GEO Audit in Burgos — Discover how (or whether) AI sees your business

An SEO audit analyses technical and content factors on your website: speed, indexation, structure, backlinks and metadata. A GEO audit, on the other hand, analyses how AIs see you from outside your website: what ChatGPT, Gemini and Perplexity say about you, what sources they use to build that image and what gap exists between how you describe yourself and how the models describe you. They are complementary analyses: the SEO audit diagnoses your visibility in Google; the GEO audit diagnoses your visibility in the new generation of search engines.

LLM visibility is the measure to which your company appears as a reference when someone asks an AI about your type of business or sector. For a company in Burgos, it means that when someone asks "what marketing agency do you recommend in Burgos?" or "what is the best restaurant for corporate events in Burgos?", your name appears in the response. It is not a numerical ranking: it is a qualitative presence measured with the AI Share of Voice — the percentage of relevant prompts in which you appear.

The main GEO metric is the AI Share of Voice: of all the relevant prompts we run in a period, in what percentage does your company appear? This metric is complemented by: the sentiment of mentions (does it speak well, badly or neutrally?), information accuracy (is the data correct?), frequency of citation as an authoritative source, and comparison with competitors. With these metrics we build the AI visibility map of your company.

2025 data shows that between 30% and 40% of searches in age groups under 35 are conducted through conversational AI instead of traditional search engines. In informational, comparison or recommendation queries, AI already surpasses Google for many user profiles. For sectors where the purchase decision involves prior research — such as professional services, renovations, health or technology — a significant portion of potential clients goes through AI before making contact.

This is the most common scenario we find in GEO audits: the audited company does not appear in relevant prompts, but its direct competitors do. This happens because competitors have more presence in the sources the models use: more mentions in media, more reviews on authoritative platforms, more quality content on their websites. The result is that every time a potential client in Burgos asks an AI about your type of service, they receive your competitor's name, not yours.

AI models do not have their own opinion about your company: they synthesise information they have processed from multiple sources. If those sources are scarce, the AI says it does not have enough information. If they are contradictory, it gives imprecise responses. If they are negative, the image it builds is negative. Controlling the sources that feed the AI's perception of your company is the foundation of GEO work, and the audit is the first step in understanding what those sources say.

We launch between 20 and 40 representative prompts in each AI model, designed to simulate the real questions a potential client in your sector in Burgos would ask. We document the complete responses: whether your company appears, how it describes you, what attributes it highlights and which it omits, whether the information is correct and up to date, and whether it explicitly recommends your services.

AI models with real-time search, like Perplexity, explicitly cite sources in their responses. For models that do not directly cite sources, we analyse your company's mention profile on the internet: sector directories, media outlets, review platforms, social networks and your own website. The sources that appear in the first search results for your name are, with high probability, the ones feeding the AI models' perception of your company.

Once the current state is documented, we compare the image AIs have of your company with the image you want to project: main services, differentiators, geographic area and credentials. The differences are visibility gaps. Some are presence gaps (the AI simply does not talk about you), others are content gaps (it talks about you but with incomplete information) and others are positioning gaps (it talks about you but not as a sector reference).

Schema Markup in JSON-LD helps AI models understand your website in a structured way. In the audit we verify what Schema you have implemented, whether it is correctly configured and whether it includes all relevant data about your business: name, type of business, services, service area, opening hours and credentials. Schema errors or their absence are common and have a direct impact on how AIs process and interpret information from your website.

A common technical error is inadvertently blocking the bots of the leading AIs in the robots.txt file. We verify that GPTBot (ChatGPT), Google-Extended (Gemini), PerplexityBot and other relevant bots have correct access to your website. We also check whether the llms.txt protocol is implemented and correctly configured to guide models towards the most relevant content. These blocks are silent: they give no visible error but make your website invisible to certain AI models.

We design between 20 and 40 specific prompts for your sector, services and geographic area. We launch them in ChatGPT (GPT-4o), Google Gemini and Perplexity AI and document the complete responses with screenshots. This process gives us the initial visibility map: which prompts you appear in, which you do not, and what each model says when it does mention you. The number of prompts where you appear out of the total is the starting point for AI Share of Voice.

We identify the sources that AI models with real-time search cite when talking about your sector, and we analyse your company's mention profile in the highest-authority sources on the internet. This analysis reveals which are the information channels for your company in LLMs and which are the gaps: authoritative sources in your sector where your company should be present but is not.

With data from the previous two phases, we create the visibility gap map: differences between your company's current image in AIs and the desired image. We classify gaps by type (presence, content, positioning) and by urgency. We also identify specific content opportunities: questions in your sector that AIs answer without citing your company and that you could answer better than current sources.

The GEO audit report includes: the complete visibility map across the three audited LLMs, screenshots of the most relevant prompts, source and mention analysis, the prioritised list of gaps, and the GEO action plan with recommended actions ordered by estimated impact and implementation effort. The report is an actionable working document, not a theoretical analysis.

The visibility map shows, for each of the prompts launched, whether your company appears in the response and how. It includes the global Share of Voice (percentage of prompts with mention) broken down by model and by prompt type. This map is the snapshot of your current AI visibility and the reference point against which to measure the progress of GEO actions.

The source analysis identifies which platforms, media and directories are being used by AI models to build the image of your company and your sector. This list is strategic: present sources are the ones to maintain and optimise; relevant sources in your sector where your company is not present are the channels to work on to increase AI visibility.

The GEO action plan prioritises actions according to their balance between expected impact and implementation effort. Those with the greatest impact and least effort go first: technical corrections to robots.txt and Schema markup, profile updates on authoritative platforms and specific Q&A content creation. Higher-impact, higher-effort actions are planned for the medium term.

AI Share of Voice is the main GEO tracking metric. It is calculated as the percentage of relevant prompts (selected during the audit) in which your company appears in the audited models' responses. We recommend measuring this metric quarterly to evaluate the progress of GEO actions. The initial audit establishes the starting benchmark; follow-up audits measure evolution and allow strategy adjustment.