Despite the volume of commentary around AI search, there is still a great deal of uncertainty around how these systems surface brands and influence decision-making. At the same time, evidence is starting to emerge that user behaviour is changing in meaningful ways, particularly as AI-generated summaries and recommendation engines become more integrated into search platforms and workplace tools.
For B2B organisations, this creates a practical communications challenge rather than simply a technical SEO issue. Visibility is becoming less dependent on traditional rankings alone and more influenced by how consistently a brand’s authority, expertise and relevance are reinforced across the wider web.
The numbers are already striking
Ofcom research shows AI-generated summaries now appear in around 30% of Google searches. Magenta’s own research found that 66% of UK decision-makers already use tools like ChatGPT, Microsoft Copilot and Perplexity AI when researching suppliers. And audience intelligence software SparkToro found that nearly 60% of Google searches in Europe now end without anyone clicking through to a website at all, a figure that rises further when AI summaries are present.
For brands that have spent years measuring visibility through traffic, rankings and click-through rates, that’s a complicated picture. Increasingly, discovery is happening before a user ever reaches your website.
These tools are doing more than answering questions
It’s worth stepping back from thinking about ChatGPT or Claude purely as chatbots. Most of these platforms now include live web search, browsing capabilities, document analysis, product comparison features and workplace integrations. Some are moving towards agentic functionality, meaning they can interact with websites and software directly on a user’s behalf.
In practice, that changes what they’re actually doing. Instead of helping someone find information, they’re starting to interpret, filter and synthesise it for them.
A buyer no longer needs to run ten searches, open fifteen tabs and spend an afternoon comparing options. They can ask:
- “What are the best workplace management platforms for enterprise businesses?”
- “Which refrigeration providers specialise in pharmaceutical environments?”
- “Compare the leading ESG reporting platforms for UK manufacturers?”
The AI retrieves, compares and summarises, then hands back an answer. In some cases, it may go further: browsing websites, comparing pricing and reading reviews. The user is increasingly a step removed from that process.
So, what does this mean for brands?
This is where the concept of Generative Engine Optimisation (GEO) comes in. Broadly, it refers to improving a brand’s visibility within AI-generated answers and recommendations, rather than just search rankings.
There’s no agreed playbook for this yet, and anyone selling you a guaranteed “AI optimisation” fix is probably overstating what they know. The platforms themselves change constantly, and nobody outside the major AI companies fully understands how these systems prioritise and retrieve information.
That said, the fundamentals still apply. Technically sound websites, clear architecture, structured data, authoritative content, strong digital PR, topical expertise and reputable backlinks matter as much as they did before. But AI visibility adds something new on top of that.
Writing for two audiences at once
For the first time, brands need to genuinely write for two audiences simultaneously: humans and machines. Human readers need persuasion, clarity and trust. AI systems need information that’s structured, extractable and easy to interpret.
Vague marketing language, bloated copy and jargon-heavy pages become a liability when AI systems are trying to retrieve and summarise information quickly. Content that explains things clearly, backs up claims with evidence and presents information consistently across multiple sources is far more likely to surface in AI-generated responses and far more useful to real readers too.
It’s not just about your website
This is one of the bigger shifts, and it has real implications for PR and communications teams.
Modern AI search systems don’t just retrieve one result. Many break queries into multiple related searches, pulling from reviews, articles, business listings, forums, industry publications and third-party sources before generating a response. Google has described this as “query fan-out”.
What that means in practice is that AI visibility increasingly depends on the wider ecosystem of information around a brand, not just what’s on your homepage. Trade media coverage, analyst commentary, reviews, expert opinion and original research all help reinforce what AI systems understand about you, and whether you appear credible enough to cite or recommend.
Brand authority is no longer built solely through owned channels. It’s inferred from repetition, corroboration and consistency across the wider web.
Measurement is getting harder
Traditional SEO metrics were built around clicks and traffic. AI-generated search experiences make both harder to track.
A buyer might discover your brand through an AI-generated summary and never visit your website directly. They may see your company cited in a response, leave, and come back days later through a branded search or direct enquiry. Google has confirmed that AI Overview traffic is blended into existing Search Console reporting, which makes it difficult to isolate what’s actually driving performance.
Visibility is becoming more distributed just as attribution becomes less precise. That’s uncomfortable for teams used to clean reporting, but it’s the reality most are now working with.
What this all means
Search isn’t disappearing. Google still dominates discovery and will do for some time. But the old model – rank highly and win clicks – is giving way to something broader and harder to reduce to a single metric.
Visibility increasingly depends on whether AI systems can understand your brand clearly, find consistent signals of authority across the web, and retrieve information that feels trustworthy enough to include in an answer. That’s partly a technical challenge. But it’s also, fundamentally, a communications one.
AI isn’t just another channel in the B2B buyer journey – it’s quietly rewriting the whole script. Read our event summary blog on protecting brand visibility, and our guide, Search Forward: How AI is reshaping the B2B buyer, to uncover how tools like ChatGPT, Copilot and Perplexity are influencing who buyers discover, what they trust and how they decide.
