01
Discoverability
Does the machine know you exist? We measure whether you appear, how often, and across how many prompts and models.
Reality Check
Find out whether AI knows, understands and recommends your brand.
We compare your brand with nine others using 10 buying questions and 3 models. We preserve the 30 answers and code them into 300 comparable records, then turn them into a ranking, evidence and concrete priorities.
€299 · one-time payment · delivered in 10 business days
Diagnostic preview
The path to being chosen
01
AppearDoes your brand enter the answer?
02
Be understoodDoes it explain what you do and for whom?
03
Have proofCan it find reasons to support the claim?
04
Build confidenceDo the sources and models agree?
05
Be chosenDoes it recommend you over the alternatives?
The same test across the category
10brands
10buying situations
3models
300comparable records
The problem
AI already talks about your brand. The question is whether it gets it right.
At TOTEM, we reviewed three months of search data. The two queries that showed the website most often were not about the agency, but about another company with a similar name.
4,920
appearances for searches about another company
16%
of all visibility that quarter
0
real visits generated
When we reviewed the case, we found something very specific: the website identified the company poorly. Basic details were missing —the full name, location, founders and references that distinguished it from similarly named businesses. Google therefore connected it with searches that had nothing to do with what the company offered. The Reality Check looks for this kind of mismatch: not only whether you appear, but whether systems understand who you are and compare you with the right companies.
Read the full caseAI Brand Preference
From existing to being chosen, in five layers.
The AI Brand Preference Index is the methodology behind the Reality Check. It goes beyond counting mentions to reveal where the path from being known to making the shortlist breaks down.
02
Understanding
Does it understand what you offer and for whom? We review categories, audiences, use cases and associated attributes.
03
Evidence
Can it find independent proof? We identify owned sources, media, reviews, communities, comparison sites and consultancies.
04
Trust
Are there observable trust signals? We analyse source quality and diversity, support for claims and agreement across models.
05
Preference
Do you make the shortlist? We measure position, recommendation and which brands take your place when you are absent.
Methodology
The same test for every brand.
We choose ten real buying situations and ask them in the same way across ChatGPT, Gemini and Perplexity. That lets us see what changes by brand, question and model.
What we compare
Your brand against its category
Your brand and nine relevant alternatives. You can name up to three priority competitors; the rest are agreed before the work begins.
10 brands
Your brandBCDEFGHIJ
What we ask
Ten buying decisions
Questions a customer might ask while discovering, comparing or choosing, without suggesting which brands should appear.
DiscoverCompareTrustChoose+6
Where we test it
Three assistants, the same questions
We repeat the test in ChatGPT, Gemini and Perplexity and preserve the answers, sources and date.
The same test in
ChatGPT
Gemini
Perplexity
What you receive
A map of what helps or holds back your brand
We record who appears, in what position, with which arguments and supported by which sources, then turn the patterns into concrete priorities.
In every answer we observe
- Who appears
- Their position
- The arguments used
- The sources cited
30 answers · 300 comparable records
We compare answers; we do not pretend to read a model's mind
The report preserves the questions, answers, sources, models and date. Its conclusions come from that observable evidence.
The deliverable
The result, not the agent.
A clear snapshot of how AI answers about your category today, which brands win and which fail to appear. Every conclusion can be checked against the evidence.
01
Ranking of ten brands
Visibility, position, recommendation and performance across all five layers of the index.
02
Matrix of 300 comparable records
Appearance, position, attributes and reasons by brand, prompt and model, with the observed answers preserved.
03
Source Share
How evidence splits across brand websites, media, reviews, communities, comparison sites and consultancies —and which domains dominate the category.
04
Consistency across models
Where the three models agree or differ, and which third parties occupy the shortlist when your brand is absent.
05
Actionable priorities
Which messages, proof and sources to strengthen first, with methodology, date and limitations documented.
Ranking + matrix + Source Share · evidence included · ready to share
Illustrative example
This is what a real finding looks like
The demo shows the result, not the automation behind it. The brand, category and data in this example are fictional.
Illustrative example
Category result · 3 models
Bright Peak appears in 8 of 10 shortlists. [Your Brand] appears in only 2 and never in the top three. Four category brands fail to appear in any model.
