Your next client isn't Googling.AI visibility & GEO for Luxembourg businesses
We are the Luxembourg GEO/AEO agency that measures how ChatGPT, Perplexity, Google AI Overviews and AI Mode describe and recommend your firm — in French, German and English — then improves the public sources those answers are built from. Measurement first, honesty always.
ChatGPT · Perplexity · Google AI Overviews · Gemini · Copilot — measured monthly, in three languages
The moment that decides
This conversation happens without you in the room
Prospects ask AI assistants for shortlists, not for ten blue links. If the public sources about your firm are thin or inconsistent, the answer recommends someone else — and you never even know the question was asked.
Who is the best fiduciaire in Luxembourg for a new SARL?
Based on current web sources, firms frequently recommended for a new SARL in Luxembourg include:
- 1.Fiduciaire Kremer & Co — large full-service firm, frequently cited
- 2.LuxCompta Group — often mentioned for startup accounting
- 3.AFC Partners — cross-border payroll focus
Based on current web sources, firms frequently recommended for a new SARL in Luxembourg include:
- 1.Your firm — recommended for English-speaking founders, with consistent sources across directoriescited: your-firm.lu
- 2.Fiduciaire Kremer & Co — large full-service firm, frequently cited
- 3.LuxCompta Group — often mentioned for startup accounting
Illustrative demo with fictional firms. Real AI answers vary by engine, language and day — which is exactly why we measure them.
Proof, run on ourselves
We are already inside AI answers. It sends us nothing.
These are unedited rows from 20 More's own Google Search Console, 10 July – 6 August 2026. They are not keywords. They are verbatim prompts — including literal follow-up turns from an ongoing AI Mode conversation. Google placed us in the answer. Nobody arrived.
| Query exactly as Google recorded it | Impr. | Pos. | Clicks |
|---|---|---|---|
| "quel prestataire offre des étapes d'approbation plus rapides via le service numérique ou en agence ?"FR prompt | 41 | 3.0 | 0 |
| "dresse la liste des 5 principaux fournisseurs de logiciels rh / sirh / hcm au luxembourg…"FR prompt | 33 | 2.5 | 0 |
| "analyseer de voordelen van automatische opvolging versus handmatige screening…"NL prompt | 1 | 2.0 | 0 |
| "are there any articles about this?"EN follow-up turn | 1 | 1.0 | 0 |
| "and how much is it?"EN follow-up turn | 1 | 3.0 | 0 |
| "and in luxembourg?"EN follow-up turn | 1 | 11.0 | 0 |
- 10,407
- impressions in 28 days — a growing share of them on AI surfaces that were never clickable
- 83% / 93%
- of searches end with no click at all, on AI Overviews and in AI Mode respectively
- 0.80%
- our own blended click-through rate — the number you get when you win the answer and lose the click
Position 1 has stopped being the finish line
Someone asked an assistant "and how much is it?" and Google logged this site at position 3 inside that conversation. There was never going to be a click — that is how the surface works. What still counts is being named, accurately, as the provider worth calling. That is the entire job: make the answer say your name, in your prospect's language, for the question they actually ask. We run this measurement on ourselves every month. We will run it on you.
Source: Google Search Console API, property 20more.lu, 10 July – 6 August 2026. Queries reproduced verbatim; long prompts truncated. Zero-click benchmarks: Seer Interactive, 2026.
Definitions
GEO, AEO, AI visibility — what the words actually mean
The category acquired three names in eighteen months and a great deal of noise with them. Here is the working definition we use, and the evidence behind it.
- Generative Engine Optimization (GEO)
- Getting a generative engine — ChatGPT, Gemini, Perplexity, Copilot — to reproduce your firm accurately when it writes an answer. The unit of success is a sentence inside the answer, not a position in a list of links. Gartner expects roughly 60% of commercial queries to be influenced by AI answers by the end of 2026.
- Answer Engine Optimization (AEO)
- The narrower craft of making a single, self-contained answer extractable from your page — for an AI Overview, a featured snippet, or a spoken reply. In practice the two terms are used interchangeably; we treat AEO as the on-page half of GEO and off-site entity work as the other half.
- Mentions are not citations
- Being named in an answer and being linked as its source are separate outcomes with separate levers. Semrush's 2026 AI Visibility Index — 126 million AI search prompts — found the overlap between mentioned brands and cited domains can fall as low as 30% on Gemini. We report mention share and citation share separately, always.
