AI Search, zero-click SEO and Google’s May update

Google’s had a busy month.

In May, it published new guidance on how websites can appear in generative AI features such as AI Overviews and AI Mode, announced a bigger shift towards AI-led Search at Google I/O, began rolling out the May 2026 core update, AND it expanded features such as Preferred Sources, designed to help users see more content from publishers and websites they actively trust.

Cue the usual panic.

SEO is dead.
Clicks are dead.
Ranking is dead.
Everything is AEO now.

Except, that’s not quite what Google’s saying.

Google’s own guidance is much less dramatic and much more useful: the fundamentals of SEO still matter, but search behaviour is changing. AI search experiences are becoming more conversational, more answer-led and more selective about which sources they surface.

For website owners, the biggest change might not be to their rankings, but something subtler: more searches ending without a click.

Your content might still appear. Your impressions might even rise. But if Google answers more of the question directly in AI Overviews or AI Mode, fewer users may need to visit your site straight away.

That makes SEO a little more complex.

Here’s what the latest Google updates mean, why zero-click search matters, and what you should do next.

What changed in May?

There were a few separate Google updates and announcements in May, so it’s worth untangling them.

First, Google published its official guide to optimising for generative AI features in Search. This covers how websites can perform in AI Overviews and AI Mode, and it makes one thing clear: AI search visibility still depends on strong SEO foundations.

Then, at Google I/O, Google announced a broader shift towards a more AI-led Search experience. Its new AI Mode capabilities are designed to help users ask longer, more complex questions, refine their searches and complete tasks directly within Google.

Google also began rolling out the May 2026 core update on 21 May 2026. Core updates are broad changes to Google’s ranking systems, designed to improve the quality and relevance of search results.

Finally, Google announced new ways to help users find original, high-quality content, including Preferred Sources appearing within AI Overviews and AI Mode.

So yes, things are changing.

But this isn’t a single “AI update” that wipes out SEO overnight. It’s a series of changes that all point in the same direction: Google wants to surface useful, trustworthy, original content in increasingly AI-shaped search experiences.

Why zero-click search is the bit brands will feel first

The biggest practical change for most website owners won’t be a sudden disappearance from Google.

It will be a reporting pattern that looks more confusing than it used to.

You may see impressions increase while clicks stay flat. You may see rankings hold while click-through rate drops. You may see informational blogs generating fewer visits, even though the content is still being surfaced, summarised or used to support an AI-generated answer.

That’s zero-click search.

A zero-click search happens when someone gets what they need from the search results page without clicking through to a website. This has been happening for years through featured snippets, knowledge panels, maps, etc. But AI Overviews and AI Mode push that behaviour further because Google can now summarise fuller answers directly in the search experience.

Pew Research Center found that users who saw an AI summary clicked a traditional search result in 8% of visits, compared with 15% of visits where no AI summary appeared. In other words, when an AI answer appears, users are roughly half as likely to click through to a traditional result.

A 2026 study of Google AI Overviews also found that AI Overviews appeared for 13.7% of trending queries overall, rising to 64.7% for question-form queries. And question-led, informational searches are often exactly the kind of queries brands target with blogs, guides and educational content.

So this isn’t a brand-new behaviour. But AI makes it more visible, more powerful and more relevant to SEO reporting.

For brands, that means a drop in organic clicks doesn’t always mean your SEO has failed. Sometimes it means the search results page is doing more of the work before a user reaches your website.

That’s frustrating, yes. Especially for publishers, blogs and advice-led content that rely heavily on informational traffic.

But it also means SEO measurement needs to get more sophisticated.

Instead of only asking, “How many clicks did this page get?”, brands should also ask:

Are we appearing for more relevant searches?
Are impressions increasing across valuable topics?
Are branded searches growing?
Are users coming back later through direct, paid, social or referral channels?
Are organic visitors more qualified when they do click?
Are we being cited, referenced or recognised as a trusted source?
Are conversions holding even if top-of-funnel traffic softens?

Zero-click search doesn’t make SEO pointless. It changes the job SEO needs to do.

Some content will still need to win the click. Commercial pages, comparison pages, product pages, case studies and high-intent service pages all still need to bring users onto your site and move them towards action.

