Why winning at AI visibility means looking at the bigger picture - and why most GEO (Generative Engine Optimisation) advice gets this wrong.
When talking about AI, one question is now central to almost every client conversation we're having: how do we become visible in AI-generated search, and how do we prove it's working?
The answers are more complex than most GEO content circulating right now suggests. The measurement traps are already forming. We want to give you the straight picture.
Content versus communication
You can't optimise your way into a single answer
There's a version of Generative Engine Optimisation that's simple, reassuring, and largely insufficient. It goes something like this: find the pages your buyers care about most, restructure the content to lead with clear answers, add schema markup, and wait for citations to roll in. Tick the technical boxes, watch the numbers move.
The problem isn't that this work is wrong, much of it is genuinely useful. The problem is that it treats AI visibility as a content optimisation challenge, when it's actually a communications challenge. And the two demand fundamentally different responses.
The importance of earned media
Research comparing Google's AI Mode and AI Overviews found that only 13.7% of their cited URLs overlap[1]; yet they reach the same conclusions 86% of the time. Different sources, same answer. This reflects how modern AI systems work: they synthesise signals from multiple authoritative sources, building a weighted consensus rather than retrieving one "best" result.
Owned media accounts for just 13.7% of AI citations; 84% come from earned media, with journalism and corporate blogs making up more than half. No matter how well a single page is optimised, if the wider signal environment doesn't support the same narrative, AI systems will look past it. This works directly against the "one great piece of content" logic still driving too much industry thinking.
It's a real challenge for communicators who've built strategies around the prestige placement or flagship launch. Those still matter, just no longer enough on their own.
What AI systems reward is what good B2B communicators have always known how to build: consistent, expert-led, multi-source credibility across the ecosystems buyers touch points - expressed over time, not in a single burst.
96% of B2B companies have near-zero AI visibility (2x AI Visibility Index, 2026)
A modern AI visibility programme for industrial B2B
Five interconnected layers. No single one is enough on its own
At EMG, we have built our AI Visibility Programme around five layers, specifically designed for global industrial B2B companies in chemicals, materials science, plastics processing, packaging, and industrial automation. Each layer works individually. The power comes from their integration.
- Diagnose - know where you stand before you act. We benchmark your AI visibility first: querying ChatGPT, Perplexity, Google AI Overviews and Copilot with the real questions buyers ask, mapping the sources feeding AI consensus in your sector, and auditing accuracy against competitors. Almost no quick-fix GEO tool does this groundwork.
- Build Authority - content structured for AI retrieval and human value. Not a content refresh. We unlock existing technical assets; application notes, processing guides, performance dataetc. From gated PDFs into formats AI can find, read, and cite. Embedded statistics, expert quotes, and third-party citations each measurably lift citation probability.
- Expand Footprint - cross-platform consensus across earned media. Citations on four or more authoritative platforms raise the odds of AI citation by 280%. We build presence in the trade publications, standards bodies, analyst citations, and LinkedIn content that feed your buyers' AI tools.
- Expert & Entity - expert positioning and category ownership. AI systems learn to associate named experts with topics over time. We build that through expert profiling, thought leadership, clear entity definition, and spokesperson training - making technical experts citable, not just quotable.
- Measure & Adapt - ongoing monitoring and share of AI voice. AI visibility isn't a one-off project. Quarterly audits, citation tracking, and feedback loops keep content current and gaps closed.
All this brings us to a point the industry urgently needs to hear.
80% of buyers rely on zero-click AI results in at least 40% of searches (Bain, 2025)
Measuring GEO
The 7 AMEC GEO Principles and why rigour matters more than reassurance
The measurement conversation around AI visibility is at an inflection point and the early signs are not all encouraging.
The industry is already seeing a rush to celebrate GEO metrics that are dangerously similar to AVEs. Citation numbers and "visibility percentages" are already being raised as the silver bullet, but visibility metrics without connection to awareness, trust, behaviour, or business outcomes risk becoming the next generation of vanity metrics.
A handful of queries, run once, in one language, on one platform, at one point in time - even if the numbers look good - tells you almost nothing reliable about your actual AI presence.
This is why EMG welcomes AMEC's new framework for GEO measurement[2] - and we have adopted its principles in our own client work.
AMEC's 7 GEO Principles:
- AI-led discovery should be measured against communication objectives and stakeholder information needs, not as a standalone metric.
- GEO measurement must assess the upstream information environment before interpreting AI-generated answers.
- Search and content readiness should be evaluated as evidence of whether reliable information can be found, understood, and cited.
- Observed AI outputs are directional indicators and should be tested transparently across tools, prompts, markets, languages, and time.
- GEO measurement should distinguish visibility from outcomes, and connect AI discovery to awareness, trust, behaviour, and impact.
- Reliable, trustworthy, and current sources matter more than volume, promotion, or short-term visibility.
- Ethical GEO improves the public information environment, it should not manipulate, disguise, or flood it.
At EMG, our measurement framework tracks share of AI voice, citation source quality, accuracy monitoring, and the downstream connection to buyer engagement and commercial outcomes, not citation counts in isolation.
3x faster AI search adoption among B2B buyers vs. consumers (Forrester, 2025)
Featured Spotlight
Stuart Bruce - PR Futurist, Founder of Purposeful Relations, earns our website newsletter spotlight this quarter. In his newsletter, Stuart posted:
“Communications is standing on a precipice. Historically, we have not been an innovative industry. Both the scale and the speed of AI adoption needs to be a lot faster. AI is impacting you because organisations around you are using it - your competitors, your peers, your users, your customers.
If you’re not understanding those shifts that are happening in society, in the economy, you are going to be left behind. We’ve got to get our act together quickly.’
Source: PR Futurist May newsletter
In closing
AI hasn't invented a new discipline - it's given us a powerful new tool, and a precise way to prove that good PR makes a measurable difference.
The brands that win won't be the ones with the most optimised landing page. They'll be the ones with genuine, broad-based credibility across their buyers' ecosystems, built through consistent, expert-led communication over time. That's what EMG helped industrial B2B companies do for 35 years; AI has only raised the stakes for getting it right.
Contact our team if you'd like to know where your organisation stands in AI-generated search, our AI Readiness Audit is the place to start: a factual, scored diagnostic with a prioritised roadmap, delivered in two weeks.
