The digital marketing industry is having one of those moments where everyone pretends they saw it coming.

Agencies that spent years selling page-one rankings, keyword maps, backlink packages and monthly “visibility reports” are now staring at a search environment where the answer may appear before the website.

The citation may matter more than the click, and the customer journey can begin and end inside a machine-generated summary.

It has turned a once-familiar game into something stranger, faster and far less forgiving.

For years, the rules were relatively simple. Build a website. Optimise the pages. Chase rankings. Publish enough content to look alive. Win links. Report traffic. Repeat until the client either renewed the contract or lost patience.

That model is now under serious pressure.

Google’s AI Overviews, AI Mode, ChatGPT, Perplexity, Gemini and other answer engines are not just sending users to websites. Increasingly, they are reading the web, summarising it, selecting sources, and deciding which brands appear credible enough to mention.

For marketers, that changes almost everything.

• The old question was: “Where do we rank?”
• The new question is: “Are we being cited at all?”

The Click Is No Longer Guaranteed

The first shock has been traffic.

Pew Research Center analysed tens of thousands of Google searches and found that users were less likely to click through to websites when an AI summary appeared in the results. In searches with an AI summary, users clicked a traditional search result in 8% of visits.

Without an AI summary, that figure was 15%

That is not a minor wobble. That is a commercial warning light.

Even more uncomfortable for publishers, brands and agencies: users clicked links inside the AI summary itself in only 1% of visits.

In other words, being included is useful, but it does not automatically mean traffic will arrive.

This is the part many marketing decks still dance around. AI search can deliver brand exposure, credibility and influence, but it can also remove the click that made the old reporting model easy to sell.

For agencies built around organic traffic graphs, that is a problem. A big one.

The New Authority Game

The second shock is that AI systems do not appear to treat every type of content equally.

Muck Rack’s May 2026 Generative Pulse study found that earned media accounted for 84% of AI citations across ChatGPT, Claude and Gemini responses. Journalism alone made up 27% of cited sources. Paid and advertorial content accounted for just 0.3%

That finding should be pinned to the wall of every agency still telling clients that a bundle of sponsored blog posts on low-grade “high authority” websites will solve their AI visibility problem.

The machines appear to be sniffing out credibility.

Not perfectly. Not always. But enough to make the old link-selling economy look increasingly tired.

This is where the digital marketing industry starts to split into two camps.

One camp is doing the hard work: proper research, original reporting, technical SEO, schema, brand building, expert commentary, digital PR, content quality, entity optimisation and credible third-party coverage.

The other camp has discovered a new menu of buzzwords.

GEO. AEO. LLMO. AI SEO. AIO optimisation. Search everywhere optimisation. Answer engine optimisation. Citation engineering.

Some of it is useful. Much of it is traditional SEO wearing a metallic jacket and pretending it has just stepped out of a venture-capital pitch meeting.

Agencies Are Being Forced Back To School

The better agencies know the ground has shifted.

They are learning how AI systems select citations, how brand entities are understood, how publisher mentions influence answer engines, how technical accessibility affects AI crawlers, and how to measure visibility when the referral traffic is incomplete or hidden.

They are also learning a harder lesson: content volume is no longer a moat.

The internet is now drowning in cheap articles, synthetic “expert” quotes, recycled research, fake thought leadership and mass-produced LinkedIn sludge. Publishing more words is not a strategy if the words say nothing.

HubSpot’s 2026 State of Marketing report found that 61 per cent of marketers believe AI has created the industry’s biggest disruption in 20 years.

It also found that 80 per cent of marketers use AI for content creation and 75 per cent use it for media production.

That explains the noise.

Everyone has the tools. Far fewer have the judgement.

The agencies that will survive the shift are not the ones producing the most content.

They are the ones producing the clearest evidence of trust: original data, named experts, editorial mentions, accurate technical pages, consistent brand signals and content that answers real questions better than the next 20 results.

The Buzzword Milk Run

Of course, no industry can experience disruption without someone immediately building a three-tier pricing package around it.