But some content will increasingly work higher up the journey. It may build brand recognition, answer early questions, support AI visibility and help your brand become part of the decision set before someone’s ready to enquire.

The goal here is useful visibility that creates trust, demand and eventual action.

The bigger zero-click lesson from Rand Fishkin

The rise of zero-click search raises an obvious question: if fewer people are clicking through to websites, what does that mean for websites?

Rand Fishkin, co-founder of SparkToro and one of the most vocal commentators on zero-click search, argues that the issue is bigger than declining referral traffic from Google.

His view is that brands can no longer build their entire marketing strategy around platforms sending users back to their website. Google, social platforms and AI tools increasingly answer questions, summarise information and keep users within their own environments.

That doesn’t mean websites no longer matter. It means they need to become more valuable.

Websites can no longer succeed simply by publishing information that search engines and AI assistants can easily summarise. On top of this, they need to also deliver experiences, interactions and brand value that cannot be replicated in a search result.

In a SparkToro article on surviving the zero-click era, Fishkin argues that the websites performing best are often those that provide something users cannot get directly from Google. That might be a product, service, proprietary asset, interactive tool, community, original data, subscription offering or distinctive brand experience.

In other words, the question is: “What do we offer that’s still worth visiting?”

That’s a much more useful way for brands to think about AI Search.

If your website only answers simple informational queries, it’s more vulnerable to zero-click behaviour. But if it helps users complete a task, compare options, explore original data, book a service, use a tool, access expertise or engage with a trusted brand, there’s still a compelling reason to visit.

This is where SEO, brand and website strategy begin to overlap.

Strong organic visibility remains important. But in a zero-click environment, traffic is no longer the only measure of influence. Your content may shape what people see in Google, LinkedIn, Reddit, YouTube, ChatGPT, Perplexity or Gemini before they ever reach your website.

That means brands need to think beyond publishing content and waiting for traffic. They need to distribute ideas across the channels where their audience already spends time, build recognition before the click, and create experiences that make visiting their website worthwhile.

The 10 blue links haven’t disappeared, but they’ve lost centre stage

Traditional search was built around a familiar pattern: type a query, scan the results page, click a link. That pattern still exists, but it’s no longer the whole experience.

AI Overviews and AI Mode can answer questions directly in the search results, summarise information from multiple sources and let users ask follow-up questions without starting a new search.

Google still surfaces links. Its own documentation says AI Overviews and AI Mode include supporting links and can help users discover relevant pages across the web.

But links are no longer the whole search experience.

They now sit inside a broader answer system, alongside summaries, citations, product results, local information, images, videos, follow-up questions and, increasingly, AI-led actions.

For website owners, that means ranking alone may not guarantee the same level of traffic it once did.

SEO hasn’t been replaced by AEO/GEO

You’ll probably see a lot of people talking about AEO, GEO, LLM optimisation or AI visibility.

Some of that language is useful. Some of it is just old SEO advice with a fresh acronym and a LinkedIn carousel.

Google’s position is pretty clear: SEO is still relevant for generative AI Search because AI Overviews and AI Mode are rooted in Google’s core Search ranking and quality systems.

That means the basics still matter:

Your site needs to be crawlable.
Your pages need to be indexable.
Your content needs to be genuinely useful.
Your technical foundations need to be clean.
Your structured data should match what users can actually see on the page.
Your site should offer a good user experience.

You don’t need to chase secret AI hacks. Google says you don’t need AI-specific schema, special markup or an llms.txt file to appear in AI Search features. You need a website Google can access, understand and trust.

Generic content is losing its hiding place

This is the biggest shift for most brands.

Google’s AI optimisation guidance repeatedly points towards useful, unique, non-commodity content. In plain English: content that says something worth saying.

That’s bad news for websites filled with:

Thin service pages
Rewritten competitor content
Generic “ultimate guides”
AI-generated filler
Blogs written purely to target keywords
Advice with no examples, evidence or expertise
Category clichés dressed up as insight

AI can already generate generic information very quickly. So if your content sounds like everyone else’s, it becomes easier to ignore.