- Why this is fixable at all
- Yext analysed 6.8 million AI citations across ChatGPT, Gemini and Perplexity: about 86% came from sources a brand controls or influences — 44% first-party websites, 42% listings and directories, 8% reviews and social. The engines are not inventing your reputation. They are reading it off surfaces you can repair.
Also sold as: AI visibility, AI search optimization, LLM SEO, AI SEO. In this market the same service is bought as référencement IA in French and KI-Sichtbarkeit in German — one discipline, three vocabularies, which is exactly why we benchmark in all three.
- +120%
- more clicks per impression when a page is cited inside a Google AI Overview (Seer Interactive, 5.4M queries)
- 4.4×
- higher conversion rate measured for AI-referred visitors vs organic search (Semrush)
- 68%
- of US Google searches now end without any click (SparkToro, 2026)
Vendor research, correlational — we quote it with its caveats, and measure your case instead of promising you these numbers.
Trusted By
Per-engine playbook
Every engine reads different sources. We work per engine.
Winning Google Maps doesn't put you in ChatGPT. Each assistant grounds its answers in its own mix of indexes, directories and review platforms — so the work is source-by-source, engine-by-engine.
ChatGPT
01What it readsBing index, Bing Places, Foursquare, Yelp
What we optimizeBing Webmaster + Places, the profiles this engine actually reads, fresh answer-first pages
Perplexity
02What it readsOwn crawler plus Yelp and TripAdvisor partner data
What we optimizeYelp completeness and review velocity, citable third-party sources
Google AI Overviews / Gemini
03What it readsGoogle index, Business Profile, classic search signals
What we optimizeGBP completeness, content covering fan-out sub-questions, schema
Microsoft Copilot
04What it readsBing index, Bing Places
What we optimizeIndexNow, structured data — confirmed by Microsoft to help Copilot grounding
Offer & pricing
Clear scope, published prices
Start free. Every step ends in a decision, not a subscription trap.
Snapshot
Free
See how AI answers about you today
- —10–15 real prompts across 2 engines
- —French + English, Luxembourg context
- —1-page scorecard vs one competitor
- —robots.txt AI-crawler check
Diagnostic
€950
Know exactly where you stand
- —60–120 prompts × 4 engines × 3 runs
- —Mention share and citation share, reported separately
- —Entity, directory and review-footprint audit
- —90-day prioritized action plan
- —Fixed price — larger or multi-entity firms quoted separately
Foundation Sprint
€2,500–6,000
Fix the sources engines rely on
- —Entity/NAP cleanup across registries and directories
- —Bing Places, Google Business Profile, Yelp/Foursquare
- —Schema, AI-crawler fixes, IndexNow
- —3–5 answer-first pages in FR/DE/EN
Growth Retainer
from €1,500/mo
Stay in the answers as they shift
- —Monthly multi-engine benchmark with confidence bands
- —2–4 expert-reviewed pages per month
- —Mention and citation building
- —Quarterly entity re-audit
Authority tier — original research assets, digital PR, YouTube program, CSSF-compliant approval workflows — on request. 3-month pilot first; no long lock-ins.
Straight talk
What we won't sell you
The AI-visibility market is full of snake oil. Our differentiator is that we tell you what doesn't work.
- Guaranteed placements in AI answers — nobody controls the models, and anyone promising placement is guessing with your money.
- llms.txt as a magic lever — server logs show ~97% of these files are never fetched by AI crawlers. We deploy it as free hygiene, nothing more.
- Keyword-stuffed "AI content" — the Princeton GEO study measured that it actively hurts AI visibility.
- Fake reviews, review gating, or schema that says things your visible pages don't.
- Perfect attribution — we report share of voice, citations and AI-referred traffic as directional evidence with confidence bands, not fantasy dashboards.
Methodology
A methodology you can audit
Mentions ≠ citations
Being named in an answer and being linked as a source are different games — only a minority of top-mentioned brands are also top-cited. We track and report both, separately.
Confidence bands, not screenshots
Every prompt runs three times across different days, per engine. AI answers are volatile; a single-run screenshot is marketing, not measurement.
Proxy honesty
We benchmark through official APIs and hand-check a monthly sample against the real consumer apps — and disclose the divergence rate in every report.
Trilingual by default
Your prospects ask in French, cross-border staff in German, expat executives in English. The same firm gets different answers per language. We measure all three.
This is the same method behind our published AI Visibility Benchmark for Luxembourg fiduciary firms — 405 real answers across ChatGPT and Perplexity, with the method open for anyone to challenge.
