Knowledge on AI visibility and agentic search

Explaining rather than selling.

This section answers questions without turning into an offer at the end. The service pages sell, this section explains. Mixing the two makes both worse.

Agentic Search The four disciplines

Romana and Florian Hieß back to back, each carrying a stack of technical books

Overview

Every topic by area.

If you do not know where to start: the boundaries article sorts out the four terms that constantly get mixed up.

Fundamentals

Measuring and visibility

Technology and sources

Agentic systems

How these texts are built

So they stay quotable.

Every piece opens with a definition in two sentences that works without the rest of the page. That is not a quirk: this is precisely the passage a system lifts out when it answers the question.

Where we measured something, a date sits next to it. Where we are guessing, that is stated too. This field is too young for certainties, and texts that pretend otherwise age badly.

What is not here: invented figures, study results without a source, and forecasts about market share. When we quote a number you can check where it came from.

Common questions

About this section.

Where do I start if I have never heard of any of this?

With the boundaries article on GEO, AEO, LLMO and AI SEO. It sorts the four terms out and says which one matters for you. After that one further page is usually enough.

Are these texts written with AI?

Partly with AI support, as set out on the AI transparency page. Technical claims, assessments and anything that counts as an assertion come from people and get checked before publication.

Why do some pages carry a review date?

Because this field moves fast. A text without a date does not tell you whether it is from last week or the year before last. Pages that go stale quickly carry the date visibly at the top.

May I quote from these?

Gladly, with attribution and a link. If you want to take whole passages, get in touch first. What we do not want is verbatim reuse without a note, because then two pages compete for the same answer.

Is there a newsletter?

No. Anyone on the list for a specific topic, such as advertising in AI systems, gets an email when there is something to say. There is no regular send-out without an occasion.

Why are some topics here rather than under services?

Because they explain rather than sell. Most knowledge pages have a service page covering the same subject from the offer side. They link to each other.

How often does something new appear?

Irregularly. We write when we have measured or learned something worth writing down. An editorial plan producing pages by volume would be counterproductive in this field.

Is the content available in German too?

Yes, every page exists in both languages. The switch sits in the header. The German version is its own translation rather than a machine output.

What if a text is no longer correct?

Then write to us. We would rather correct it than leave a wrong claim standing, precisely because these texts can be cited by systems. The contact route is in the footer of every page.

Can I fetch these texts programmatically?

Yes. Every page also exists as Markdown at the same address plus index.md, and an ordered short version of the site sits at llms.txt.

Next step

Let us find out where you stand in AI answers.

One conversation, 30 minutes, no sales pressure. You leave with an honest read on whether this is worth the effort for you.

Look at the audit first Request a call

Romana and Florian Hieß, she holding a tablet, he holding a football