AMEC GEO Principles: why AI visibility needs evidence, not hype

The launch of the AMEC GEO Principles is an important moment for communications, research and reputation measurement.

It matters because generative engine optimisation, or GEO, is already moving quickly from an emerging discipline into a commercial priority. More organisations are asking how they appear in AI-generated answers, whether they are being recommended, how competitors are being described, and what they can do to improve visibility in tools such as ChatGPT, Perplexity, Google AI Overviews, Copilot and Gemini.

That interest is understandable. AI is changing how people find information. It is also changing how they form impressions, compare options and decide what deserves further attention.

But this is also where risk starts to creep in.

Whenever a new visibility channel emerges, there is a temptation to reduce it to a score, a ranking or a tactical game. GEO could easily become another version of that: a narrow race to appear more often in AI answers, regardless of whether the information being surfaced is accurate, useful or meaningful.

The AMEC GEO Principles are important because they push the conversation in a better direction.

They make clear that AI-led discovery should not be measured simply by whether a brand appears. It should be assessed by whether stakeholders encounter information that is accurate, credible, current and trustworthy. That is a much higher bar, and a more useful one.

At Sensu Insight, this aligns closely with what we are seeing in our own research into AI and decision-making. The issue is not just that AI tools are becoming another place where brands may or may not appear. The bigger shift is that AI is starting to influence what people consider in the first place.

AI is shaping decisions earlier than many organisations realise

A lot of discussion around GEO focuses on visibility.

• Is the brand mentioned?
  
• Is it cited?
  
• Does it appear ahead of competitors?
  
• Which sources are being used?

These are important questions. But they only become meaningful when connected to how people actually make decisions.

Our research found that 31% of consumers already use AI when researching products and services, rising to 49% among those aged 25 to 44. In business contexts, the shift is even more pronounced, with 74% of professionals using AI in their role. 

The influence is strongest early in the decision-making process.

Among business decision-makers, 32% say AI has the greatest influence when identifying possible solutions, and 28% say it matters most when comparing options. Only 5% say AI has the greatest influence at the final decision stage. 

That distinction matters.

AI is not necessarily replacing decision-makers. It is shaping the field of options they consider. It helps define the problem, introduces possible solutions, frames comparisons and narrows the list of organisations that are explored further.

In consumer markets, the pattern is similar. AI is increasingly used to research, compare and shortlist products and services. Around 36% of consumers say AI-generated answers influence their decisions at least sometimes, and 18% say they have chosen a product, service or brand based on an AI recommendation. 

This means GEO is not just a technical search issue. It is becoming part of brand awareness, consideration and reputation.

If an organisation is absent, poorly described or inaccurately compared at the point where options are being formed, it may never reach the later stages of the decision process.

The real GEO question is not “are we visible?”

The simplest version of GEO asks:

Are we appearing in AI-generated answers?

That is useful, but incomplete.

The better questions are:

• Are we being understood accurately?
  
• Are we being associated with the right topics?
  
• Are we appearing for the questions that matter to our audiences?
  
• Are credible sources supporting the way we are described?
  
• Are AI tools reflecting our current position, or an outdated version of us?
  
• Are competitors being framed more clearly or more credibly?

This is where the AMEC GEO Principles are especially valuable. They move the discussion away from isolated AI outputs and towards a broader evidence base.

The principles identify three connected evidence domains: the upstream information and reputation environment, search and content readiness, and downstream AI output tracking. 

That structure is important because AI-generated answers are only the visible end point.

They are shaped by a much wider information environment, including media coverage, owned content, expert commentary, reviews, forums, public records, stakeholder discussion and other sources that help define what appears credible and relevant.

In other words, AI visibility is not created only inside AI platforms. It is shaped by the information those systems can find, interpret and trust.

GEO should start with the public information environment

One of the biggest misconceptions about GEO is that it is mainly about optimising prompts or trying to influence AI answers directly.

That view is too narrow.

AI systems do not create brand understanding from nothing. They draw on what already exists across the public domain. They interpret it, compress it and present it back as a simplified answer.

That means the most important question for many organisations is not:

How do we make AI say something different?

It is:

What does the available evidence already say about us?

This is what we describe as the public information environment: the combined body of information that shapes how an organisation is discovered, understood and compared.

It includes a company’s own website and content, but it is not limited to it. It also includes trade media, national media, sector commentary, customer reviews, social discussion, employee perspectives, partner references, industry listings, thought leadership and competitor context.

If that environment is clear, consistent and well supported, an organisation is easier for AI tools to interpret.

If it is fragmented, outdated or contradictory, the risk of weak or inaccurate representation increases.

This is why GEO belongs as much in reputation research and communications strategy as it does in search optimisation.

Visibility without credibility is a weak measure

The AMEC principles also arrive at a useful moment because the GEO market is already becoming noisy.

A growing number of tools can show whether a brand appears in AI-generated answers. Some simulate prompts across different platforms. Others monitor citations, source patterns or answer variation.

These tools can be valuable. But they do not remove the need for interpretation.

