Let AI cite
you as the source.
Your customer now types the question into a chat rather than a search box, and gets the answer in one go. If your brand is not in that answer, you lose them without a click ever happening. GEO is what changes that.
Visibility AuditHow Visibility Works
Being inside the answer.
When someone types a question about your sector into a chat, the model builds an answer from sources it trusts and lists them underneath. Our goal is to be on that list.
Here is the striking part: the large majority of pages cited in AI summaries do not sit near the top of the classic results for the same query. A good ranking is no guarantee you will appear in the answer - it is a separate discipline.
Scope of Work
What do we do?
Visibility Audit
We put the real questions from your sector to the platforms and measure how many answers your brand currently appears in, and who else does. That is the starting point.
Entity Schema
With Organization, Article, Author, FAQ and HowTo schemas we mark up who your brand is, who wrote what, and what you know - in a language machines understand.
Authority Content
Deep, long-form content the model can summarise and quote. Not generic filler; content written by a named expert that actually makes a claim.
Original Data
Your own research, your own figures. This is the content type most open to being quoted - because it exists nowhere else.
Crawler Access Management
robots.txt and llms.txt configuration for GPTBot, ClaudeBot, PerplexityBot and Google-Extended. Which content is open and which is closed stays under your control.
Third-Party Mentions
Your brand mentioned consistently across trade publications, forums and professional platforms. Models look not at one source but at agreement between sources.
Topic Cluster Architecture
Connected topic clusters rather than scattered pages. Depth in one area sends a stronger signal than breadth on a single page.
Technical Foundation
Clean semantic HTML, fast Core Web Vitals, correct canonicals and redirect hygiene. A page that cannot be read cannot be quoted.
Citation Tracking & Reporting
The defined question set is asked regularly; how many answers you appear in, where you are losing ground and traffic arriving from AI platforms are reported monthly.
Entity Network
The model looks at consistency,
not at a single source.
When an AI model "knows" your brand, what has really happened is that it saw the same information across independent sources. When your name, what you do and where you are is written the same way everywhere, trust builds; when it conflicts, the model skips you.
Every one of these points has to tell the model the same story. If one place carries a different brand description, a different service list or an old address, the model usually uses none of them - because it cannot decide which is right.
The Technical Side
What do we tell the crawlers?
We define explicitly what AI crawlers may read on your site. The content you want cited is opened up, while price lists and customer-only areas stay closed.
- Per-crawler access rules in robots.txt
- Declaring the content to be surfaced through llms.txt
- Organization, Article and Author markup in JSON-LD
- Semantic HTML - heading hierarchy and meaningful elements
- Core Web Vitals and mobile performance optimisation
- Canonicals, redirects and duplicate content clean-up
# ai crawlers User-agent: GPTBot Allow: /services/ Disallow: /trade-pricing/ User-agent: PerplexityBot Allow: / User-agent: Google-Extended Allow: /
# Your Brand > A clear one-sentence definition. ## Services - [Service name](/service): what it covers. ## Research - [Sector report](/report): original data.
Process
How do we proceed?
Measurement
We build your question set and record your current visibility on the platforms, along with your competitors.
Foundation
Schema markup, crawler access and gaps in the technical foundation are closed.
Production
Authority content and original data are produced; third-party mentions are built.
Tracking
The same question set is asked regularly and the citation rate is reported monthly.
FAQ
Questions we get asked
What is the difference between GEO and SEO?
SEO aims for a ranking on the Google results page. GEO aims to be cited as a source inside the answer the AI produces. The distinction matters: most pages cited in AI answers are not even near the top of the classic results for the same query. Authority and entity clarity decide it, not ranking.
Which platforms do you work on?
Chat-based models such as ChatGPT, Google Gemini, Perplexity and Claude, along with answer surfaces like Google AI Overviews and AI Mode. The technical side is common to all of them; the content and citation side differs by platform behaviour.
If I already do SEO, do I need GEO as well?
They do not replace one another. A solid technical SEO foundation is a precondition for GEO, but it is not enough on its own. When a user types the question into a chat rather than a search box, the answer is produced without a click ever happening. If your name is not in that answer, you lose that visitor entirely.
Is it risky to open my site to AI crawlers?
Not when it is done in a controlled way. We decide separately, through robots.txt and llms.txt, which sections are open to crawlers such as GPTBot, ClaudeBot, PerplexityBot and Google-Extended. Price lists and customer-only areas stay closed, while the content you want cited is made reachable.
How do you measure the result?
We put a defined set of questions to the platforms on a regular basis and record how often your brand is cited in the answers. Alongside that we report referral traffic from AI platforms and the change in branded search volume.
How long before we see results?
Technical changes and schema markup take effect within a few weeks. A lasting rise in how often you are cited usually takes 3-6 months, because the models' content refresh cycles and the build-up of third-party brand mentions both take time.
Related Services
Stronger together
Does AI know who you are?
Let us ask the questions from your sector and see together where your brand appears.
Request a Visibility Audit