Loading…
AI Visibility · Packaged
When ChatGPT, Claude, Perplexity or Google's AI Overviews answer a question about your brand, the answer is only as good as the sources behind it. We document your entity in the sources AI already trusts — so when it speaks about you, it gets it right.
The honest difference
The market is full of pseudo-AI services promising to 'force' an AI to mention you. That is either an illusion — models don't have a paid memory slot — or brute-force prompt injection, which gets banned. We do the opposite: we improve the public-source layer these systems read, so accuracy goes up and so does the probability of an accurate citation.
What we will not do
What we actually do
The frame in one line: we don't make an AI talk about you — we make sure that when it does, it's accurate. That's trust-visibility (you sit inside the trusted-source tier), not attention-hacking.
The stack AI reads
ChatGPT, Perplexity, Gemini and Google AI Overviews share a backbone of high-trust sources. Each layer below adds a distinct signal — and each one compounds the last. We sequence them deliberately, cheapest and most foundational first.
| Layer | What it is | Why AI leans on it | Tier |
|---|---|---|---|
| Wikidata | The machine-readable entity | A Q-number is the entity ID behind most confident AI answers and Google's Knowledge Graph. The cheapest, highest-leverage way to exist as a known thing, not a string of letters. | Backbone |
| Wikipedia | The encyclopedic anchor | The single most-cited source across ChatGPT, Claude and Perplexity, and a measurable share of every major model's training corpus. Only viable where notability genuinely holds — we assess first. | Backbone |
| Wikimedia Commons | The licensed image layer | Correctly-licensed logo and imagery that feed Google's Knowledge Panel and image understanding — so the visual an AI surfaces is the one you chose. | Backbone |
| Multilingual Wikipedia / Wikidata | Per-market trust | An AI answering in Spanish reaches first for Spanish sources. Additional-language entities extend accurate coverage into the markets that matter — where a page is genuinely defensible. | Backbone |
| Supplementary knowledge bases | Corroboration & source density | Open knowledge bases and community wikis add breadth and corroborate the core record. We are candid: these do not carry Wikipedia's trust weight — we use them to reinforce the backbone, never to substitute for it. | Supplementary |
We don't sell “presence on eight wiki farms.” The backbone is where the trust is; supplementary placements only ever corroborate it. Where your sourcing supports a real Wikipedia page, that remains the strongest single signal of all.
Three packages
Standardised so you know exactly what you get and what it costs. Prices are per engagement, in EUR; the currency toggle shows an approximate USD.
€700
one-time · ~2 weeks
Small businesses, solo founders and pre-product-market-fit startups where a full Wikipedia page is not yet defensible.
Outcome
A real entity in the knowledge graph. A Google Knowledge Panel becomes eligible (typically initiates within 30–60 days where the criteria are met), and the first measurable AI-citation events begin to appear.
Most popular
€1,500
~3–4 weeks · 6-month monitoring
SMEs with a non-classic Wikipedia case — crypto, niche B2B, regional brands — that want a multi-platform presence.
Outcome
A broad knowledge-graph footprint, measured AI-citation events across ChatGPT, Claude and Perplexity over 2–3 months, and a more stable Google Knowledge Panel.
€3,500
~6–8 weeks · +€1,500/yr support
Corporate marketing and PR, IPO candidates, and regulated industries (fintech, biotech, fashion) that need a measurable program.
Outcome
An entity-accuracy program with measurable trend reporting and a client-facing dashboard — the version you can show a board.
All work is source-first and disclosed. Every outcome above is framed as a probability we can measure, not a guarantee we can't keep — see our guarantees.
How clients progress
Step 1
Most clients begin with Starter — a low-risk way to put a real entity in the graph and watch the first AI citations appear.
Step 2
Standard adds corroborating sources and media density, then six months of monitoring, so the record holds and compounds.
Step 3
Enterprise turns it into a measured, multilingual program with monthly reporting and a defence plan — for brands where accuracy is a board-level concern.
How we prove it
Most AI-visibility reports ask you to take their word for it. We capture real citations live and hand you evidence you can verify in your own browser.
Linked to the live source
Every finding links to the source the AI cited — a specific Wikipedia article or Wikidata record — not a screenshot you can't check.
A re-runnable query
Each finding ships with a prompt you can paste into Perplexity or ChatGPT to see the citation appear, for yourself.
A trend, on Enterprise
Monthly reporting charts how mentions evolve across ChatGPT, Claude and Perplexity — so the probability going up is something you can see, not assert.
Want the why behind all of this? Start with Wikipedia, Wikidata & AI Search and the AI Visibility hub.
Frequently asked questions
Send your brand brief. We'll show you what AI says about you today, and which package closes the biggest gap — within 48 hours.