AI visibility · three packages · from €800

AI visibility packages: ChatGPT, Perplexity and Google describe you from reliable sources

An assistant describes a brand from the sources it trusts: Wikidata, Wikipedia, Commons, independent media. If you are not there, it guesses. We build those sources and measure what AI says afterwards.

WikiBusines: Starter €800 (about 2 weeks), Standard €1,500 (3–4 weeks) and Enterprise €3,500 (6–8 weeks); the AI audit is a separate product and is not credited.

Reviewed by Roman Melnyk, Chief Editor · 13 November 2026

Request an AI visibility audit →Packages: €800 / €1,500 / €3,500

Outcomes as a measurable probability, not a guarantee: model makers control the answers
Each package includes everything below it; fixed price, the audit is a separate deliverable
Every finding in the report has a prompt and a source you can repeat yourself

sources → answer

auto

What AI finds about you

company-n.comown website

“Innovative solutions for business.” No dates, address or founders on the site.

WikidataQ-item

instance ofbusinessinception2017headquartersLondonfounded byE. Carter, J. Walker

Wikimedia CommonsCC BY-SA 4.0

Company N logo.svg · 1200 × 400

Trade mediaindependent

“SMB payment platforms market, 2025”: Company N among the five largest by client count.

ChatGPT

What is Company N and where does it operate?

I have no reliable information about “Company N”. Judging by its website, it is a software vendor founded probably in 2018 or 2019, headquartered in Berlin. I suggest checking the official website.^[1]

1company-n.com · home pageown source only

industry: guessedyear: wrongcity: wrongfounders: missing

What AI finds about you

company-n.comown website

“Innovative solutions for business.” No dates, address or founders on the site.

WikidataQ-item · 23 properties

instance ofbusinessinception2017headquartersLondonfounded byE. Carter, J. Walker

Wikimedia CommonsCC BY-SA 4.0

Company N logo.svg · 1200 × 400

Trade mediaindependent

“SMB payment platforms market, 2025”: Company N among the five largest by client count.

Gemini

What is Company N and where does it operate?

Company N is a London-based fintech company founded in 2017 that builds a payment platform for small businesses. Founders: Emily C. and James W.. It operates in the UK and Germany.^[1]^[2]

1Wikidata · Q-item with 23 propertiestrusted source
2Wikimedia Commons · logo, CC BY-SAimage

industry: accurateyear: accuratecity: accuratefounders: accurate

What AI finds about you

company-n.comown website

“Innovative solutions for business.” No dates, address or founders on the site.

WikidataQ-item · 23 properties

instance ofbusinessinception2017headquartersLondonfounded byE. Carter, J. Walker

Wikimedia CommonsCC BY-SA 4.0

Company N logo.svg · 1200 × 400

Trade mediaindependent

“SMB payment platforms market, 2025”: Company N among the five largest by client count.

Perplexity

What is Company N and where does it operate?

Company N is a London-based fintech company founded in 2017 that builds a payment platform for small businesses. According to a 2025 industry review, it is one of the five largest SMB payment platforms in the UK; it also operates in Germany.^[1]^[2]^[3]

1Wikidata · Q-item with 23 propertiestrusted source
2Trade media · payments market review, 2025independent
3Knowledge base · company profilecorroboration

facts: accuratemarket position: sourced3 independent sources

20+sourced properties in the Wikidata item

3assistants in the report: ChatGPT, Claude, Perplexity

6months of monitoring in Standard

5Wikidata languages in Enterprise: EN / UA / ES / PT / FR

01 The stack AI reads

How AI visibility is built: layers of sources from the base up

ChatGPT, Perplexity, Gemini and Google AI Overviews draw on a shared base of sources. We build the layers in order, from the base up.

Wikidata: the machine-readable entity

A Q-number is the entity ID behind most confident AI answers and Google's Knowledge Graph. One item that search, assistants and every Wikipedia language point to.

instance ofbusinessinception2017headquarters locationLondonindustryfinancial technologyofficial websitecompany-n.com

Included in

StarterStandardEnterprise

Wikipedia: the encyclopedic anchor

One of the sources assistants cite most often and part of every major model's training corpus. A page is possible only where notability rests on independent sources; Enterprise adds 2–3 language editions where a page is defensible.

Company NTypePrivateIndustryFinancial technologyFounded2017, LondonFoundersE. Carter, J. WalkerWebsitecompany-n.com

Included in

StarterStandardEnterprise

Wikimedia Commons: licensed images

A freely licensed logo and photo feed the Google Knowledge Panel and the images in answers. We attach the file to the Wikidata item with the “logo image” property.

