The digital marketing world has not slowed down. Google rolled out its second core update of the year just six weeks after the first, AI Mode hit its one-year mark with some revealing usage data, and paid search started to look a lot more conversational. On the social side, Meta is quietly building a subscription business while ChatGPT ads inch closer to performance marketing territory.
Here is everything worth knowing from this month, and what it means in practice.
Google released the May 2026 broad core update on 21 May, the second core update of the year, arriving just 6 weeks after the March 2026 core update completed. The rollout window is up to two weeks and on the 1st of June it was official over. Early data pointed to significant volatility around 30 May, with some sites hit hard. Google’s description is the standard one: a regular update designed to surface relevant, satisfying content for all types of sites.
The unusually short gap between March and May is worth flagging. The 3 to 4 month spacing that has been typical in recent years appears to be giving way to a faster cadence.
Read more: Search Engine Land coverage | Search Engine Roundtable on the hard hit
A significant set of changes to how Google surfaces content across AI Overviews and AI Mode. Google is bringing preferred sources to AI Mode and AI Overviews, letting users spot links from sources they have already selected. Google reports that users have chosen over 345,000 unique sources globally, and people are twice as likely to click through to a preferred source. Alongside this, Google is rolling out a perspectives carousel for developing topics and expanding the highly cited label to more web article links, making it easier to identify primary reporting that other articles are referencing.
Read more: Search Engine Land | Search Engine Journal on the 345k sources milestone
Google published its first substantive first-party data on AI Mode behaviour, timed to coincide with Google I/O. AI Mode has surpassed one billion monthly active users globally, with queries more than doubling every quarter since launch. The average AI Mode search is three times longer than a traditional search, and follow-up queries rose over 40% month on month in the US. More than one in six AI Mode searches are now multimodal, using voice, images, or video. Planning queries grew 80% faster over six months, and decision questions starting with ‘which’ increased 40%.
Read more: Search Engine Journal
This data comes from an analysis of 846,000 real Google search sessions. When an AI Overview is present, cursor positions spread across 83% of a viewport compared to 66% without one. Users kept their cursors still 44% of the time versus 29% without an AI Overview, and reverse-direction scrolling accounted for nearly half of all scroll movement, suggesting active evaluation rather than linear reading.
AI Mode behaves like a closed loop. In 88% of AI Mode tasks, users accepted the AI’s shortlist as-is. 74% picked the item ranked first. 64% clicked nothing at all. AI Overviews also collapse the difference between search intent types: without one, only 12% of navigational searchers are still on the SERP at 21 seconds. With one present, that figure rises to 45.8%.
Read more: Search Engine Land
New Similarweb data shows that since OpenAI began surfacing more prominent links in ChatGPT answers, referral traffic from the platform is up 150%. Pageviews per visit are up 24% and time on site is up 11%, suggesting this traffic arrives with specific intent rather than casual browsing.
Read more: Search Engine Roundtable
Google Analytics now assigns an ‘ai-assistant’ medium when a referrer matches a recognised AI assistant, and groups those sessions under a new AI Assistant default channel group. Property owners no longer need to build custom regex patterns to separate AI assistant visits from referrals. Google names ChatGPT, Gemini, and Claude as examples of recognised referrers, though the full list has not been published.
One of the most practically useful research pieces of the month. Evertune reviewed the 6,000 most-cited URLs per model across ChatGPT, Copilot, Gemini, Google AI Mode, AI Overviews, and Perplexity for March and April 2026. Of roughly 25,000 unique URLs, half were listicles. Across nearly 400 million citations from all models, 63% pointed to listicles. The vast majority were ranked lists such as ‘Top 5 CRM Tools’, making up 71% to 86% of all listicles depending on the model. Cited pages typically ranged from 1,000 to 2,000 words, averaged 18 words per sentence, linked frequently, and used structured headings throughout.
Google published its official optimisation guide for generative AI features in Search on 15 May, and the most notable section is the mythbusting. Google explicitly states that site owners do not need to create llms.txt files, do not need to chunk content into small pieces, and do not need to rewrite content specifically for AI systems. Google’s position is direct: create non-commodity, people-first content. For Google Search, ignore tactics like chunking content, creating unnecessary AI text files like llms.txt, or pursuing inauthentic mentions.
A nuanced one worth understanding carefully, given the guide above. Google has added an experimental llms.txt audit to Chrome Lighthouse under a new Agentic Browsing category. The audit checks for a machine-readable summary at the domain root. Importantly, the audit does not produce a traditional 0 to 100 score, and a missing file returns ‘Not Applicable’ rather than a failure.
The distinction matters: Google Search says llms.txt is not needed for AI features. Chrome’s Lighthouse checks it for agentic browser readiness. These are different concerns for different use cases.
John Mueller answered a question about why some websites use multiple XML sitemaps instead of a single file. His answer: segmenting sitemaps by content type makes it significantly easier to diagnose crawling and indexing issues. When one section of a site has a problem, a segmented sitemap pinpoints exactly which content type or area is affected rather than requiring you to dig through a monolithic file.
Google Ads are increasingly appearing within AI Overviews, which means thinking about paid search purely in terms of traditional keyword coverage is no longer enough. Shopping, Performance Max, and AI Max for Search are the main campaign types to watch, but feed quality, landing page content, and creative all play a role in how ads perform within this surface.
Google’s latest AI ad updates point toward a significant shift in what an ad actually does. Rather than directing users to a landing page to find their answers, ads may begin handling part of the qualification process directly within Search. Business Agent for leads, for example, could let users ask questions and get responses directly within a Search Ad before they ever click through to a website.
Google has launched a built-in lead management dashboard inside Google Ads, giving advertisers a central place to track Google-hosted form leads, lead stages, and quality signals without needing to export data to a CRM or third-party tool.
Google appears to be testing new branded search controls in AI Max campaigns, which could let advertisers decide whether brand traffic is included, excluded, or handled separately. This is still in testing and has not been confirmed as a full rollout.
OpenAI is reportedly preparing conversion-focused ad formats for ChatGPT, with formats designed around specific actions like purchases, appointment bookings, and contact form submissions. Advertisers may also be able to use tracking tools such as pixels and API connections to measure results in a way that is more comparable to Google or Meta.
Meta has officially launched paid subscription plans for Instagram, Facebook, and WhatsApp, with more plans in testing for creators, businesses, and Meta AI users. Some of the business-focused plans could include improved profile visibility, enhanced analytics, stronger link options, and more ways to drive people to websites or shops.
Read more: TechCrunch
Step back from the individual updates and one theme runs through all of them: the gap between organic, paid, and AI search is closing fast. What earns you a citation in ChatGPT is connected to what gets your ads into AI Overviews, which is connected to how your content performs in AI Mode. These are no longer separate conversations.
The fundamentals have not moved. Helpful content, clear site architecture, accurate data, and a strong brand presence across the web are still the inputs everything else runs on. What is changing is how many surfaces those fundamentals now need to work across, and how quickly.
At Verkeer, staying across these shifts and translating them into practical action for our clients is what we do every day. If any of this month’s updates have raised questions about your own visibility, we would love to talk.