Services
We do not sell a list of services. We sell an order of work: measure first, fix second, check again third.
Until you know what the machines say and where they take it from, any fix is made blind. The measurement shows which work is needed and which is not.
Where to start
Find your situation.
Do you know what assistants say about you?
Start with the measurement (AI Visibility Baseline). →Does an assistant confuse you with another company or person?
Measurement (AI Visibility Baseline), then work on the Wikidata record. →Have the answers gone stale after a rebrand or a change of structure?
The deeper measurement (AI Visibility Baseline, Premium). →Does your site say one thing and outside sources another?
A structured data check of the website. →Is Reddit already turning up in assistant answers?
A Reddit exposure measurement (Reddit Exposure Baseline). →Are you considering an article in the English Wikipedia?
A source eligibility assessment (Wikipedia Eligibility Report). →Do you want a Wikidata record but do not know whether it is possible?
A feasibility check (Wikidata Readiness Audit). →Are you looking for a supplier to work on your own projects?
The page for agencies. →
Five directions
Measurement (AI Visibility Baseline)
We check what ChatGPT, Gemini, Perplexity and Google AI Overviews say about you right now. Each question runs at least five times across different days, we count the share of correct answers and fix the result to a date. Three levels of depth, from an automated report to a full analysis with a plan of work.
More on the measurementWikidata records
Wikidata is an open base of facts about organisations, products and people. Google and AI assistants read it directly, and a record there does not depend on your website. We correct existing records (Wikidata Repair), create missing ones where sources support the facts (Complete Wikidata Creation), and link them together: the organisation, its units, products and executives (Corporate Entity Architecture). For universities and researchers we connect academic identifiers (Advanced Institutional Entity).
More on Wikidata records →Website data (structured data check)
This is not a redesign. We check whether what the site tells machines matches what outside sources say. From the results we write a precise brief for your development team and check the outcome once it is live.
More on website work →Owned surfaces
Extra open resources that assistants can cite: a personal site for an executive, a product resource, a reference page, a project or research page. We build them on our platform, where the markup and the links are part of the foundation. The owner of every surface is named openly.
More on owned surfaces →Reddit
We check whether Reddit discussions turn up in assistant answers (Reddit Exposure Baseline). We set up an official presence and, where it makes sense, build your own community with the affiliation disclosed. This direction works above all for companies with an international audience: the platform runs in English.
More on Reddit →Separately: the Wikipedia eligibility check (Wikipedia Eligibility Report)
We check whether there are enough independent publications for an article in the English Wikipedia. This is not the writing of an article. It is a verdict on the odds and a plan for strengthening the source base.
More on the assessment →
Limits
We do not control assistant answers directly. Nobody controls them. We fix the input those answers are built from and measure what changed against the date of the first measurement.
An audit changes nothing, writes no content and guarantees no place in assistant answers. It is a measurement and a set of priorities.
We do not promise a Google Knowledge Panel and we do not push negative material out of search results. What we guarantee is the work performed and a measurable change in the input data.
The work is not possible when there is nothing verifiable to support the facts, when the community would read it as self-promotion, or when we are asked to hide who commissioned it.
Common questions
- Which service should I start with?
- The measurement (AI Visibility Baseline). It shows what the machines say and where it comes from, and it decides what work follows. Sometimes the answer is "fix the markup on the site, nothing else is needed", and that is a result too.
- Do I have to choose a direction up front?
- No. Send the website address and describe the task briefly. The direction follows from the measurement, not from your guess about what is broken.
- Can I order only the fixes, without the measurement?
- You can, though we advise against it. Without a measurement it is unclear which layer to repair, and the work often lands somewhere other than the problem.
- What does it cost?
- The terms depend on scope: the number of questions and languages in the measurement, the depth of analysis, the state of the site and the records. We set them out after a short brief, and for that the website address is enough.
- What do we get at the end?
- A dated report in Word and PDF: which answers are stable, where the machine gets it wrong, where the error comes from and what to do about it. The higher levels add a consultation.
How to open the conversation
Send us your company website and describe the task briefly. Our first reply contains an observation about your data, not a request for a call.