SVG

File:Company N logo.svg
Own work · uploaded by the rights holder
CC BY-SA 4.0

Included in

StarterStandardEnterprise

Multilingual: trust per market

An AI answering in Spanish pulls Spanish sources first. Labels and descriptions of the item in other languages open accurate answers in the markets you need.

ENCompany N · fintech companyUAКомпанія N · фінтех-компаніяESCompañía N · empresa fintechPTCompanhia N · empresa de fintechFRSociété N · entreprise fintech

Included in

StarterStandardEnterprise

Supplementary knowledge bases: corroboration

Open knowledge bases and industry wikis repeat the same facts and corroborate the core item. They do not carry Wikipedia's weight. One placement in Starter, four in Standard.

ProfileCompany NFounded2017HQLondon, United KingdomSourceWikidata Q-item, industry report 2025

Included in

StarterStandardEnterprise

Wikipedia AEO: the encyclopedic backbone →Wikidata and the Knowledge Graph →Are your sources enough for a page →

02 What we actually build

Wikidata item for AI visibility: 20+ properties with sources

Every package starts with a Wikidata item in which each statement has a source. On the left the item fills in; on the right, why each group of properties matters.

Company N (Q000000000)British financial technology companyAlso known as: Company N · CompanyN

instance ofbusiness1 reference · company registry
legal formlimited liability company1 reference · company registry
official nameCompany N LLC1 reference · company registry
inception20171 reference · company registry
headquarters locationLondon1 reference · official website
countryUnited Kingdom
street address1 Example St, London1 reference · official website
industryfinancial technology1 reference · industry report 2025
member ofnational fintech association1 reference · association member list
founded byE. Carter · J. Walker1 reference · interview, 2024
chief executive officerE. Carterstart time: 2017
employees120point in time: 2025 · 1 reference
product or material producedpayment platform for small business1 reference · official website
operating areaUnited Kingdom · Germany1 reference · industry report 2025
official websitehttps://company-n.comlanguage of work: English
logo imageCompany N logo.svgCommons
X usernamecompanynidentifier
Instagram usernamecompany.nidentifier
LinkedIn company IDcompany-nidentifier
Crunchbase organization IDcompany-nidentifier
OpenCorporates IDua/00000000identifier
Legal Entity Identifier0000000000000000000Nidentifier
Google Knowledge Graph ID/g/11n0000000identifier

23 statements · 12 references

1
Type and name
“instance of”, legal form and official name: the assistant stops confusing the company with a person or product of the same name.
2
Founding date
The most frequent error in AI answers about brands. One sourced statement closes it.
3
Location
Headquarters, country and address decide in which markets and languages you are mentioned.
4
Industry and membership
Industry and trade associations tie you to category queries, not only to your name.
5
People and size
Founders, chief executive, headcount: the right answer to “who founded it” and “how big is it”.
6
Products and markets
What you make and where you operate, as statements that can be quoted.
7
Website, logo, social
The official domain, the Commons logo and social profiles stitch the item to your own channels.
8
Identifiers
LinkedIn, Crunchbase, OpenCorporates, LEI, Google Knowledge Graph: external IDs link the item to other databases.

For comparison: items of well-known brands carry 30–45 properties: monobank 34, Revolut 31, Nova Post 45 (Wikidata, November 2026).

03 Three packages

AI visibility packages: three packages at a fixed price

Fixed scope and price per engagement, in EUR with taxes included. Each higher package includes the one below.

AI Visibility Starter

Starter

€800one-time · about 2 weeks

Small businesses, solo founders and pre-product-market-fit startups for whom a Wikipedia page is not yet defensible.

A Wikidata item with 20+ sourced properties: founding, industry, location, people, products, identifiers, official links
Wikimedia Commons: your logo and one hero image, correctly licensed
One placement in a knowledge base relevant to your sector
One-time delivery, no ongoing commitment

OutcomeAn entity in the Knowledge Graph and first measurable AI citations. A Google Knowledge Panel usually appears within 3–4 months; Google itself creates it.

Order

Standard

€1,500one-time · 3–4 weeks · 6 months of monitoring

Mid-market companies without a classic Wikipedia case: crypto, niche B2B, regional brands.

Everything in Starter
Three more knowledge-base placements
One placement in a tier-2/3 outlet
6 months of monitoring and a quarterly refresh of properties, images and links

OutcomeA broader Knowledge Graph footprint, measurable citations in ChatGPT, Claude and Perplexity, a steadier Google Knowledge Panel.

