Help us prepare for our appointment
The role of an SEO specialist has changed significantly over the last two years. Simply optimising relevant keywords and anchor text is no longer enough: the article, or part of the content, must now be selected by a chatbot (ChatGPT, Claude, Gemini, Perplexity, etc.) and included directly in its responses. Ranking on the first page of Google is no longer the goal to aim for in 2026!
This is the crux of SEO for AI, also known as GEO (Generative Engine Optimisation): simply aiming for the top spot in search results is no longer enough; you need to become the source that the model chooses to cite and include in its generated response.
This development is fundamentally transforming the profession and requires us to test, adapt, challenge our own assumptions, and remain open to the changes brought about by conversational AI. Seeking support from an agency specialising in AI SEO is becoming essential to maintain – or even improve – your online ranking, whether on Google or in the responses provided by generative AI. We’ll explain which criteria to take into account and which professionals to turn to so you can jump on the bandwagon!
Key takeaways:
In summary, SEO is evolving towards GEO: to remain visible, it is essential to adapt your content and measurement strategy to meet the requirements of generative search engines and strengthen your overall credibility.
Behind the acronym GEO lies a fairly simple idea: to present your content in the form of a buffet from which artificial intelligence systems can draw. A traditional search engine simply indexes pages; a generative model, on the other hand, goes further: it extracts fragments, cross-references them with other sources, and turns them into a unique response that it presents as its own.
🧑🍳To continue the metaphor: you are no longer the restaurant that the customer chooses and where they sit down to eat. You are the supplier who stocks the kitchen of a top restaurant (let’s call it the TrattorIA). The end customer will never see your name on the bill, but if your products are good, well labelled and easy to incorporate into a recipe, the chef will choose you over a less reliable competitor.
LLM stands for Large Language Model. It is the engine that powers ChatGPT, Claude, Gemini and Perplexity. In practical terms, it is a system trained on vast amounts of text, capable of generating a response by predicting, word by word, the most likely and coherent continuation based on what it is asked.
Two things to bear in mind to understand why this is a game-changer in terms of visibility:
🔍 An LLM doesn’t ‘Google’ information. It either draws on its training data or retrieves fresh data in real time using RAG (Retrieval-Augmented Generation) – a bit like ‘checking my notes before answering’.
✍️ Forget the list of links: it gives you a brand-new answer, put into its own words. Your content can feed into its answer… without anyone ever knowing that you’re the source 🫠 (unless it deigns to credit you).
Traditional SEO is based on a well-known strategy: well-placed keywords, a solid link profile, and hierarchical headings (H1, H2, H3…) that help Google understand the page’s structure. SEO for AI takes a completely different approach.
On Google, you get a click. On ChatGPT, at best you get a mention in the middle of a sentence that the user will never trace back to its source. The challenge has therefore changed: your page must not only exist, but also be clear and reliable enough for a model to decide to rely on it rather than on a competitor.
Every AI has its own secret formula for retrieving information. Some rely on a near real-time index, whilst others use more static databases, each with their own specific trust criteria. Content tailored to appear in Google’s AI Overviews therefore offers no guarantee whatsoever on Perplexity, which sources its answers from elsewhere and in a different way. It is precisely this kind of nuance, search engine by search engine, that Ad’s up details in its analysis of the launch of AI Overviews and AI Mode in France.
Traditional SEO isn’t dead – far from it – it’s simply changing the way it’s done. The problem is that an agency well-versed in title tags and internal linking doesn’t necessarily have the instincts to navigate this new terrain, where the rules haven’t even been fully established yet.
| Traditional SEO | SEO for AI / GEO | |
| Objective | Rank in positions 1–3 on Google | To be mentioned in the generated response |
| Unit of measurement | Ranking, organic traffic | Citation rate, share of voice in AI responses |
| Content format | Long-form articles, keywords | Standalone sections, structured data, Q&A |
| Sign of trust | Backlinks, domain authority | E-E-A-T, multi-source consistency, schema.org |
| Rate of change | Algorithm updated a few times a year | Models updated continuously |
| Monitoring tools | Search Console, SEMrush | AI citation tracking tools, test prompts |
An agency specialising in LLM optimisation knows how to balance both types of SEO within the same content, because by 2026, neither has completely replaced the other. This is precisely the approach championed by Ad’s up through its GEO expertise: not to pit SEO against GEO, but to make them work together within a single visibility strategy.