Source Share: media 34% · reviews 27% · communities 18% · official websites 12% · other sources 9%.
Top priority
Build independent evidence around the speciality models already attribute to category leaders; the official website alone is not creating enough authority.
Illustrative excerpt. The full report preserves 30 answers and codes them into 300 brand–prompt–model records.
Who is behind it
AI with brand judgement.
Felipe Stalman is an industrial engineer researching how AI systems discover, understand and recommend brands. At TOTEM Branding he connects that research with strategy and positioning, and shares what he learns as an international speaker.
Applied research
AI and brands
International speaker
Brands and organisations that have trusted Felipe
An entry point, not a cut-down report
What the free brand check reveals and what the Reality Check adds
The free brand check shows whether something deserves attention. The €299 diagnostic broadens the evidence, explains the competitive picture and prioritises what to do.
AI brand check · Free
Reality Check · €299
Questions
AI brand check · Free1 real decision
Reality Check · €29910 buying prompts
Competition
AI brand check · FreeWho appears naturally
Reality Check · €29910 brands compared
Sample
AI brand check · Free18 answers reviewed
Reality Check · €29930 answers · 300 comparable records
Reading
AI brand check · FreeInitial signal
Reality Check · €299All 5 index layers
Sources
AI brand check · FreeOne initial clue
Reality Check · €299Source Share and key domains
Action
AI brand check · Free1 first priority
Reality Check · €2995–10 recommendations
Interpretation
AI brand check · FreePersonalised synthesis
Reality Check · €299Report + 20-minute call with Felipe
The 20-minute video call with Felipe is included in the €299 price to interpret the report, answer questions and decide what to tackle first.
Clear price
€299
One-time payment. No subscription.
Committed delivery time
10 business days
from payment confirmation and receipt of the complete information
The diagnostic includes
- 10 brands × 10 prompts × 3 models
- 30 answers · 300 comparable records
- Ranking, Source Share and verbatim evidence
- 5–10 prioritised recommendations
- 20-minute interpretation video call with Felipe
If the diagnostic falls short of what you need, Felipe reworks it at no extra cost.
Felipe reviews the request and confirms the scope with you first. You then receive an invoice and the agreed payment method. Nothing is charged on this page.
Frequently asked questions
What to know before you start
Can I not do this myself for free?
Yes. If you only want to check whether you appear, you can open several assistants, ask a few questions and save the answers. The harder part is turning those isolated checks into a consistent comparison: choosing the right brands, writing questions that do not lead the answer, recording positions and sources, separating a one-off result from a pattern, and deciding what to change afterwards. The Reality Check does that work for you and delivers it with documented evidence, priorities and a call to interpret them. If you only need an initial signal, start with the free check.
What is the AI Brand Preference Index?
It is the methodology behind the Reality Check. It measures Discoverability, Understanding, Evidence, Trust and Preference across ten brands, ten prompts and three models. The report keeps the metrics separate so the result remains interpretable rather than relying on one opaque score.
Is this just SEO analysis under another name?
No. Traditional SEO looks at how you appear in search engines. This diagnostic analyses how language models describe, compare and recommend you when a customer asks a question in natural language.
How long does it take?
The committed delivery time is 10 business days from payment confirmation and receipt of all the required information.
Is the call with Felipe included?
Yes. The price includes a 20-minute video call to interpret the report together, answer questions and decide which actions to tackle first. It is not a sales call and does not imply hiring a follow-on service.
How does payment work?
Once the engagement is confirmed, Felipe sends an invoice and the agreed payment method. This may be a bank transfer or a secure payment link. You do not enter card details on this website and there is no automatic charge.
What if the AI responses change later?
Variation between runs is normal. The report includes the analysis date and observed evidence, giving you a clear reference point. If you later need follow-up monitoring, it can be considered separately.
What if the report doesn't work for me?
Felipe reworks it at no extra cost until it covers what you need.
Request
Request your Reality Check
Share your brand and up to three priority competitors. Felipe will propose the rest of the category and confirm all ten brands with you before running the diagnostic.
€299 · One-time payment. No subscription.
10 business days