The accompanying AMEC practitioner guidance is clear that downstream AI outputs should be treated as directional evidence. Results vary by platform, query, user context, market and time. They can show patterns, but they cannot provide a definitive performance score on their own. 

That reflects what we see in practice.

The same prompt can produce different brand recommendations across different AI platforms. A brand may appear strongly in one environment and barely appear in another. Some tools rely more heavily on live retrieval. Others reflect broader semantic associations, search authority, citations or conversational relevance.

That does not make measurement impossible. It means measurement needs to be designed carefully.

A credible GEO audit should include:

• a governed library of prompts linked to stakeholder questions
  
• branded and unbranded queries
  
• clear market, language and audience definitions
  
• repeat testing over time
  
• evidence of sources, citations and answer framing
  
• risk flags for inaccuracies, omissions or outdated claims
  
• analysis of the upstream information environment
  
• a clear distinction between visibility, outcomes and impact

Without this, organisations risk mistaking a snapshot for a strategy.

Ethical GEO matters because AI answers shape trust

One of the most important aspects of the AMEC GEO Principles is the ethical guardrail.

This says GEO should improve the availability of accurate, useful and verifiable information. It should not involve flooding the web with low-quality content, disguising promotion as independent evidence, manipulating reviews or forums, or treating AI outputs as factual without verification. 

That matters because AI discovery does not only affect organisations. It affects the quality of information available to everyone.

If GEO becomes a manipulation exercise, the information environment gets worse. Search results become more cluttered. Reviews become less trustworthy. Forums become less useful. AI-generated answers become more vulnerable to distortion.

The better version of GEO is different.

It helps organisations explain themselves more clearly, correct outdated information, provide evidence for their claims, strengthen credible third-party validation and make useful content easier to find and interpret.

That is good for organisations, but it is also good for audiences.

It means people are more likely to encounter information that helps them make informed decisions.

What organisations should do now

The practical implication of the AMEC GEO Principles is that organisations should avoid rushing straight to tactics.

Before asking how to improve AI visibility, they should understand how they are currently represented.

That means looking at three connected areas.

1. How the organisation appears in AI-generated answers

This includes whether the organisation is mentioned, how prominent it is, how it is described, which competitors appear alongside it, what sources are cited, and whether there are inaccuracies or omissions.

This should be measured across relevant platforms, prompts, markets and audiences. It should also be repeated over time, because AI outputs are not static.

2. How strong the upstream information environment is

This means reviewing the public evidence base that AI tools are likely to draw from.

• Are there clear explanations of what the organisation does?
  
• Is the language consistent across owned and third-party sources?
  
• Is there credible media coverage or expert commentary?
  
• Are reviews, forums or social discussions reinforcing or undermining the desired position?
  
• Are outdated descriptions still visible?
  
• Are competitors easier to understand?

This is often where the most useful insight sits.

3. Whether content is ready to be found, understood and cited

Search and content foundations still matter.

Many AI discovery tools use web search or hybrid retrieval, so clear, structured and current content remains important. That includes strong headings, useful question-led content, clear service descriptions, visible expertise, credible internal linking and pages that answer the questions stakeholders are actually asking.

This is not about producing more content for the sake of it. It is about making reliable information easier to find and interpret.

GEO measurement should connect to outcomes, not stop at outputs

The final reason the AMEC GEO Principles matter is that they connect GEO back to proper evaluation.

A mention in an AI answer is not, by itself, impact.

It may be an important signal. It may show that an organisation is visible for a relevant question. It may indicate that certain sources are shaping AI-generated responses. But it does not prove awareness, trust, consideration or commercial effect on its own.

GEO measurement needs to connect outputs to out-takes, outcomes and impact.

That means looking at what audiences may see and take away, but also how this relates to search behaviour, website engagement, enquiries, lead quality, stakeholder feedback, brand tracking, reputation measures and commercial indicators where relevant.

This is where GEO becomes more than a visibility dashboard.

It becomes a way to understand how organisations are being discovered, interpreted and considered in a changing information environment.

From AI visibility to AI-era reputation

The launch of the AMEC GEO Principles should help the industry avoid a narrow and tactical view of generative engine optimisation.

The opportunity is not simply to help brands appear more often in AI-generated answers.

The bigger opportunity is to help organisations understand and improve how they are represented across the information ecosystem that AI tools draw from.

That means GEO should be treated as part of reputation, research, content strategy and communications evaluation.

• It should ask whether an organisation is visible, but also whether it is understood.
  
• It should measure AI outputs, but also the evidence shaping those outputs.
  
• It should improve discoverability, but not at the expense of trust.
  
• It should support commercial goals, but with clear standards and ethical boundaries.

AI is already shaping how decisions begin. It is influencing what people see, which brands they explore, and which organisations are considered credible enough to investigate further.

That makes GEO important.

But the organisations that benefit most will not simply be those that chase AI visibility. They will be those that build a stronger, clearer and more trustworthy public information environment.

That is why the AMEC GEO Principles are useful. They help move the conversation from hype to evidence.

And that is where the real value of GEO begins.