Across the market, agencies are now selling “AI search optimisation” audits that look suspiciously like SEO audits with three extra slides.

Some are charging for “LLM visibility strategies” that amount to little more than asking ChatGPT whether a brand appears in a generated answer. Others are repackaging digital PR as “citation acquisition” and acting as though journalists were invented last Tuesday.

There is a useful service underneath some of this. Brands do need to understand how they appear in AI-generated answers. They do need to monitor mentions across AI systems. They do need stronger editorial authority. They do need technically clean websites that machines can understand.

But the hype is running ahead of the discipline.

The uncomfortable truth is that much of this “new” work still depends on old fundamentals: authority, trust, clarity, relevance, crawlability and proof.

A weak brand with thin content will not magically become a trusted AI citation because an agency has renamed its SEO package “Generative Engine Optimisation Pro Max.”

The Measurement Problem

The other challenge is measurement.

Traditional SEO reporting was never perfect, but it had familiar numbers: rankings, impressions, clicks, sessions, conversions, backlinks and domain metrics.

AI search muddies the water.

A brand may be cited inside an AI answer without receiving a click. A customer may ask ChatGPT for a shortlist, compare products inside an answer engine, then arrive later through direct traffic, branded search or even a sales call.

The influence happened, but the attribution trail is broken.

Semrush has estimated that AI search traffic in digital marketing and SEO-related topics may overtake traditional search traffic by early 2028.

It also reported that AI search visitors, where tracked, were worth 4.4 times the average traditional organic search visitor based on conversion rate.

That is the paradox.

There may be fewer casual clicks, but the clicks that do arrive can be more qualified. Users who come from AI systems may already be further through the decision process.

They have asked the question, received the shortlist, compared options and clicked only when they are closer to action.

For clients, this means the old obsession with raw traffic needs to calm down. Visibility, citation share, branded demand, assisted conversions, direct enquiries and source credibility are becoming more important measures of performance.

Google Has Not Left The Building

Despite the excitement around ChatGPT and Perplexity, Google remains central to the story.

Google said in May 2026 that AI Mode had surpassed one billion monthly users, with queries more than doubling every quarter since launch. The company also described its new AI-powered Search box as the biggest upgrade to Search in more than 25 years.

That is not a side experiment. That is the main stage.

Recent academic research has also found that AI Overviews can appear frequently, especially for question-based searches.

One study of trending queries found AI Overview activation at 13.7 per cent overall, rising to 64.7% for question-form queries. Another study using real-user representative queries found AI Overviews generated for 51.5% of queries.

The direction is clear. Search is becoming less like a directory and more like a front-page editor with an algorithmic brain.

That raises enormous questions for publishers, businesses and the agencies advising them:

Who gets cited? Who gets ignored? What happens when the answer is wrong? How does a small business compete when a machine selects the sources before the user sees the open web?

Nobody has fully solved those questions. Anyone claiming otherwise is probably selling a course.

The Winners Will Look More Like Editors Than Hackers

The next phase of digital marketing will reward a different kind of operator.

• Less spammer. More editor.
• Less keyword stuffing. More evidence.
• Less fake authority. More genuine expertise.
• Less “we can get you 100 placements.” More “we can make your business worth citing.”

That is the major shift. AI search has not killed SEO. It has exposed the weakest parts of it.

The shortcuts are becoming more obvious. Thin content looks thinner. Fake experts look faker. Paid link farms look like what they always were: artificial credibility machines built for an older version of Google.

The agencies that understand this are already changing their language and their methods. They are combining technical SEO with digital PR, newsroom-style content, data-led reporting, authority building and AI visibility tracking. They are not abandoning search. They are broadening it.

The agencies that do not understand it are still sending cold emails offering “high DA guest posts” as though the market has not moved on.

It has.

AI has not ended digital marketing. It has made it harder to fake.

And for an industry that has spent years rewarding volume, shortcuts and jargon, that might be the most entertaining development of all.