Your customer no longer types a company name. They ask AI who to trust.
We check whether ChatGPT, Gemini, Claude and Perplexity mention your company in queries that drive purchase decisions. We show who they recommend instead of you — and what needs to change.
Free report by email within 24 h. No subscription.
› Which law firm specialises in construction disputes in the Tri-City?
Illustrative example. Real results in the report for your domain.
120+
categories measured in Poland
4 800+
queries in the database, updated weekly
4 models
ChatGPT · Gemini · Claude · Perplexity
2 400+
companies recognised in responses
Where the decision is made
Before the customer reaches your website, they already have a shortlist of three companies.
They describe the problem, budget or deadline. The model sorts the options and gives reasons why some are better than others. The company website is visited later — or not at all, if the company wasn't in the response.
This doesn't replace human decisions. But it changes who gets to participate in the selection at all.
We measure the first three. Steps four and five require access to your analytics and CRM.
Scope
Not just tech and B2B. Every category where customers compare.
If someone can ask AI "who should I choose", the category qualifies.
Professional services
› which law firm handles shareholder disputes?
Health & clinics
› where to get dental implants without overpaying?
E-commerce
› what coffee machine for an office of 15?
IT & SaaS
› what invoicing software for a sole trader?
Construction & renovation
› who does anhydrite screeds near Warsaw?
Manufacturing & B2B
› who produces short-run cardboard packaging?
HoReCa & local
› where to have a good breakfast on Sunday?
Education & training
› which accounting course is recognised by employers?
Metrics
Mere name presence tells you nothing.
A company can be mentioned and still lose — because the model describes it inaccurately, cites an outdated source, or adds a caveat. We separate five things because each requires a different response.
Whether the company appears for a given query, in which model, and how consistently across repetitions.
What attributes the model assigns to the company and whether they match the actual offering.
Who appears alongside or instead of you, and what reasoning the model gives for that choice.
Which sites the model draws on to build its response — the only element you can realistically influence.
Comparison of successive measurement waves on the same query set.
Why don't we report "positions"?
The model generates a response — it doesn't read a ranking. The same query can produce a different result an hour later. That's why we work with query sets and repeatable measurement waves, not screenshots.
How we work
From diagnosis to infrastructure. Three phases, each with its own decision point.
Visibility audit
A query set built for your category and customers, run across four models. You get hard numbers instead of impressions: share of responses, competitor list, cited sources, errors in the company description, and priorities.
→ You know where you stand and whether it's worth investing.
AEO/GEO implementation
Specific instructions: what to add to the website, what data to structure, where independent confirmation must appear, which comparative content makes sense. Implemented by your team, ours, or jointly. After an agreed window, we repeat the measurement.
→ A series of changes with a hypothesis and a way to verify it.
Verification & recommendation platform
We're building our own tool where a company confirms its data and AI agents can read it without guessing. The goal: visibility determined by what the company actually does, not how much content it has published.
→ A lasting position in the layer AI models draw from.
Each phase ends with a decision: we continue, your team takes over, or we close the project. Phase 3 is not a prerequisite for phases 1 and 2 to make sense.
Technology
One screenshot is an anecdote. A system turns it into decision material.
The platform stores queries, full responses, cited sources, recognised brands, and measurement history. This lets us distinguish a one-off occurrence from a repeatable pattern.
Query generator
Sets built automatically from search data, website content, and customer language. The control core stays constant between waves.
Category index
Preventive crawl of entire industries. Historical data exists before the client signs up.
Response archive
Full text of every response with a run ID, model, and date. Every conclusion leads back to source material.
Citation analysis
Source classification by page type: offer, ranking, article, forum, directory, comparison.
Alerts
Notification when a new brand appears in your category or your company description changes.
API & integrations
Data to your own BI, GA4 and GSC for traffic cross-referencing. MCP server for teams working with agents.
Differentiator
Visibility without verification is just louder marketing.
AI visibility tools measure what the model says. We go further: we build a company profile where key facts can be independently confirmed. Models increasingly reach for sources that can be verified. We want to be one of them.
Verification is voluntary and developed in stages. It doesn't assess company quality — it confirms facts.
Open data
We publish who AI mentions in Polish categories.
We regularly release summaries from the index: which brands appear in purchase queries, which sources models rely on, and how this changes between waves. Free of charge and with no option to buy a placement.
Law firms
Poland, wave 07/2026
View →
Dental practices
5 largest cities, wave 07/2026
View →
Software houses
Poland, wave 07/2026
View →
Summaries show presence in model responses. They are not a quality ranking or recommendation.
You're deploying AI. Someone should check what's coming into the company with it.
AI systems open an attack surface that classic IT audits don't cover: prompt injection, data leakage through integrations, uncontrolled agent permissions. Add post-quantum cryptography horizons and NIS2, CRA, and AI Act requirements.
Advisory & audit
Status assessment, risk analysis, and action map for AI systems and cryptography.
Implementation
Security design and implementation, including migration to post-quantum cryptography.
Dedicated R&D
Solutions built for specific problems when off-the-shelf products aren't enough.
Offer
Start at the level that fits your situation.
Visibility report
Free
You want to know if the problem exists
Category audit
quote
You see the symptom, not the cause
AEO/GEO implementation
quote
You have a diagnosis and a team ready to act
Continuous monitoring
subscription
AI is a real acquisition channel
For agencies and companies with multiple brands — white-label platform access. Write to us →
Background
CompanyNow was built within the AIQ group.
The AIQ group builds tools for businesses in law, tax, accounting, and security, operating in Poland, Ireland, and Spain. CompanyNow is the answer to a question clients started asking: since purchase decisions are moving to language models, who is watching how those models describe my company?
We develop the technology and measurement methodology ourselves. We publish the rules we count by and the limitations we know about.
AIQ GROUP
Holding — law, tax, accounting, security. Poland, Ireland, Spain.
AIQ CYBER
AI system security, post-quantum cryptography, NIS2/CRA/AI Act audits.
COMPANYNOW
Company visibility in language model responses. Measurement, implementation, verification.
Questions we get most often.
AI visibility measures how often and how accurately AI assistants (ChatGPT, Gemini, Claude, Perplexity) mention and recommend your company when users ask purchase-related questions.
First step
Check who AI recommends to your customers.
Enter your domain. We'll build a query set for your category, run it across four models, and send you the report. Free, no subscription, no sign-up required.
Report sent within 24 business hours. We don't call without notice.