Does ChatGPT name your company?
More buyers now ask an AI assistant instead of scrolling search results. We measure whether your company appears in those answers, identify what is blocking it, and do the work to change it.
Visibility audit · repeatable measurement · technical and content implementation
Don't take our word for it
Pick an industry, copy the question and paste it into ChatGPT or Perplexity. If your company is missing while competitors are named, that is exactly the problem we work on.
Model answers are non-deterministic — the next attempt may differ. That is why an audit is never a single test, but repeated measurement across dozens of questions.
Which {category} tool would you recommend for a 50-person company?
What are the best alternatives to {competitor} in Europe?
Compare {category} platforms — pricing and key features.
Replace the parts in braces with your own — product, city, specialism.
Three things we do
Each can be bought separately. Most engagements start with the audit, because without measurement there is nothing to discuss.
AI visibility audit
A one-off, self-contained report: where your company is cited, who is cited instead, and what is standing in the way.
- 30–50 real buyer questions across ChatGPT, Perplexity, Gemini and Google AI Overviews
- Technical audit: crawler access, rendering, structured data, indexation
- Review of the third-party sources models actually cite
- Prioritised action plan
Visibility monitoring
The same question set measured on a cycle. Not a snapshot, but change over time reported together with the spread between runs.
- Monthly measurement using one fixed method
- Delta and variance reporting, not a single flattering number
- Competitor comparison
- Alert when visibility drops
GEO, AEO and SEO implementation
The work that actually moves the measurement: technical fixes, content structure, and presence in the sources models draw on.
- Structured data (schema.org) and technical remediation
- Content shaped so a model can quote it cleanly
- Presence in industry listings, directories and comparison pages
- Classical SEO as the foundation
Four steps
Define the questions
We don't guess. Questions come from your sales conversations, Search Console data and competitor analysis — phrased the way buyers actually ask.
Measure the baseline
We run the question set through the assistants repeatedly and record who is cited and which sources the model used. That becomes your reference point.
Remove the obstacles
Technical blockers first, because they invalidate everything downstream. Then content structure. Then presence in third-party sources.
Measure again
Same questions, same method. The report shows the difference and the spread, so effect can be told apart from noise.
What we don't promise
GEO is a young field and a lot of it is oversold. We would rather state the limits up front — it makes the engagement easier to hold us to.
No guaranteed placement in AI answers
There is no ranking to buy and no stable algorithm to optimise against. Model answers are non-deterministic. Anyone guaranteeing a spot in ChatGPT is promising something they do not control.
A single measurement proves nothing
The same question asked twice returns different answers. We measure repeatedly and report the spread, rather than one number that looks good on a slide.
llms.txt is not a lever
There is no public confirmation that major model providers read it. We implement it because it costs nothing — not because it works. We don't bill for it separately.
Much of the work sits outside your website
Models cite Wikipedia, directories, comparison pages and forums. Optimising your own site is typically half the job, sometimes less.
Three ways to work with us
Pricing is quoted individually — it depends on the size of the site and the number of questions under measurement. Below is what each option covers.
Audit
one-offFull report and action plan. No forward commitment.
- 30–50 questions across 4 engines
- Technical and content audit
- Third-party source analysis
- Prioritised plan
- Walkthrough call
Ongoing
monthlyMeasurement, implementation and content. Scope matched to size.
- Monthly measurement and delta report
- Technical implementation
- 2–4 content pieces per month
- Third-party source work
- Direct contact, no account-manager layer
White-label
for agenciesFor SEO agencies and software houses that need GEO in their offering.
- Audits and reports under your brand
- Per-client billing
- Technical support for your team
- Sales material
We quote in EUR, net of VAT, usually within one working day of your enquiry.
Frequently asked questions
How is GEO different from SEO?
SEO aims at a click on a link in the results page. GEO aims at your company being named inside the answer the model generates — often with no click at all. The techniques overlap, because models largely draw on the same pages search engines rank well, but the goal and the measurement differ.
Are GEO and AEO the same thing?
In practice yes. GEO (Generative Engine Optimization) is the more common term in the US and UK; AEO (Answer Engine Optimization) is used more in continental Europe. Both mean optimising for AI-generated answers.
How do you measure visibility in ChatGPT?
We agree a set of 30–50 questions your buyers actually ask, then run them repeatedly through ChatGPT, Perplexity, Gemini and Google AI Overviews. We record whether and how your company appears, who appears instead, and which sources the model used. Because answers are unstable, results are always reported with the spread between runs.
Can you guarantee ChatGPT will recommend us?
No. No mechanism exists that would guarantee it, and answers shift between sessions and across model updates. We can remove specific obstacles, improve the odds and measure the change honestly. If a provider offers a guarantee, ask how they intend to settle it.
How long does it take?
Technical fixes ship in weeks. Change in model answers depends on when the model or its search layer revisits the sources — a realistic horizon is several months. Work on third-party listings usually shows up faster than work on your own content.
Do we have to drop SEO?
The opposite. Models largely rely on the pages search engines rank well, so SEO is the foundation of GEO rather than its competitor. We treat both as one body of work.
Can you show results from other clients?
We are early in this service line and do not yet have closed case studies we could honestly point to. Instead we show the method and our own measurement — including the visibility of our own site, which we build with the same method. Client results will be published with numbers and methodology once they exist.
Which languages and markets do you cover?
Measurement can be run in any language, because it consists of questions posed to the assistants. Our own delivery — calls, reports, correspondence — is handled in English and Polish. We are based in Poznań, Poland and work with clients across the EU.
Send us a domain and we'll show you where it stands
Every quote includes a preliminary measurement: a handful of questions from your industry run through the assistants, so you can see the starting point before committing to anything.
We reply in English. Our team works in English and Polish — if you prefer another language, we will still correspond in English rather than promise something we cannot deliver well.