GEO consulting: AI visibility that starts with measurement
I measure where your brand stands in ChatGPT, Gemini, Perplexity and Google AI Overviews answers using fixed prompt sets. Then I implement the content, schema and access work that closes the gaps.
In shortGEO consulting measures whether your brand is cited in AI answers and implements the work that improves it. It suits brands whose customers ask AI before they decide and who see competitors in those answers. The deliverables are target prompt sets, a GEO visibility matrix, a source analysis, and implemented content and schema improvements.
When do you need this service?
- Your customers ask AI when choosing a product or service
- Competitors are recommended for category questions and your brand is not mentioned
- AI gives incomplete or wrong information about your brand
- You want to track visibility with a repeatable measurement, not with screenshots
What does it cover?
Prompt sets
I group the questions your audience asks AI into four fixed sets: service intent, brand queries, local intent and comparisons. The concept itself is covered on the what is GEO page.
Source and competitor analysis
For every question I record which brands are recommended and which sites the answer relies on. That shows whether the gap is content, third-party sources or access.
Entity and schema
Your brand should be described the same way on your site, profiles and directories. I build Organization, Person, Service and FAQPage schemas around that definition and link your profiles with sameAs.
Quotable content blocks
I write short blocks that answer the target question in the first sentence. Each block has to make sense on its own, because AI takes the relevant passage, not the whole page.
llms.txt and AI bot access
I check whether robots.txt, firewall or CDN rules block crawlers such as GPTBot, OAI-SearchBot and PerplexityBot, and I prepare an llms.txt file. For a first check, use the GEO readiness check.
Re-measurement
I rerun the same prompt set at regular intervals and look at the trend across rounds, not at a single answer. Two anonymized cases are on the results page.
How does the work run?
- 01
Baseline measurement
I define the prompt sets with you and take the first measurement on ChatGPT, Gemini, Perplexity and Google AI Overviews. On timing, see when to start GEO.
- 02
Source analysis
I list the competitors and sites that get cited and sort the gaps into content, entity, third-party sources and technical access.
- 03
Content and schema implementation
On priority pages I implement answer blocks, schema, llms.txt and bot access rules, or hand your developer the code.
- 04
Re-measurement and report
I measure again with the same set, report what changed in the matrix and set the priorities for the next round.
What do I deliver?
- Target prompt sets in four groups
- A GEO visibility matrix: yes, partial, rival or none for each question group and AI system
- A breakdown of cited sources and recommended competitors
- Schema code, an llms.txt file and robots.txt recommendations
- Answer block briefs and a change report after every round
How do I measure success?
- How the yes, partial, rival and none cells in the GEO matrix change between rounds
- The number of questions your brand is cited for, and which of your pages get cited
- Referral sessions from AI systems in GA4
For two of my clients, ChatGPT recommends the brand directly in category queries. Details are in the case studies (Case 05 and 06).
One old rule blocks the GPTBot training crawler together with OpenAI's search bots. ChatGPT cannot read your pages, so it cannot cite them. Whether to allow GPTBot is a separate decision.
BeforeUser-agent: GPTBot
User-agent: OAI-SearchBot
User-agent: ChatGPT-User
Disallow: /User-agent: OAI-SearchBot
User-agent: ChatGPT-User
Allow: /Frequently asked questions.
How long does GEO consulting take?
The first cycle fits into my 90-day process: baseline measurement in weeks 1-2, plan in weeks 3-4, implementation in weeks 5-8 and re-measurement in weeks 9-12. Answers change slowly, so I recommend running GEO as ongoing work.
Does SEO have to be finished before starting GEO?
No, but the technical foundation has to be sound. AI systems draw on content that can be crawled and indexed, so I fix serious indexing problems first.
Do I need an llms.txt file?
It is not required. llms.txt is a proposal, not a standard, and it is unclear how each AI system uses it. Content, entity consistency and bot access matter more.
Do you guarantee that my brand will appear in AI answers?
No, I don't guarantee that. Answers vary by user, time and model. What I commit to is a repeatable measurement with fixed prompt sets and implementation based on it.
Which AI systems do you measure?
I measure ChatGPT, Gemini, Perplexity and Google AI Overviews. Which one gets more weight depends on the habits of your audience.
You may also find these useful
- All servicesAll seven modules and how they combine.
- Content strategyI use data to decide which topics you should cover, with which intent and in which order.
- The 90-day working processThe order and outputs of each service.
Let's measure where you stand in AI answers.
Tell me your industry and competitors. In the first call we decide which questions you need to be visible for.
What happens in the first call?
- I listen to your goalIn a short call we talk through your goals, your site's history and what has been done so far.
- We look at your site togetherI share what stands out in your technical foundation, content and AI visibility during the call.
- I send a tailored scopeAfter the call I send the scope of work and the quote in writing. No fixed packages, the decision is yours.
WorkExe Agency