How it works
How an engagement runs
- 1
Free snapshot
We run 10–15 real prompts and show you the actual answers. If you already look great, we'll tell you so and leave you alone.
- 2
Diagnostic
The full benchmark: where you appear, where competitors appear, and which public sources are driving their answers instead of yours.
- 3
Fix the sources
Entity cleanup, per-engine profiles, schema and answer-first content — the foundation sprint, with your approval on everything client-visible.
- 4
Measure monthly
Re-run the benchmark, track mention and citation share, adjust. Volatility is why this is a program, not a one-off project.
Want the background first? Read our guide to generative engine optimization (GEO) for Luxembourg businesses — and if you need broader help with AI, see how we work as your AI consultant in Luxembourg.
FAQ
Frequently asked questions
What is AEO / GEO / AI visibility?+
Answer Engine Optimization (AEO) — also called Generative Engine Optimization (GEO) — is the work of making sure AI assistants like ChatGPT, Perplexity and Google's AI Overviews describe your firm accurately and recommend it when relevant. It overlaps with SEO, but the levers differ: third-party brand mentions, per-engine directories like Bing Places and Yelp, entity consistency across registries, and answer-first content matter more than backlink volume.
Can you guarantee my firm appears in ChatGPT?+
No — and no honest agency can. AI answers are generated fresh and vary by engine, language and day. What we sell is measurement (how AI describes you today, with confidence bands) and source-readiness (fixing the public signals engines rely on). Both are in your control; the model's output is not.
How is this different from SEO?+
SEO gets you ranked among ten links; AI visibility gets you named inside the one answer people actually read. Large-scale studies show branded web mentions correlate strongly with AI visibility while backlink volume barely matters — and ChatGPT runs on Bing's ecosystem, not Google's. Classic SEO still gates entry on Google surfaces, so the two complement each other; we tell you which one your problem actually is.
Is this allowed for regulated professions — lawyers, CSSF-supervised firms?+
Yes, with discipline. For lawyers, content must stay objective and verifiable under Bar rules — exactly the kind of content AI engines prefer anyway. For CSSF-supervised firms, every published page flows through your compliance approval before going live; our workflow is built around that. We decline work for professions whose rules prohibit priority-placement services, such as doctors under the Collège médical charter.
What does it cost?+
The snapshot is free. A full diagnostic is a fixed €950 one-time (larger or multi-entity firms quoted separately), a foundation sprint €2,500–6,000 depending on scope, and the growth retainer starts at €1,500/month on a 3-month pilot. Below roughly €1,500/month, the industry's own consensus is that you are buying rebranded SEO reports.
How long before we see results?+
Source fixes — directories, entity data, schema, crawler access — land within weeks. AI engines refresh on their own cycles, so we benchmark monthly and evaluate a pilot over 90 days. You get a straight read-out either way: what moved, what didn't, and whether continuing is worth your money.
What does a GEO agency in Luxembourg actually do?+
Four things, in this order. First, measure: how does each engine answer the questions your buyers ask, in French, German and English — because Luxembourg buys in all three. Second, audit your entity footprint: whether the RCS registry, your Google Business Profile, Bing Places, LinkedIn and the local directories all agree on who you are, what you do and where you are. Third, fix the sources those answers are built from, including the on-page structure engines can actually extract. Fourth, re-measure monthly, because AI answers move week to week. Anything that stops at "we added an FAQ block to your homepage" is a rebadged SEO retainer.
How do I check today whether AI assistants mention my company?+
Open ChatGPT, Perplexity and Google AI Mode and ask what a buyer would ask — "best [your service] in Luxembourg" — in French and German as well as English, because the answers differ by language. Run each prompt three times on different days; the answers move. Record three things: whether you are named, whether your own site is cited as the source, and which competitors appear instead of you. That is a manual version of our free Snapshot, in which we run 10–15 prompts across two engines and hand you the raw answers.
Our market is small. Is AI visibility worth it?+
Often more, not less. In a market the size of Luxembourg the total monthly search volume for a professional category can be a few hundred queries — but the person asking an assistant "who should I use for X in Luxembourg" is a buyer, and the answer names two or three firms rather than ten. Being one of those names is a far larger share of a far smaller pool. If the prompt volume in your niche is genuinely too thin to justify a retainer, we will say so in the free snapshot rather than sell you one.
See how AI describes your firm — free
Book a free 30-minute review. We'll run the snapshot before the call and walk you through the actual answers — the findings are yours either way.