Order →

AI Visibility Enterprise

Enterprise

€3,500one-time · 6–8 weeks · +€1,500 a year of support

Corporate marketing and PR, IPO candidates, regulated industries: fintech, biotech, fashion.

Everything in Standard
A second media placement: two in total
Wikidata in five languages: EN / UA / ES / PT / FR
Presence in 2–3 Wikipedia language editions where a page is defensible
A citation baseline and a monthly “what AI says about you” report across ChatGPT, Claude and Perplexity
A source-correction plan for inaccurate or negative AI answers
Yearly support at €1,500 a year: updates, monitoring, defence

OutcomeAn entity-accuracy program with monthly reporting and a client dashboard: the version for the board.

Order

What's included	Starter	Standard	Enterprise
Wikidata item20+ sourced properties	20+	20+	20+
Wikimedia Commons imageslogo and hero image	✓	✓	✓
Knowledge basescorroborate the core item	1	4	4
Media placementstier-2/3 outlets	–	1	2
Monitoring and quarterly refreshproperties, images, links	–	6 mo	6 mothen yearly support
MultilingualWikidata in 5 languages, 2–3 Wikipedia editions	–	–	✓
LLM citation measurementbaseline and monthly report	–	–	✓
Plan for an inaccurate answersource correction	–	–	✓
Yearly supportupdates, monitoring, defence	–	–	€1,500 / year

work plan by weekStandard · 3–4 weeks

w1w2w3w4w5w6w7w8

Source auditwhat exists, what is missing

Wikidata and Commonsitem, sources, files

Knowledge bases1 in Starter, 4 in Standard

Media1 in Standard, 2 in Enterprise

Multilingual5 languages, 2–3 Wikipedia editions

Citation measurementbefore the start and after

MonitoringStandard 6 mo, Enterprise a year

source workmonitoring and refreshcitation measurement

Starter: what you receive

The Q-number and link to the Wikidata item with all statements and sources
Commons files with their licence
The link to the knowledge-base placement
A short report: what was done, what to check in a month

Standard: in addition

The list of placements and the publication with URLs
A quarterly refresh of the item with a change list
Alerts on third-party edits to the item for 6 months

Enterprise: in addition

A baseline: what three assistants say about you before the start
A monthly report with the prompt and source of every finding
A source-correction plan
A client reporting dashboard

Prices include taxes, per engagement. The AI visibility audit (€490) is a standalone deliverable and is not credited toward packages. We state outcomes as a probability we can measure: our guarantees →

The full price list →AI audit: what AI says about you today →Booster Pack: five wiki platforms →

04 Separate channels

Reddit and Quora for AI visibility: discussions assistants cite

Assistants often take answers from Reddit threads and Quora answers. These channels are bought separately from the packages.

one-time campaigns

Posts and comments from disclosed accounts in subreddits where your topic is already discussed. No purchased votes.

Package	Price	Volume
Newbie	€1,500 one-time	4 weeks · 4 posts, 8 comments
Seasoned	€2,300 one-time	6 weeks · 8 posts, 16 comments
Ultimate	€3,800 one-time	8 weeks · 12 posts, 24 comments

Reddit packages →

Quora

monthly retainer

In-depth expert answers to the questions your buyers search, from a profile with a real biography. Answers rank in Google and get cited by AI.

Package	Price	Volume
Foundation	€1,190 / mo	5–7 answers a month
Authority Builder	€1,955 / mo	10+ answers a month
Enterprise	€3,230 / mo	20+ answers a month

Quora packages →

05 The honest difference

What we do for AI visibility: public sources, not a paid slot in a model

Models have no paid memory slot. We work with the public sources they read.

What we don't do

We don't force AI to mention you in every answer: no one honestly can
We don't inject content into a model's training or retrieval: that is not for sale
We don't guarantee that a given AI will say a given thing: providers control their own logic
We don't create a Wikipedia page where there are no independent sources: it gets deleted and the trace stays

What we actually do

Document your entity in the sources AI already relies on
Make the facts accurate, complete and neutral, with a source for every statement
Measure real AI citations over time
Tell you before the start if your sources are too thin for a higher step

In short: we don't make AI talk about you; we make sure that when it does, it is accurate.

