Your buyers are asking AI. Is your name in the answer?
We measure how many buying-relevant questions ChatGPT, Perplexity and Google AI name you in — and make that number go up. You see your baseline before you pay a cent.
Buying-relevant questions in your prompt set — measured across every AI system
90 days
Until your Share of Answer has measurably moved
€1,490
Per month, cancel monthly, no minimum term
Measured across every AI system that matters
ChatGPT
Perplexity
Google AI Overviews
Gemini
Claude
Copilot
Test it yourself — 60 seconds, right now
We won't show you someone else's case study.
Open ChatGPT in a new tab and ask one of these — with your industry and your city. You don't need us for this part.
★ Your self-test
Ask AI what your buyers ask it.
Not your company name — it might know that. Ask the way someone asks who doesn't know you yet:
→ "Who are the best [your industry] in [your city]?"
→ "Which [provider] specialises in [your niche]?"
→ "I need [your service]. Who do you recommend, and why?"
→ "[Competitor A] or [Competitor B] — which is better?"
Is your name in the answer? Then you probably don't need us. If it isn't — AI is recommending someone else, right now, to someone ready to buy. On every one of those questions. Every day.
Buyers who used to Google ten providers and compare them now get handed three names. If you're not one of the three, you never find out you were in the race.
🥉
The shortlist forms without you
AI names two to four providers. That's the new shortlist. There is no position five — there's "named" and there's "doesn't exist".
👻
It's invisible in your analytics
Someone who finds you via an AI answer then Googles your name or just calls. It shows up as "Direct". You never see what you're losing.
📈
Good rankings don't protect you
Ranking #1 on Google and still not being named in the AI answer is the norm, not the exception. They're two different systems.
🔒
The lead is hardening
AI systems cite whoever is already established as the authority. Whoever gets into the answers now will be considerably harder to displace in two years.
★ First things first
We're not promising you traffic.
That isn't modesty. It's the most honest sentence on this page — and the reason you'll still be with us in month 12.
✕ What others sell
"AI traffic converts 23× better!"
The number is often even true. What's never included is the denominator. Run it for a site with 10,000 visits a month:
One lead every three months. And in month 6 the client opens Analytics, sees 40 visits from ChatGPT, and cancels. Rightly — they were sold a channel and handed a rounding error.
✓ What we sell
Share of Answer
AI search barely sends clicks — it answers the question instead of forwarding it. The influence happens inside the answer. The click doesn't. So we measure the influence, not the click:
Month 1: named in 0 of 30 questions
Month 3: named in 12 of 30 questions
Main competitor: 8 → 5
Reproducible, in front of you.
The metric we promise is the metric you'll judge us on. So we promise one we can prove — and one that shows what actually happened.
The work
How you get into the answer.
AI systems don't invent their recommendations. They pull them from sources they consider trustworthy. Our job is to get you into exactly those sources — and make you findable and unambiguous once you're there.
🎯
Prompt set & baseline
We define 20–30 questions your buyers actually ask — not keywords, sentences. That becomes the yardstick for everything after.
🔍
Citation gap analysis
Which sources do AI systems actually cite in your category? Where does your competitor appear and you don't? We work from measurement, not assumption.
🧩
Entity & schema work
So AI recognises you as a distinct company and doesn't confuse you with a similar-sounding one: consistent data, structured markup, a clean company profile.
✍️
Content that can be quoted
Fact-dense, clearly structured, directly answering. Not more blog posts — content a model can cleanly lift a sentence from and attribute to you.
🌐
Citation building
Presence on exactly the sources demonstrably cited in your category — trade directories, comparison sites, forums, review platforms, industry press.
📊
Share-of-Answer report
One page a month: where you were, where you are, what moved for your competitors. With the original answers attached as evidence. No 40-page deck.
Transparency
What we don't do — and why.
GEO is a young market with a lot of promises and very little evidence. Almost every competitor of ours sells these three things. We don't.
✕
Set up & "maintain" llms.txt
Sold across the industry as a deliverable, sometimes even as a recurring monthly line item. Google's John Mueller on it: "no AI system currently uses llms.txt" — you can see in your server logs that they don't even check for it. He compares it to the keywords meta tag.
If that changes, we'll set it up in twenty minutes. As long as it does nothing, we won't bill for it.
✕
Wikipedia & Wikidata entries
A popular sales promise. Wikipedia requires paid editors to disclose their client, notability cannot be purchased, and Wikidata explicitly prohibits commercial providers from creating items about clients for marketing purposes.
That's not a deliverable, it's a reputational risk in your name.
✕
Traffic and revenue forecasts
AI visibility can't be attributed click by click — there's no query data, no click data, and the models don't answer deterministically. Anyone who builds you a revenue forecast from it made it up.
We measure what's measurable: whether you get named. We don't assert the rest.
Sources for this page
John Mueller (Google) on llms.txt: "FWIW no AI system currently uses llms.txt" — Search Engine Roundtable
Define the prompt set, measure across every AI system, Share of Answer against three competitors, citation gap, technical check. You get the report plus a one-hour call — and know exactly where you stand. Even if you don't continue.
Week 2–6
Foundation
Entity and schema work, content restructured for quotability, first citations on the sources that actually count in your category. This is the work that compounds later.
Week 7 onwards
Measure, push, repeat
Monthly measurement of the same prompt set. What works, we double down on. What doesn't, we drop. Every month you get one page with the original answers as evidence.
Your investment
What it costs.
Most providers in this market hide their pricing behind a phone call. Here's ours.
Entry
990€one-time
AI Visibility Audit. The way in — and the only honest way to say what you actually need.
All prices excl. VAT · The audit (€990) is credited in full against your first month if you continue with a retainer · Cancel monthly, no minimum term, no setup fee · Swiss clients are billed in CHF
Our promise
Share of Answer moves in 90 days — or we keep working.
