the new technical SEO?

Two posts published in the last few weeks are worth reading together. Suganthan Mohanadasan (co-founder of Snippet Digital and Keyword Insights) and Joost de Valk (founder of Yoast) have both independently published detailed walkthroughs on making websites readable by AI agents. They come from strong technical backgrounds, and the fact that they’ve both landed on similar frameworks in the same window is worth paying attention to.

The starting point for both is a distinction that’s easy to miss. A site being in Google’s index is not the same as a site being legible to the agents that increasingly sit between a user and the open web. AI agents – whether browsing autonomously, grounding answers in live content, or acting on behalf of users in tools like Claude Code or Cursor – interact with your site differently to a search crawler or a human. The question “can Google find and rank this page?” is not the same question as “can an AI agent read, interpret, and act on this page reliably?”

The two layers

Suganthan breaks this into two layers.

Layer 1 – the page itself. Semantic HTML, stable layout, accessibility fundamentals, correct use of elements like <button>, <nav>, and <main>. This has overlap with traditional technical SEO and accessibility work. Most of it is not new.

Layer 2 – everything around the page. The protocol stack that makes a site’s structure, permissions, and capabilities discoverable without loading HTML. llms.txt, robots.txt with AI bot rules, Link headers pointing to key resources, Markdown negotiation, and a set of emerging well-known files for MCP server discovery, A2A agent cards, and API catalogues.

Joost’s post covers the same territory, framed around a practical exercise: he ran his own site through isitagentready.com, scored 25 out of 100, and methodically worked through the gaps. His write-up is useful for showing what “doing the work” actually looks like – including which items are genuinely useful, which are optional for most sites, and which the scanner flags incorrectly because it can’t distinguish a deliberate choice from a gap.

What the evidence actually says

Both writers are honest about where the evidence is weak. Suganthan is particularly direct: an Ahrefs study across 1,885 pages found no significant 30-day citation uplift from structured data. Neither that study nor any equivalent has been published for llms.txt, MCP, or A2A. Google’s AI optimization documentation explicitly states you don’t need new machine-readable files to appear in generative AI search. That’s a direct contradiction of what much of the agent-readiness industry is selling.

What the protocol work does do, on current evidence, is reduce the cost and friction of being read. Suganthan’s own Cloudflare logs showed 8,060 AI crawler requests in a seven-day window from at least eight organisations. Markdown negotiation dropped page payload from around 85KB HTML to 16KB. The agents are already coming. Making it cheaper and cleaner for them to read you is a reasonable argument. Whether that translates to citations is a separate question the data doesn’t yet answer.

One finding from Suganthan’s crawler data is worth highlighting specifically. OpenAI represented 43% of crawl volume across his site, but generated just three human referrals over seven days. Google generated 2,570 referrals. Crawl rate and surfacing rate are very different things, and the gap between them varies significantly by platform.

Where to start

The minimum viable checklist that both writers converge on:

AI bot rules in robots.txt
An llms.txt file
Semantic HTML and accessibility fundamentals
A sitemap with Link headers
Markdown negotiation

The more advanced protocol work – MCP server cards, A2A agent cards, WebMCP, API catalogues – is defensible for sites that expose tools or APIs, and optional-to-skip for content sites and agencies.

The schema.org analogy is useful here. In 2011, structured markup was niche infrastructure with uncertain direct value. By 2017 it was table stakes. Whether agent protocols follow the same path isn’t guaranteed, but the shape of the bet is similar. Early implementation may compound. Or the standards might fragment. Both writers acknowledge this honestly, and the minimum viable checklist is hedged against the latter.

Three tools for auditing agent-readiness

Three tools have emerged that score your site against this emerging set of standards, each with a different angle.

isitagentready.com (Cloudflare) checks the protocol layer specifically: robots.txt, sitemap, Link headers, Markdown negotiation, AI bot rules, Content Signals, MCP server cards, A2A agent cards, WebMCP, API catalogues, OAuth metadata, and agentic commerce protocols. It scores out of 100, lets you customise which checks run, and outputs recommendations formatted for pasting into a coding agent. The default scan covers a broad set, some of which won’t apply to most sites.

agentchecker.ai takes a different approach: it sends a real AI agent to browse your website and tests whether the agent can actually navigate and complete tasks. The focus is on practical usability – structured data, navigation, forms, content hierarchy – rather than protocol compliance. Useful for identifying UX-level friction that a protocol scan won’t surface.

