In a nutshell: Traditional SEO is no longer enough. Today, your customers ask their questions directly to ChatGPT, Perplexity or Google AI Overview. To exist in these new spaces, you need to measure your AI visibility. This article explains how to audit your presence in LLMs, which metrics to track (Share of Model, sentiment) and which tools to use to drive your GEO (Generative Engine Optimization) strategy.
For years, the rule was simple: to be visible online, you had to rank on the first page of Google. But the nature of online search has changed. With the rapid rise of generative AI, users no longer click through lists of blue links. They have conversations with assistants that deliver direct, synthesised, ready-to-use answers.
The question is no longer just "Where do I rank on Google?" but rather "Does ChatGPT recommend my brand?" That is where AI visibility measurement comes in. If you are new to the topic, this guide will give you the tools to understand, analyse and audit your presence in LLM-generated responses (Large Language Models).
Why measuring your AI visibility has become essential?
You have probably heard that SEO is dead. It is not, but it is going through a deep transformation. Traditional search engines now integrate AI-generated summaries (such as Google AI Overview), pushing organic results further down the page. The result: click-through rates are collapsing on many queries.
At the same time, millions of users turn to tools like Perplexity or Claude to get product or service recommendations. These models are not designed to send traffic to your website; they are built to provide a complete answer within their own interface.
In this context, strong SEO no longer guarantees that an AI will cite you. You can rank first on Google and be completely invisible in ChatGPT. Measuring your AI visibility tells you whether your brand actually exists in these new audience touchpoints, and more importantly, how it is perceived there.
A quick glossary to understand AI visibility
Before going further, here are the key terms you need to know to navigate the world of AI search:
- LLM (Large Language Model): A language model trained on vast amounts of text, capable of understanding and generating natural language (e.g. GPT-4, Claude 3, Gemini).
- GEO (Generative Engine Optimization): The practice of optimising your presence and content to be cited and recommended by AI-powered search engines. It is the natural successor to SEO.
- Prompt: The query or question a user submits to an AI model.
- Citation vs. Mention: Being "cited" means the AI uses your content to build its answer (sometimes without naming you). Being "mentioned" means the AI explicitly names your brand as a relevant reference.
- AI Overview: The Google feature that displays an AI-generated summary at the very top of search results.
The 3 key metrics to analyse your AI ranking
AI visibility cannot be measured with the same metrics as traditional search. Forget simple positions 1, 2 or 3. Here is what actually matters:
1. Share of Model (LLM Share of Voice)
This is the AI equivalent of the classic share of voice. It measures how often the various language models mention your brand compared to your competitors across a set of strategic queries. It tells you whether you are perceived as the clear leader, a credible alternative, or simply absent from the conversation.
2. Sentiment analysis
Unlike Google, which simply ranks links, an AI "reads" the web to form an opinion. It is capable of judging your brand. If your online reputation is poor on review platforms, the AI may actively discourage users from choosing your services. Tracking the sentiment (positive, neutral, negative) associated with your mentions is therefore critical.
3. Source citations (links)
This is the new holy grail. Engines like Perplexity or Google AI Overview strive to back up their answers with clickable source links. Appearing in those sources generates traffic that is lower in volume than classic SEO, but far more qualified - because the user arrives on your site already convinced by the AI.
How to audit your LLM visibility step by step ?
Running an AI visibility audit requires a structured approach. Here are the steps to get started:
- Map search intent: Identify the questions your prospects are actually asking. Stop thinking in isolated keywords ("accounting software") and start thinking in conversational queries ("What is the best accounting software for a small business in the UK?").
- Define your test prompts: Build a list of queries covering different angles: informational (What is...), comparative (Brand A vs Brand B), and reputational (What are the reviews of...).
- Query the different models: Submit these prompts to ChatGPT, Gemini, Claude and Perplexity. Keep in mind that responses can vary from one session to the next.
- Analyse the results: Note whether your brand is mentioned, at what point in the response, with what sentiment, and whether your competitors are more visible than you. Also identify the third-party sources (media, comparison sites) the AI relies on to build its answer.
While this approach can be done manually for an initial test, it quickly becomes time-consuming. That is why dedicated tools have emerged to automate the process.
Which tools should you use to track your AI positions?
To run a serious GEO strategy, you need the right tooling. The market is structuring itself quickly around dedicated solutions.
| Tool type | Primary use case | Example solutions |
|---|---|---|
| GEO platforms | In-depth share of voice analysis, cited sources tracking and actionable recommendations. Ideal for running a real strategy. | BotRank.ai, Rankscale |
| Enriched SEO suites | For teams already using SEO tools who want to layer in early AI signals without changing their reporting workflow. | Semrush (AI Visibility), SearchAtlas |
| Simple monitoring tools | Basic brand mention tracking with alerts, without deep competitive analysis. | Otterly, PeecAI |
To get started, we recommend testing a native platform like BotRank, which lets you quickly visualise your share of voice against competitors in a clear, readable format, while identifying the influence sources you should target.
Frequently Asked Questions (FAQ)
Is traditional SEO dead now that AI has arrived?
No, SEO is not dead - it is evolving. Strong technical SEO and high domain authority remain essential prerequisites for being picked up by AI models. That said, SEO alone is no longer sufficient: you need to add a GEO (Generative Engine Optimization) layer to ensure you are properly understood and cited by language models.
Why does my site, which ranks first on Google, not appear in ChatGPT?
LLMs cross-reference multiple sources. If your site is technically hard to parse (too many ads, render-blocking JavaScript) or if your reputation is poor on third-party review platforms, the AI may simply choose to ignore you, even if Google ranks you well. AI models prioritise trustworthiness above all else.
How can I increase my visibility in AI-generated answers?
Three main levers: 1) Make your site technically sound (structured data, fast loading times). 2) Structure your content with clear summaries and FAQs to make it easy for AI to extract. 3) Build a presence in trusted third-party sources (comparison sites, specialist media, forums), since those are the sources AI consults when formulating its answers.
