Buying advice increasingly happens in chat. Whoever gets named there has half won the decision, because people rarely keep searching after three suggestions.
Your contact for products in AI answers
Anfrage
Schreib mir kurz, worum es geht. Ich melde mich innerhalb von 24 Stunden mit einer ehrlichen Einschätzung, ob und wie ich dir helfen kann.
Message sent!
Thank you for your inquiry. I'll get back to you as soon as possible.
Lieber direkt sprechen? 30 Minuten Termin wählen
Two shops with measurable mentions in AI answers: 653 and over 200 AI citations.
18,000+ Clicks/Day (Peak) SEO takeover of an established gold and silver marketplace. Structured SEO program led to 834,000+ clicks and €91,000 traffic value.
Read case study
2,000+ Clicks/Month Shopify shop for plant-based horse products completely rebuilt. Sustainable growth achieved purely through technology and smart content, without backlinks.
Read case studyAblauf
From what the AI recommends today to solid product data.
I put your audience's buying questions to the models and record which products and shops get named, and with what reasoning.
Then I look at whether your product data is machine readable at all: schema, attributes, availability, prices and variants.
Missing properties get added and marked up so a model can attribute them reliably instead of guessing them from running text.
Comparisons, use cases and honest boundaries give a model the arguments it needs to recommend your product.
The same buying questions at fixed intervals, so it becomes visible whether your mentions change and for which questions.
The more a product needs explaining, the more likely someone asks the AI first.
When questions have to be resolved before a purchase, people now ask the AI rather than a comparison site. Appearing there as the fitting solution skips the entire comparison.
Own products have no manufacturer text that already sits everywhere. That is an advantage, but only if you fill the gap with your own verifiable details.
In niches the data available to models is often thin, because few write about it in depth. That is exactly what makes it comparatively easy to become the cited source.
Pick an area to see what it actually involves.
I collect the questions actually asked before a purchase in your category and test them across several AI systems. The result shows which products get recommended today and why.
Price, availability, dimensions, material, compatibility: whatever is not marked up a model has to guess, and in doubt it leaves your product out. Product schema is the baseline here.
„High quality“ is worthless to a model, „925 sterling silver, 2.4 mm thick“ is usable. I make sure your product copy consists of verifiable details.
Models like to recommend when they have a reason. Honest comparisons, including the cases where another product fits better, provide exactly that reason.
A model has to know your brand as a thing in its own right, not just as a word in a shop. That requires consistent details and mentions beyond your own pages.
Category pages that only list products are no good as a source. With an orientation section explaining the selection criteria they become the content an AI refers to.
What AI crawlers cannot load does not exist for them. Shops with heavily script-driven pages in particular lose content here without noticing.
Product recommendations from models change constantly. Fixed buying questions at regular intervals show whether the work is landing and where you are still missing.
Through the A/B tests and optimizations we increased our conversion rate by 23%. The collaboration was structured and results-driven.
Richard supported us strongly with SEO and A/B testing. For some of our products we achieved a 13% conversion uplift while making performance more stable.
Great work. Richard supported us with various SEO perspectives, including AI Search. I can recommend him without reservation.
Häufige Fragen
Because it understands it better and finds a reason for it. What matters are machine-readable properties, content explaining selection criteria, and mentions outside your own shop. Pure ad copy gives a model nothing to refer to.
It is the foundation, but not sufficient on its own. Schema makes sure your product can be identified. Whether it gets recommended is decided by content explaining who it fits and who it does not.
No. Both draw on the same technical and editorial basis. A shop with messy indexing and thin product pages will do badly in Google and with an AI alike. GEO comes on top, see E-Commerce SEO.
Through repeated tests with the same buying questions. I document whether your products get named, in which position and with what reasoning. Over the months that becomes a development line, even without rankings in the classic sense.
Often precisely then. With few products that need explaining, the necessary depth is realistic to produce, whereas large catalogues first need a system for it. In niches the competition for the cited source is also small.
The providers keep building shopping features into the chat interfaces. Which form will prevail is open. What is certain is that all of them require clean, machine-readable product data. That is exactly what we create now, regardless of which interface wins in the end.
GEO stands for Generative Engine Optimization: optimizing to be recommended in AI answers from ChatGPT, Perplexity and Google AI. Here is how it works.
GEO & AI SearchHow your brand gets recommended as a source in ChatGPT and Perplexity: clear facts, strong entities, trust signals and clean technology.
GEO & AI SearchA practical GEO checklist: in 8 steps, make your brand visible as a source in ChatGPT, Perplexity and Google AI Overviews. Ready to work through.
Aktuell Kapazität für neue Projekte
Tell me what you sell. I will test a few real buying questions and tell you who gets named today.
Check product visibility