Glippy (by Jan-Willem Bobbink) covers 16 GEO dimensions across 240+ checks, including information density, entity authority, citability, factual verifiability, content freshness, and agent interactivity. It’s available as a Chrome extension and has a free tier. More content-focused than the other two, with meaningful overlap with classic SEO and GEO best practice.

All three are worth running as a baseline audit for any site. The caveat from Suganthan applies: a red mark on any of these tools doesn’t necessarily mean “implement this now.” Some checks don’t apply to your site type. Treat the output as a structured checklist to investigate, not a prescription.

Away Resorts ranks 3rd in UK staycations – ahead of brands with far higher awareness

Carrie Rose at Rise at Seven published an analysis this month of the UK staycations category using a new category leaderboard in collaboration with Ahrefs which scores brands across three dimensions: Google discoverability, AI search visibility, and brand demand. The leaderboard has tracked over 300 categories since launch, and travel has been added in time for peak holiday season.

Away Resorts – a Distinctly client – sits 3rd in the UK staycations category. That puts them above Sykes Cottages, Hoseasons, Haven, and Center Parcs, which sits as low as 14th despite being one of the most recognised names in the sector.

The finding that stands out most from Carrie’s analysis is what’s holding each brand back. Away Resorts scores very strongly on Google visibility but is pulled down by brand demand – the volume of people actively searching for the brand by name. That gap reflects a business that has built genuine search infrastructure ahead of where its broader brand awareness currently sits. Haven is flagged as the most interesting case: strong brand demand, reasonable AI visibility, but Google discoverability that’s dragging them down and preventing them from converting that awareness into category leadership.

Carrie’s broader point is about what it actually takes to be a category leader. Google discoverability alone isn’t enough – you need brand demand too, and that’s where Away Resorts has room to grow. But the flip side of that argument is just as interesting: doing great SEO can put a brand at the top of these leaderboards precisely because it compensates for the awareness gap. Away Resorts is outranking household names in a competitive category because of search investment, not despite lacking the brand recognition of those above them.

It also reinforces something we’re seeing more broadly across search: brand signals – visibility, awareness, authority – are playing a growing role in how Google determines who ranks. The days of outranking a household name purely on technical SEO and content volume are getting harder. Google is increasingly rewarding brands that people already know and look for.

But Away Resorts sitting 3rd above Haven, Hoseasons, and Center Parcs shows that the game isn’t over for brands with awareness gaps to bridge. What it takes is being nimble enough to move quickly when the landscape shifts, willing to test things that bigger, slower organisations won’t commit to, and consistent enough to build search infrastructure before the brand awareness catches up. That combination – strong Google visibility built ahead of brand scale – is exactly what this data reflects.

The brands that wait until they have the awareness of a Center Parcs before investing in search are the ones that end up 14th on a leaderboard they should be leading.

Google’s “Preferred Sources” comes to AI Mode and AI Overviews

Google has officially rolled out Preferred Source labels within AI Mode and AI Overviews, alongside a new carousel giving users a way to surface content from publications they already trust directly within AI responses.

Google has said that people are twice as likely to click through to a Preferred Source, and that users have already selected more than 345,000 unique sources – a signal worth paying close attention to.

Alongside that, Google has also introduced a “Highly Cited” badge on search results, to help users identify original reporting and influential coverage, making it easier to spot articles that many other stories have cited and find the primary reporting that others are referencing.

The Highly Cited label is particularly relevant for digital PR strategies. It creates a visible reward for original research, data studies, and reactive commentary that earns widespread pickup from authoritative publications. Being the source that others cite is now literally badged within search results.

Google warns against inauthentic brand mentions and compares them to paid links

This one is a direct signal to the industry. Google has issued explicit guidance warning against the practice of seeking inauthentic mentions to gain visibility in AI Mode and AI Overviews.

Google’s statement was direct: “Our generative AI features can show what’s being said about products and services across the web, including in blogs, videos, and forum discussions. However, seeking inauthentic ‘mentions’ across the web isn’t as helpful as it might seem. Our core ranking systems focus on high-quality content while other systems block spam; our generative AI features depend on both.”

Google analyst Gary Illyes went further, comparing the pursuit of inauthentic mentions directly to paid linking practices. That comparison matters. It validates what we’ve always argued: that building genuine authority through earned editorial coverage, original research, and trusted commentary is the right approach – not because it’s the principled one, but because it’s the one Google’s systems are built to reward. Shortcuts that worked in link building don’t transfer to AI visibility, and Google is now saying so explicitly.

Enjoyed this roundup? Stay updated with the latest organic trends, tips, and exclusive insights by subscribing to our newsletter.