What performs better is content with information gain.

That might mean:

First-hand experience
Original research
Expert commentary
Internal data
Customer insights
Case studies
Strong points of view
Practical examples
Sector-specific advice

For example, for us, a generic blog on “how to improve website conversions” is easy to replicate.

A blog showing what we learned from redesigning a complex B2B SaaS website, which pages mattered most, what messaging changed, what happened after launch and what we’d do differently next time is much harder to copy.

That’s the kind of content AI systems have more reason to cite.

It’s also the kind of content humans actually want to read.

E-E-A-T still matters

Google has talked about E-E-A-T for years: experience, expertise, authoritativeness and trustworthiness.

In an AI-led search environment, those signals become even more important.

If Google is going to summarise information and recommend sources, it needs confidence in where that information comes from.

That means brands should make expertise visible.

Add author bios where they’re useful. Show who created or reviewed the content. Use expert quotes. Include evidence. Reference reputable sources. Link to original research. Show real project experience. Keep content up to date.

This matters even more for topics where accuracy has a real-world impact, such as finance, legal, healthcare, security, sustainability and technical decision-making.

Helpful content doesn’t just answer the question.

It proves why the answer should be trusted.

Technical SEO still matters

AI Search might sound futuristic, but it still depends on very practical technical foundations.

Google needs to crawl your site. It needs to index your pages. It needs to understand what each page is about. It needs to process your headings, links, media, schema and visible page content.

So no, technical SEO hasn’t gone away.

If anything, machine readability matters more when AI systems are retrieving and synthesising information from across the web.

A few useful checks:

Can Google crawl your important pages?
Are key pages indexable?
Do your pages have clear, descriptive titles and headings?
Is your internal linking structure logical?
Does your structured data match the visible content?
Are images and videos properly optimised?
Is important content visible on the page, not hidden in inaccessible formats?
Is the site fast, usable and stable across devices?

The goal is simple: make it easy for people and machines to understand your site.

Brand authority is becoming a search advantage

One of Google’s most interesting May updates was the expansion of Preferred Sources into AI Overviews and AI Mode.

Preferred Sources allow users to choose the publishers and websites they want to see more often. Google says users are twice as likely to click through to a Preferred Source. That matters because it connects brand trust directly with search visibility.

If someone recognises your brand, trusts your content or actively prefers your website, you have a better chance of winning attention when Google presents multiple sources.

This is where SEO and brand strategy start to overlap.

Strong brands tend to earn:

More branded searches
More direct traffic
Better recall
More backlinks
More press mentions
More repeat visits
More trust at the point of decision

You can’t separate brand from search anymore.

Being visible is useful. Being recognised is better.

And in a zero-click environment, that recognition matters before someone ever reaches your website.

Search behaviour is fragmenting beyond Google

Google still dominates search. According to StatCounter, Google held around 91% of the UK search engine market in April 2026, with Bing at just under 6% and DuckDuckGo at less than 1%.

So no, brands shouldn’t suddenly abandon Google. But it would be a mistake to assume every search journey begins and ends there.

As Google adds more AI-generated summaries and answer-led experiences, some users are looking for alternatives. DuckDuckGo, which positions itself as a privacy-focused Google alternative, reported a rise in app installs after Google’s latest AI Search announcements. Business Insider reported that DuckDuckGo’s US app installs increased by an average of 20.8% week on week after Google’s AI Search changes, while iOS installs rose by an average of 33%.

That suggests that some users want more control over how they search, especially around privacy, tracking and AI-generated answers.

But search is also fragmenting in another way. People are increasingly searching across platforms that don’t behave like traditional search engines at all. They use YouTube to learn, Reddit to compare real opinions, TikTok to discover products, LinkedIn to assess expertise, Perplexity or ChatGPT to explore topics, and Amazon or marketplaces when they are closer to buying.

Your audience’s discovery journey is no longer contained within one search results page.

A potential customer might first see your brand mentioned in a LinkedIn post, compare opinions on Reddit, ask ChatGPT for options, search your name on Google, read a case study, then return directly weeks later. If you only measure Google organic clicks, you will miss a lot of that journey.