There’s no need for a three-hour audit. These seven points are enough to distinguish an agency that really knows its stuff from one that’s just jumping on the bandwagon.
1. She talks about technical matters without using unnecessary jargon 🥸RAG
, embeddings, crawling by AI agents – sure, it’s all well and good for a forum of SEO experts, but these concepts need to be applied with a bit of common sense.
2. It measures mentions, not just rankings
🥇If reporting still revolves solely around Google rankings, the agency hasn’t really pivoted towards GEO. The metric that matters is your actual presence in AI responses to your strategic search queries.
3. She knows how to structure content so that it can be reproduced as it is
👯Schema.org, well-structured FAQs, self-contained sections, no waffle – she structures her article like someone with Virgo rising, the masters of organisation and prioritisation.
4. It shows results 📸Case studies
, screenshots of quotes obtained from ChatGPT or Perplexity, figures. We want proof!
5. She takes E-E-A-T seriously 🫶
I’ll have a double E, I’ll have an A, I’ll have a T 🤸 for Expertise, Experience, Authority and Trustworthiness. Generative models largely follow the approach established by Google. Content written by an identifiable expert remains a step ahead of mass-generated text.
6. She follows the news closely enough to anticipate 👀
Developments are happening almost every month. Check whether the agency publishes content on these topics itself. Ad’s up regularly analyses these shifts, as in ‘GEO Strategy: Rethinking Your Media Mix in the Age of AI Overviews’, which quantifies the actual impact of AI Overviews on organic clicks. The agency’s SEO experts’ LinkedIn accounts are also good indicators ;)
7. She explains her method
🤌You need to be able to track, over time, what has been done and why. Vague reporting makes it impossible to assess effectiveness.
💡 Advice from our experts:
A good onboarding meeting should enable you to confirm the following points:
At Ad’s up Consulting, the aim is not to sell GEO as a marketing gimmick, but to integrate it into a comprehensive visibility strategy, alongside SEO, SEA and social media advertising. The agency has, in fact, developed a dedicated GEO service, designed to answer a very practical question that more and more advertisers are asking themselves: ‘Does my business exist, in the eyes of an AI?’
This approach is based on almost daily monitoring of the sector — it can be seen, for example, in the GEO ‘Best Agency 2026’ comparison: a comparison of the 10 agencies specialising in AI visibility, which asks precisely the right questions to help you choose a reliable partner in this still-emerging field.
And because generative search won’t replace traditional search overnight, Ad’s up continues to support its clients with traditional SEO and SEA in parallel, ensuring a visibility strategy that doesn’t rely entirely on a single channel.
📣 Case study: increasing the share of traffic from LLMs tenfold
Business Challenge
A brand in the natural cosmetics sector wanted to boost its visibility on generative AI engines. The aim: to become a benchmark in its market, to be better understood and more frequently recommended in the responses generated by LLMs.
Ad’s up Consulting’s approach
Our business impact
GEO encompasses the techniques that enable content or a brand to be understood, cited and recommended by generative AIs such as ChatGPT, Perplexity, Gemini or Google’s AI Overviews.
Yes, absolutely. Generative engines still rely heavily on web content indexed in the traditional way. SEO and GEO complement each other; they are not mutually exclusive.
The budget varies depending on the volume of content to be produced or optimised, the size of the website and the level of monitoring required. A reputable agency should be able to offer you an initial audit to establish a realistic budget.
The simplest way is to manually test several prompts related to your business and your competitors. Specialised AI citation tracking tools also exist to automate this monitoring on a larger scale.
Some can, provided they have developed genuine technical expertise in the field. The risk is ending up with a GEO service that has simply been tacked on to an existing portfolio, without a dedicated methodology of its own.