Wikipedia, Wikidata and AI search →What we guarantee and what we don't →

06 How we prove it

Measuring AI citations: a prompt, an answer and a source for every finding

Every finding in the report has a prompt, an answer and a source you can repeat in your own browser.

citation report · Company N

auto

Query	ChatGPT	Claude	Perplexity
What is Company N?brand	doesn't know	doesn't know	doesn't know
Payment platforms for small business in the UKcategory	not mentioned	not mentioned	not mentioned
Who founded Company N?fact	doesn't know	doesn't know	doesn't know

Answers with a source

0of 9 queries × assistants

Finding cardPerplexity · baseline

“I have no reliable information about Company N. Judging by its website, it is a software vendor…”

promptWhat is Company N and where does it operate? Name your sources.sourcecompany-n.com (own source)facts1 of 4 accurate

3assistants in every snapshot: ChatGPT, Claude, Perplexity

3 typesof query: brand, category, fact; the same wording every month

4 statesof an answer: doesn't know · mentions without source · wrong · cites a source

1 promptper finding, so you can repeat the result yourself

Tied to a live source

Each finding links to the source the AI cited: a Wikidata item, a Wikipedia article or a publication, not a screenshot.

A query you can repeat

Each finding comes with a prompt for Perplexity or ChatGPT. Answers vary between snapshots, so we record the date.

Trend in Enterprise

The monthly report shows how mentions in ChatGPT, Claude and Perplexity change. The baseline is part of the AI audit.

AI audit from €490: the baseline →AI Visibility hub →

07 Which step fits

AI visibility packages in practice: three typical situations

Most clients start with Starter and move up after the first results.

startup before product-market fit

Not a single independent source

A website and social accounts, no press, no defensible Wikipedia page. Assistants confuse the name or invent facts.

no entity→Wikidata item, Commons logo, knowledge-base profile

Within the first month a machine-readable entity exists; “what is …” queries get a sourced answer.

Starter · €800

regional brand, crypto, niche B2B

Mentions exist, Wikipedia is not an option yet

A few trade-media pieces without facts about the company. Assistants mention the brand without sources and with wrong dates.

scattered mentions→Wikidata, 4 knowledge bases, a publication, 6 months of monitoring

Independent sources multiply; category queries start citing media; third-party edits to the item get a response.

Standard · €1,500

corporation, fintech, IPO candidate

Accuracy is a board-level question

Presence exists, but across several markets and languages where answers are not controlled. Reporting and a plan for a wrong answer are needed.

scattered markets→Wikidata in 5 languages, 2–3 Wikipedia editions, monthly report, correction plan

Each market gets sources in its own language; citation trends every month; a wrong answer has a procedure.

Enterprise · €3,500 + €1,500/year

08 FAQ

AI visibility packages: frequently asked questions

Why does Wikipedia matter for ChatGPT and AI search?

Wikipedia and Wikidata are among the most consistently cited sources in assistant answers and are part of every major model's training data. When verified, neutral information about a brand exists in these sources, the model describes it from them rather than from fragments of a website. The packages build exactly this layer: Wikidata as the entity, Commons as images, media as independent sources, Wikipedia where a page can be justified.

How is AI visibility different from SEO?

SEO works on the ranking of pages in search. AI visibility works on what an assistant knows and cites about a brand when it composes an answer: the right facts, sources and images. For AI, the entity in the Knowledge Graph and trusted sources decide, not keywords on your website.

Do you guarantee my brand will appear in ChatGPT answers?

No. No vendor controls what a model will say: its makers do. We guarantee the work in the sources and the measurement: the entity is created, the media pieces are out, changes in citations are recorded in a report with prompts you can repeat.

Can I get a Google Knowledge Panel without a Wikipedia page?

Yes. A Knowledge Panel often appears based on Wikidata, consistent data on the official website, licensed images on Commons and independent mentions. It usually takes 3–4 months after the item and sources are ready.

Do you guarantee a Google Knowledge Panel?

No. Google creates and updates the panel under its own rules. We make the entity complete and consistent in every source it reads. The panel usually appears within 3–4 months, but the timing is not in our hands.

Which knowledge bases, and why not all at once?

Open knowledge bases and industry wikis where the company profile is supported by the same sources as the Wikidata item; we pick the list for your industry and country after the audit. They corroborate the core, so Starter has one and Standard four.

What happens after six months of Standard, or if someone edits the item?

Anyone can edit a Wikidata item or a Commons file. In Standard we watch the item for 6 months and refresh it quarterly; after that you can move to yearly support. In Enterprise monitoring runs for a year, with a source-correction plan for an inaccurate AI answer.

Can I start with Starter and move up to Standard?

Yes, that is the usual path. You pay the difference and nothing is redone: each higher package already includes the one below.

All questions and answers →

→ Flagship