Obviously, the practical takeaway is not to optimise for every platform at once. That way lies chaos. It’s to understand where your audience actually searches, researches and validates decisions, then make sure your brand shows up with useful, credible information in those places.

For some brands, that might mean YouTube explainers. For others, it might mean LinkedIn thought leadership, digital PR, Reddit visibility, comparison pages, industry directories, Google Business Profile optimisation, or stronger presence in AI answer engines.

Google still matters enormously. But search behaviour is becoming more distributed, more conversational and more trust-led. Your SEO strategy needs to account for that.

What should brands do now?

Don’t throw your SEO strategy away.

Upgrade it.

1. Separate visibility from traffic in your reporting

Clicks still matter, but they won’t tell the whole story.

As AI answers become more visible, brands need to look at a broader set of signals:

Organic clicks
Rankings
Impressions
CTR
Branded search growth
Engagement quality
Assisted conversions
Direct traffic
Referral traffic
AI citation visibility where measurable
Share of voice across search and AI platforms

A lower CTR does not automatically mean SEO is failing.

But visibility without commercial impact is not enough either.

The job is to connect search visibility with brand demand, trust and real business outcomes.

2. Audit your content for originality

Look at your key service pages, blogs and guides.

Ask:

Does this say anything new?
Could any competitor have written this?
Does it include first-hand experience?
Does it answer the question clearly?
Does it show evidence?
Does it help a real person make a better decision?

If the answer is no, the page probably needs more than keyword tweaks. It needs substance.

3. Structure content around real questions

AI Search is built for more conversational queries.

That doesn’t mean stuffing pages with FAQs. It means understanding how your audience actually searches when they’re trying to make a decision.

Use clear headings. Answer questions directly. Add short summaries where useful. Break complex topics into logical sections. Make pages easy to scan without flattening the thinking.

Good structure helps readers. It also helps search systems understand your content.

4. Make your expertise easier to see

If your team has deep experience, show it.

Add expert commentary, case studies, author credentials, project examples, client outcomes, original data and clear explanations of how you know what you know.

This is especially important for B2B brands, where buyers are often researching complex, high-consideration decisions.

5. Strengthen your technical foundations

Before chasing AI visibility, check the basics.

Google can’t cite what it can’t crawl, index or understand.

Make sure your technical SEO, structured data, internal links, page speed, media optimisation and content accessibility are in good shape.

6. Understand where your audience actually searches

Google is still the dominant search engine, but it’s no longer the only place people go to research, compare and validate decisions.

Look at where your audience finds information before they contact you. That might include Google, Bing, DuckDuckGo, YouTube, Reddit, LinkedIn, TikTok, industry directories, review sites, marketplaces or AI answer engines such as ChatGPT and Perplexity.

The goal is not to chase every channel. It’s to identify the places that genuinely influence your buyers, then make sure your brand appears there with useful, credible and consistent information.

7. Build brand authority beyond your website

AI Search pulls from the wider web, not just your site.

That means your authority is shaped by what exists around your brand too.

Think digital PR, thought leadership, partnerships, reviews, social proof, third-party mentions, industry commentary and high-quality backlinks.

A strong website is the crucial foundation. But a strong wider footprint matters too.

What this means for your website

Google’s May updates are absolutely not a reason to panic. They’re a reason to look more honestly at how search is changing.

AI Search will make some queries more zero-click. Users will get more answers directly from Google, and that may reduce traffic to certain pages, especially broad informational content.

But that doesn’t mean SEO has stopped working. It means SEO has to work across more of the customer journey.

Your website still needs to rank. It still needs to earn clicks. It still needs to convert visitors when they arrive. But your content may also need to build visibility before the click, give Google clear information to retrieve, and make your brand recognisable enough to choose when users do decide to visit a source.

The practical response is to strengthen the things that make your website useful in any search environment. That means clear answers, original insight, visible expertise, technical accessibility, strong internal structure and a brand people trust.

Zero-click search makes weak content easier to ignore. But it also gives useful, credible brands more ways to show up earlier in the decision-making process.

So the question is: “How do we become the source our audience (and search systems) trust enough to use?”

That’s where SEO is heading (and has been heading for a long time).