We improve the information that prospects and AI-assisted search systems use to place your company: unambiguous services, concrete project experience and an accessible website. Our focus is companies in Germany, industrial and B2B in particular. We tie the editorial work to SEO and to the technical implementation.
What may be missing from your website
A service page should not only name what you offer. It should explain what the service is for, how a job runs, and what a prospect needs in order to make an enquiry. A case study adds what you actually delivered: the starting position, the decisions, the functionality and results with a period attached.
We also check whether the details about the team, location, prices and services contradict one another. Statements of fact are confirmed with your company. Confidential client information does not have to become public for this; an approved, anonymised case can explain the process just as concretely.
How GEO fits into the ongoing work
We settle the relevant provider and service questions, document the answers of selected AI searches and check how your company is presented. Errors and missing information turn into concrete tasks for existing pages, case studies and technical settings. After implementation, we compare new checks against the same method.
We distinguish between a bare mention, a recommendation, a linked source and a demonstrated enquiry. Those signals carry different weight. A score from an external tool does not replace the distinction.
What GEO support costs
GEO can be part of our ongoing SEO support for €600 per month including 19% VAT. We use around ten working hours per month as a guide; tasks and priorities are agreed. Tasks are prioritised as needed. Extensive investigations or additional development are agreed separately. What the retainer covers
What GEO means
Generative Engine Optimisation covers the measures companies take to make their content easier for AI-assisted search and answer systems to understand and verify: unambiguous service descriptions, substantiated project write-ups and a technically reachable website, for instance.
GEO does not replace SEO. Google still describes the fundamentals of search engine optimisation for its AI search features; no special text length or AI markup is prescribed. Source: Google Search Central. The basics: What is GEO?
What we check first
We start with whether the company name, the people responsible, the services and the actual area of operation are described unambiguously. Contradictory prices or an unclear provider help neither prospects nor automated systems.
Then we review selected content: does a section answer a concrete question? Is it understandable without the rest of the page? Can a figure be tied to its period and source? Are real references kept apart from mere promises?
Technical accessibility is part of the starting point too. Rules for search crawlers and training crawlers are not identical; OpenAI's documentation describes different jobs for OAI-SearchBot and GPTBot, among others. Source: OpenAI crawler documentation. The technical side: technical SEO.
Understandable sections instead of artificial text blocks
A useful section names its topic, answers the associated question and explains the important limits. A pricing answer should say, for example, whether it concerns building a website or ongoing support. A case study should separate what was implemented from what was measured afterwards.
So we structure content by real information needs. That does not mean repeating the same claim on every page or creating a new URL for every conceivable question. Search systems decide for themselves what they retrieve, cite or take into account.
Examining visibility in a way that can be followed
An examination starts with a fixed set of relevant questions: provider questions, service questions, comparison questions. For each run, the system, date, language and available search settings are recorded.
Recorded separately: Is the brand mentioned? Is one of your own pages linked? Which source supports the statement? Is the account factually correct? Several runs help reveal variance. A single answer is not a stable market position.
alkon in a documented sample of AI search answers
On 17 September 2026, alkon was examined against German-language supplier searches for aluminium series production. Across five fully displayed Google AI Overviews, alkon was named twice. The questions were run in a signed-out session with Germany as the target market. A sixth attempt did not produce a complete answer and was not counted as a non-mention.
| Question checked (original German) | alkon named |
|---|---|
| Aluminiumteile Serienfertigung Hersteller Deutschland | Yes |
| CNC Aluminium Serienfertigung Deutschland | No |
| Hersteller Aluminium Frästeile Großserie Deutschland | No |
| Aluminiumprofilbearbeitung Serienfertigung Deutschland | No |
| Aluminiumteile nach Zeichnung Serienfertigung ISO 9001 Deutschland | Yes |
A separate Google AI Mode answer named alkon as well. A separately examined Perplexity answer did not. Neither of those single answers belongs to the denominator of the five AI Overviews. ChatGPT and the standalone Gemini application were not tested in this examination.
The sample shows that alkon was visible in certain answers. It demonstrates neither a stable 40 percent recommendation rate nor an improvement over a comparably measured earlier state. Results can change with the question, the system, the moment and the search context.
See the website relaunch and implementation for alkon
Why case studies and named authors matter
Concrete project information makes service descriptions easier to verify. The alkon case, for example, shows how requirements, design alignment and content management fit together. A named person makes clear who checked the technical claims.
Those details are valuable to readers first. Whether a given AI system later uses them as a source cannot be guaranteed. alkon project · About WebAktive
FAQ
Can you get us placed in ChatGPT?
We cannot sell a guaranteed placement or recommendation. We can work on content and technical prerequisites and examine, in a documented way, how your brand appears in selected answers.
Does every page need special GEO chunks?
No. What helps are understandable sections that answer real questions. There is no universally prescribed word count that secures inclusion in AI answers.
Is an llms.txt file enough?
No. Such a file is no substitute for content, technical accessibility and verifiable company information. For Google Search it is not required according to current documentation.
Is GEO the same as Google Maps?
No. Local search optimisation deals with regional searches and business profiles, among other things. GEO in this offer means generative search and answer systems. For local: local SEO.
Let's start with a starting point that can be checked
Send us your website and the questions where potential customers should find your company. We will work out which examination and which work make sense. Discuss your GEO starting point
Further reading: the alkon project and implementation · What is GEO? · technical SEO · what the retainer covers