A GEO audit can quickly generate a great deal of information. But this assessment is only truly worthwhile if it answers one simple question: what needs to change?

We test queries, observe a brand’s visibility in ChatGPT, Gemini or other generative engines, compare competitors and measure gaps. This is where many GEO initiatives become difficult to act on. Knowing that a hotel rarely appears, that a tour operator is seldom recommended or that a competitor appears more often is useful. But it does not yet tell us whether the problem lies in positioning, content, evidence, external sources, reviews, technical factors or simply poor alignment between several of these elements.

A useful GEO strategy must therefore connect three things: what AI systems observe, why they observe it and what the business can realistically do to improve the situation. This is the logic around which we have structured our method: See, Align, Build, Amplify.

See: measure what AI systems really understand

The first step in a GEO audit is not to produce content. It is to understand how the business appears in generated answers. Is it present for the queries that really matter to its business? Is it mentioned without being selected? For which types of customer or use case does it disappear? Which competitors are preferred, and in what contexts?

The important thing is not to look solely at average visibility. A hotel may feature prominently in general queries yet be almost absent as soon as the request becomes more specific: couples, families, budget, location, a particular occasion or type of experience. A tour operator may be well recognised for a destination but rarely suggested when a traveller adds criteria relating to price, specialisation or travel style. That is why an AI visibility audit must examine the situations in which a brand is present, relevant or preferred. Otherwise, what we measure is mostly noise.

The expected outcome of this first step is therefore not a score. It is a map of the gaps: where is the business understood as it wishes to be, and where is it not? We explore this distinction between presence and recommendation in being visible in AI does not mean being recommended.

Align: understand why signals diverge

Once the gaps are visible, we must avoid the most common reflex: immediately producing more content. The problem may lie elsewhere. A business can have a very clear website while sending contradictory signals across its ecosystem. The website may highlight one experience while customer reviews mostly discuss something else. A platform may present a property with positioning that differs from the brand’s. Partners may use outdated descriptions. Some evidence may exist but be difficult to find or rarely picked up elsewhere.

In this case, the issue is not the volume of information. It is consistency. Before adding new signals, we need to check that those already in place tell a compatible story. This alignment step matters because AI does not encounter only the business’s official messaging: it may encounter the website, reviews, platforms, media, partners, directories, local content or other external sources. A serious GEO strategy must therefore examine this entire environment.

“What genuinely supports the business’s positioning, and what undermines it?”

Build: address the gaps that really matter

Only once we understand these gaps can we build. And here again, building does not simply mean creating “pages for ChatGPT”. The work can vary considerably from one situation to another. Essential information may be missing from the website. An experience may be poorly described. Expertise may be claimed but barely demonstrated. Some evidence may exist without being accessible. The technical structure may make some content difficult to understand. The scope of the positioning may be too broad for the business to be clearly associated with specific situations.

Building must therefore be guided by the assessment. This is what distinguishes a GEO strategy from a generic list of best practices. The aim is not to “optimise for AI” in the abstract: it is to strengthen the concrete reasons why a generative engine should consider the business a good answer. This may involve content, data, evidence, more specific pages, external sources or technical elements. But every action must address an identified problem.

Amplify: strengthen the right signals, rather than simply producing more

Once the useful elements have been clarified or built, they still need sufficient visibility across the information environment. This is where amplification comes in. It may involve customer reviews, partners, media, press relations, selected editorial content, distribution or other relevant sources. But this step must come last.

Amplifying poorly defined positioning, an experience with little supporting evidence or a promise contradicted by customers does not solve the problem. It may even make it more visible. Amplification only has value when what is being amplified is clear, credible and consistent. This is also why GEO cannot be treated solely as a search optimisation issue: it touches on positioning, the experience offered, the quality of the evidence and sometimes the offering itself.

What should a GEO audit achieve?

A good GEO audit should not simply produce a ranking, a score or a visibility table. It should help an executive make decisions. By the end of the work, it should be possible to answer a few questions clearly: why are we recommended here but not elsewhere? Which competitors are better understood than we are, and why? Which signals strengthen our position? Which are weak, absent or contradictory? Which actions are likely to have the greatest impact? And how can we subsequently check whether the situation is changing?

This is when a GEO audit becomes a GEO action plan. The distinction matters: an assessment describes the situation; a strategy enables decisions. For the fundamentals, we refer readers to our reference article: what is GEO for tourism businesses?

How do you recognise a sound GEO method?

For an executive, there is ultimately a fairly simple test. A serious GEO approach must be able to do three things. First, it must measure more than mere presence in AI systems. Next, it must explain the gaps observed, even when several possible causes coexist. Finally, it must lead to prioritised actions and indicators that allow progress to be tracked.

If it stops at “you appear in 42% of answers”, it remains descriptive. If it helps you understand why, decide where to act and subsequently check whether recommendations change, it becomes genuinely useful.

See, Align, Build, Amplify

Our GEO method therefore rests on four simple questions. See: how do AI systems really perceive the business? Align: do the different available signals tell a consistent story? Build: what information, evidence or experiences need better documentation? Amplify: which signals then deserve to be strengthened through the right sources?

This sequence matters because producing more content is very easy. It is much harder to know which content, evidence or action can genuinely change how a business is understood and recommended. Yet that is precisely where the value of a true GEO strategy lies.