01 · Category map
How AI engines define your category
Where the engines draw the boundaries. Almost always different from how your team draws them.
A measurement practice, not a dashboard
GEO Assessment is the European measurement practice for how ChatGPT, Claude, Gemini, and Perplexity represent your brand. Our proprietary consulting layer turns the findings into action you can publish on Monday.
Three patterns from our assessments
of B2B categories we’ve measured, the brand AI names first wasn’t even in the buyer’s stated consideration set.
From 500+ assessments · 2026
the citation weight for B2C brands on Reddit and category-specific trade publications versus general press. Where you invest PR matters.
Source ecology analysis · 2026
of brands describing themselves as ‘modern’ I engines describe as ‘legacy’ or ‘established’ instead.
Attribute mapping audit · 2026
Anatomy of an assessment
Every assessment produces the same six outputs. They’re designed to be read in a meeting, not configured in a dashboard. Each one answers a question a different part of your team is already asking.
01 · Category map
Where the engines draw the boundaries. Almost always different from how your team draws them.
02 · Consideration set
For real buying-intent queries: the actual surfaced competitors, in order, with the proximity measured.
03 · Attribute map
Unsupervised cluster of the adjectives engines reach for. Frequently very different from your own positioning.
04 · Sentiment with context
Not aggregate polarity. The specific queries that flip a brand from ‘trusted’ to ‘expensive’ — and why.
05 · Source ecology
Ranked map of which trade publications, communities, and analysts feed the citation graph for your category.
06 · Perception delta
Side-by-side comparison of what you claim and what’s been consolidated. The work programme writes itself.
How an engagement actually runs
Week one
90 minute workshop. Your target audience, your desired perception, your search console, and your real customer language. We seed the assessment from what your buyers actually say.
Weeks two and three
We create assessment questions built on data-based conversation seeds. The assessment runs at multi-shot sample sizes across every engine that matters. The data lands in a shared workspace.
Week four onward
Reports that sit in dashboards don't change anything. We bring the patterns from dozens of prior assessments, translate findings into actions ranked by leverage, and work alongside your team continuously while you ship them.
The practice behind it
Pedigree
Built by Cocomore, the Frankfurt-based consultancy with twenty-five years of brand and content engagements across European enterprise.
Methodology
Multi-shot sampling, contextual sentiment, source-graph mapping, perception-versus-positioning analysis. Documented, not proprietary. Defensible to a research director.
Operating model
A structured operating model behind every engagement — a dedicated team per account, a defined monthly cadence, documented updates. Accountability is built into the process, not improvised.
Infrastructure
EU-resident data processing. Documented sub-processor list. GDPR-clean for procurement on the first review.
Recent work · published
We mapped three shortlists across the same set of buying-intent queries. The single restaurant the engines and Tripadvisor agreed on sits at position #8 on Tripadvisor’s list. Each LLM had its own favourites, drawn from a different set of sources you’d have to actively work to be inside.
Read the field map →Two ways to start
A concierge-assisted four-week pilot with our analyst team, or set up your first assessment yourself in the app. We’ll meet you wherever you are.