🛡
Free of charge until the number moves.
In week 1 we measure your baseline — in writing, with the exact prompt set. If your Share of Answer hasn't risen after 90 days, we keep working at no cost until it does. We can offer that because the number is measurable, not interpretable: either AI names you or it doesn't. There's nothing to spin.
★ Why there are no testimonials here
The most honest section on this page.
We won't show you case studies we can't prove.
Before building this page we read through ten German-speaking GEO agencies. Not one shows a verifiable client result. They all cite the same borrowed studies. One puts "74% Share of Answer" prominently on the page — with "illustrative representation" in small print underneath.
We could put three invented quotes from "Sandra M., consulting, Graz" right here. Half the industry does. We don't — because if we're explaining on the same page that we skip llms.txt since it demonstrably does nothing, we can hardly invent clients underneath it.
What we offer instead: your own measurement. Your questions, your competitors, your numbers — before you book a retainer. It's verifiable, it's reproducible, and you can check it yourself in the moment you're reading this sentence. No testimonial on earth is worth as much as a screenshot of your own baseline.
Who does the work
No account manager.
Founder
Christoph Sacher
ResultStars — visibility, funnels, branding
I've been doing organic growth since "GEO" and "AEO" weren't acronyms yet. In that time the rules have shifted from links to topics to AI citations — and with them, the amount of nonsense you can sell on the back of it. At ResultStars we don't sell that.
"GEO is the most verifiable marketing service there is. You run the questions before, and you run them after. Which is exactly what makes it so telling that almost nobody in this industry shows their results."
— Christoph Sacher
10+
Years experience
180+
Projects shipped
30
Questions per baseline
6
Retainers maximum
✓ Cancel monthly — no 12-month lock-in
✓ You work with me directly, not with a junior
✓ Six retainers at a time, maximum — otherwise the depth goes
FAQ
Answers to common questions.
The share of buying-relevant questions where an AI system names your brand in its answer. We define a fixed set of 20–30 questions with you, measure it monthly across ChatGPT, Perplexity, Google AI Overviews, Gemini and Claude, and compare it against your competitors. Example: "named in 12 of 30 questions" = 40% Share of Answer. It's the one metric in AI search that can be measured and verified cleanly.
SEO gets you to position 1 in the blue links. AI visibility gets you into the generated answer. They're related but different systems — ranking #1 on Google and still not being named in the ChatGPT answer is the norm. Honestly: a substantial part of the work is good, classic SEO foundation. Anyone telling you otherwise and selling GEO as an entirely new discipline is selling you snake oil. What genuinely is new: optimising for quotability rather than clicks, work on entities and knowledge graphs, choosing the sources models demonstrably cite from — and above all a completely different success metric. If your SEO foundation is missing, we'll tell you in the audit, and we'll start there.
Very little, and we say so upfront. AI referral traffic currently sits well below one percent of total traffic — the systems answer the question instead of forwarding it. The value is elsewhere: someone who learns about you from an AI answer then Googles your name, clicks through, or calls. In your stats that appears as direct or brand search. The influence is real; the click out of the AI isn't. If what you need is a channel that measurably sends visitors next month, paid is the more honest route — happy to talk about that.
Because it's real work: developing the prompt set, measuring across five systems, benchmarking competitors, analysing the citation gap, writing the action plan. A free audit is either superficial or a sales call in disguise. If you book a retainer afterwards, we credit the full €990 against your first month — so it was effectively free. If you don't, the report is still yours, action plan included, and you're free to implement it yourself or with another agency.
Worth it: when your buyers make a considered, explanation-heavy decision and a single client is worth four figures or more. Tax advisors, law firms, private healthcare, B2B SaaS, real estate, specialist consultancies. That's where the shortlist decides everything — and AI is building it right now.
Not worth it: when your average job is €200 and customers decide on price and proximity. For a typical trades business, Google Business, reviews and local ads will do more per euro. We'll tell you that in the first call, before you pay anything.
We usually see first movement in 4–8 weeks, because competitive density in AI answers is still low — considerably lower than in classic Google rankings. Solid change in Share of Answer is typically visible after 90 days. We commit to that: if the number hasn't moved by then, we keep working at no cost until it does.
Partly, yes — and the criticism is fair. Jeremy Moser (uSERP) puts it this way: if a GEO service doesn't openly tell you that most of the success is solid classic SEO, they're selling snake oil. We're telling you openly. What's genuinely new: optimising for quotability instead of clicks, entity and knowledge-graph work, selecting the sources models demonstrably cite — and a fundamentally different success metric. Anyone selling GEO as a revolution is overselling. Anyone ignoring it loses the shortlist.
No — not as a paid deliverable, anyway. Google's John Mueller has publicly stated that no AI system currently uses llms.txt, and that you can see in your server logs that the crawlers don't even look for it. He compares it to the keywords meta tag. Plenty of agencies sell it regardless — some even as a monthly maintenance item. If you want the file, we'll create it for free. The moment it demonstrably does something, it becomes standard.
Yes. No notice period, no exit fee, no contract. One email and you're out at the end of the month. We earn the relationship every 30 days — that's a very direct incentive on our side.
Yes. Most of our retainers are in Germany, Austria and Switzerland — Swiss clients are billed in CHF — but AI visibility works the same way in any language and market. AI systems answer differently by region, which is exactly why we measure your prompt set in your market and your language, not generically.
The shortlist is forming without you.
AI Visibility Audit for €990 — credited against your first month. You get your baseline across 30 buying-relevant questions, a comparison against three competitors, and an action plan. The report is yours, whether you continue or not.
Perplexity
Google AI Overviews
Gemini
Claude
Copilot
