Pages customers and AI can understand.

We build service pages, articles, comparisons, FAQs, cases and product proof that answer real buying questions.

Companies already recommended by LLMs through getSichtbar

AGS IT-ServiceAGS IT-ServiceHealthcare IT services
William WalkerWilliam WalkerPremium dog accessories
VOLTH MaterialsVOLTH MaterialsIndustrial insulation & aerogel
voids.aivoids.aiSupply-chain SaaS
SeedwiseSeedwiseGrant consulting
Campus FiveCampus FiveAI consulting & automation
NiccosNiccosShopify & e-commerce
RAY AIRAY AIExecutive assistant service
NEONANEONADesigner lighting & e-commerce
MAYKS FahrschulenMAYKS FahrschulenDriving school group
Agentur AusdruckslosAgentur AusdruckslosPerformance marketing agency
Senseven GmbHSenseven GmbHIndustrial AI & predictive maintenance
SPIELMANN Steuerberatung GmbHSPIELMANN Steuerberatung GmbHTax consulting
Zahnzentrum St. GeorgZahnzentrum St. GeorgDental practice
Ad SpecialistAd SpecialistPerformance marketing agency
PUNKT PRPUNKT PRPR agency

What we do.

Website content with AI means we use buyer questions, search data and AI answers to decide which pages are needed. AI helps draft. Humans check the facts, structure and final message.

Owned contentWebsiteCaseArticleOwned pages become citable evidence.

01

Buyer questions

Which questions do buyers ask before they contact a provider?
Which article or service page is missing?
Which proof should support this claim?

02

Evidence

Buyer-question map
Service pages, articles, comparisons and FAQs
Proof elements, cases, tables and internal links

03

Outcome

Your website gets pages that explain your offer clearly and can be used as evidence.

Why it matters

Content only matters when it answers a real question and gives AI enough structure and proof to use it.

1,200–1,800

target word count

In this range AI systems extract roughly 35% of a page. Above 3,000 words it drops to about 13% (Dejan.ai 2024).

+115%

from citing sources

Visibility gain in AI answers from supported claims (Princeton GEO Study, KDD 2024).

+41%

from expert quotes

Gain from direct quotes with name and title (Princeton GEO Study, KDD 2024).

2

readers

Humans have to understand it, AI has to be able to cite it. Both or neither.

Many pages are filler.

They mention topics, but they do not answer the question that decides the purchase.

Blog posts bring traffic but no leads.
Service pages are too generic.
Important questions have no clear answer.
Claims are not backed by proof.

Every page gets a job.

A page must answer one real question, explain the offer and show why the answer is credible.

Result

AI should not just find your company. AI should understand when your offer fits.

From question to page.

We decide which page type is needed, then build it with structure, proof and links.

01

Find question

We start with buyer questions, search data and AI answers.

02

Choose page

Not every question needs a blog post. Some need a service page.

03

Build text

AI helps with drafts. Humans check facts and message.

04

Add proof

Cases, sources, tables, FAQs and links make the page useful.

What we actually do.

Buyer-question research
Service pages, articles, guides and FAQs
Comparison pages and case sections
Internal links and answer blocks
Human review before publishing

What you get.

Content priority list
Page briefs
Ready-to-publish website content
Proof and FAQ blocks
Internal-link plan

Good fit if you want implementation.

You can provide facts, cases or examples.
You want pages that explain and sell.
You publish regularly or want to start.

Not a fit if you only want a report.

You only want a report and no implementation.
You expect a guaranteed mention in one exact AI answer.
Your offer is not clear enough to recommend yet.

When this service is especially useful.

Your website does not answer buying questions clearly.
You need service pages, articles or guides.
You want content that supports SEO and GEO together.

What this should not be confused with.

Blog agency

We do not start with an editorial calendar. We start with buyer questions.

SEO text

A text alone is not enough. The page needs answer, structure and proof.

AI generator

AI creates raw material. We turn it into a publishable page.

What it needs.

Choose the question and page type.
Build the page with answer, proof and structure.
Publish and connect it internally.

Clear answers.

Is this just AI content?

No. AI helps with speed, research and structure, but every page is built from real questions, evidence and human review. One wrong number in an article costs more trust than the faster turnaround ever saves.

How long should an article be?

Between 1,200 and 1,800 words. Research by Dejan.ai (2024) shows AI systems extract roughly 35% of a page in that range but only about 13% above 3,000 words. Three focused articles beat one long one.

Do you write blog posts?

Yes, when a blog post is the right page type. Sometimes a service page or an FAQ answers the question better and converts more reliably. We choose the format from the intent, not from a content calendar.

Do you improve existing content too?

Yes, and it is often the stronger lever. An existing article with ranking history that gains numbers, sources and an FAQ block moves faster than a brand-new page starting from zero.

What makes content citable for AI?

Supported claims. The Princeton study (KDD 2024) measures +115% visibility from citing sources, +41% from expert quotes with names and +40% from concrete statistics. Well-written text without substance adds roughly +3%.

How many articles do we need per month?

Fewer than most people assume. Four well-supported articles beat twelve thin ones. We plan around open questions in the buying process rather than a quota — a quota reliably produces content nobody needs.

Let us check whether this service should come first.

One call is enough to see the rough direction: website, content, sources, product data or measurement.