of B2B buyers use ChatGPT or Copilot for product research and purchase preparation.
Emporix 2026Exclusive data report · July 2026
The AI Visibility Report 2026
ChatGPT is recommending your competitor right now.
- 3,971 AI answers
- 4 AI systems
- 16 providers
- 30 days
94 % of B2B buyers research with ChatGPT and Copilot. 93 % of mid-sized DACH companies do not appear in those answers. We measured for 30 days who gets recommended in the DACH region, and why. No borrowed studies: our own primary data from daily AI monitoring.
- 56 pages
- Free PDF
- German language
- 10 cited studies
- Straight to your inbox
56 pages · PDFFree
Data report · 2026
3.971ausgewertete KI-Antworten
KI-Sichtbarkeits-ReportWer im DACH-Raum von der KI empfohlen wird
lilox.io90-day plan
3,971AI answers
51buyer questions
4AI systems
16providers
30days measured
10cited studies
LILOX AI visibility monitoring via Peec AI · 02.06. to 02.07.2026 · provider names anonymized
/ 01 · Why this report exists
The decision is made before your sales team even hears about the lead
The B2B buying process has changed more in 24 months than in the 15 years before. You will not see it in the trade fair calendar, you will see it in your buyers’ browsers. They ask ChatGPT or Copilot for a market overview, get three to five providers named, and contact exactly those.
chatgpt.com
“Which providers for [your category] are there in the DACH region?”
Answer
For this use case, three providers are primarily named in the DACH region. All three publish structured specifications and are confirmed by several sources:
- Competitor ANamed
- Competitor BNamed
- Competitor CNamed
- Your companyNot named
Unlike Google, there is no position 4 that still gets clicks. The AI answer is binary: either you are on the shortlist, or you do not exist for that buying process. We wanted to know what that looks like in the DACH region, so we measured it instead of guessing.
The core thesis of the report
Between “buyers ask the AI” and “providers are visible in the AI” lies the biggest sales gap since Google search was introduced.
of mid-sized DACH companies are visible in those answers at all. The channel is practically unoccupied.
GEO study DACH 2026 via Big Data Insider/ 02 · Executive summary
The seven figures you need to know
If you read nothing else in the report, read this page. Seven figures that explain why AI visibility is a sales topic and not a marketing topic.
of B2B buyers use tools like ChatGPT or Microsoft Copilot for product research and purchase preparation.
Emporix 2026now start their research more often with an AI chatbot than with Google.
G2, 03/2026of mid-sized B2B companies in the DACH region stay invisible in AI search, despite good Google rankings.
GEO study DACH 2026visibility is what the best of 16 measured providers reaches. Even the market leader is missing from 85 % of answers.
LILOX monitoringhigher conversion for website visitors from AI search compared to classic Google traffic.
Emporix 2026organic clicks for queries where Google AI Overviews appear. Classic SEO is losing reach.
Omnibound 2026of B2B buyers prefer a buying process entirely without sales contact. The decision is made digitally.
Gartner 03/2026growth in AI referral traffic year over year. The new channel is growing while the old one shrinks.
Omnibound 2026/ 03 · Your outcome
What you actually hold in your hands after reading
Not a trend paper. Six outcomes you can take into the management meeting the same day.
Clarity on where you stand today
A self-test with ten prompts and a scoring scale. In ten minutes you know whether the AI knows you, ignores you or describes you incorrectly.
Real primary data instead of trend claims
Six findings from 3,971 analyzed AI answers. Anonymized, documented so you can follow it, with the methodology disclosed.
The nine most expensive mistakes
Sorted by cost, distilled from audits and monitoring data. You can check and fix mistake 1 this week.
Technical building blocks to copy
Ready-made JSON-LD and llms.txt snippets. Your web team implements them in a few days, without a consulting project first.
A 90-day roadmap in four phases
Measure, foundation, third-party sources, prove. With five KPIs to measure progress against instead of gut feeling.
Arguments for the management team
A calculation showing what invisibility actually costs. Plus answers to the three objections that will come up in the meeting.
Six findings your competitor hopefully does not know yet.
56 pages, free, straight to your inbox. All we need is a name and an email.
/ 04 · Sample from chapter 5
Six findings from 3,971 AI answers
For 30 days we asked 51 real buyer questions daily to four AI systems and tracked 16 providers. Three findings are here. The other three, including all analyses, are in the report.
FINDING 0114.6 %
Even the market leader is missing from 85 % of answers
On Google, the top 3 occupy a query almost completely. In AI search the best provider only reaches 14.6 %. The channel belongs to nobody yet, and dominance will be distributed over the next 12 to 24 months.
FINDING 041.67
Structure beats brand size
A small, cleanly structured site achieved the highest citation rate in the entire measurement field. Directories, Reddit, YouTube and LinkedIn were among the most cited sources. Optimizing only your own website leaves half the value on the table.
FINDING 060 %
The average lies
39 % visibility on the strongest question, 0 % on 36 of 52 questions, including the purchase-relevant ones. The 4 % average hides the imbalance completely. You steer via the 10 to 15 most revenue-relevant questions.
All six findings are in the report in full.
/ 05 · The visibility paradox
“We rank well on Google, so people find us”
This is the most dangerous misjudgement of the year. AI systems read websites by different rules. SEO remains the foundation, but it now covers only half the distance.
Google search
What you have optimized for so far
AI search · in the report
What matters from now on
✕Domain authority and backlinks decide
✓The passage with the most precisely extractable answer wins
✕Ten blue links, position 5 still gets clicks
✓One answer with three to five providers, in or out
✕Short keywords as search logic
✓Whole questions with context and specification
✕Third-party sources work indirectly via backlinks
✓Directories, Reddit and YouTube feed directly into the answer
✕Substance may sit in PDFs and configurators
✓Facts in the HTML, machine-readable via JSON-LD and llms.txt
/ 06 · The self-test
The self-test: your own picture in ten minutes
Ten questions to four AI systems, and then you know where you stand. Three prompts are here, all ten plus the scoring scheme are in the report.
The prompts
Three out of ten
- Which providers for [category] are there in the DACH region?
- Recommend a manufacturer for [product] with [specification].
- Compare the leading providers for [category].
- Prompts four to ten are in the report.
Test in all four systems and on several days. Visibility varies by over 60 % between systems and by ±30 % from week to week.
The scoring
How often are you named?
- 8 to 10×: Excellent. You are among the few percent. Now defend the position against newcomers.
- 4 to 7×: Solid basis, but the AI recommends you unreliably. You are probably missing from the most important buying prompts.
- 1 to 3×: Critical. The AI barely knows you or says the wrong things. Both actively cost you inquiries.
- 0×: You are part of the 93 %. Your prospects are shown your competitors only.
/ 07 · The 90-day roadmap
From invisible to measurably visible in 90 days
Four phases with a clear result per stage. Designed for fast retrieval wins, not for a year of lead time.
DAYS 1 TO 14
Measure
Establish a baseline across four AI systems, track competitors, define the 10 to 15 revenue-relevant prompts. After that you know where you stand and who is taking inquiries from you.
DAYS 15 TO 45
Foundation
Open robots.txt for AI bots, answer-first structure on the five most important pages, FAQs in prompt wording, JSON-LD, llms.txt and the core content out of the PDFs and into the HTML.
DAYS 46 TO 75
Third-party sources
Maintain relevant directories completely, publish specialist articles with citable figures, build a presence in professional communities and on LinkedIn.
DAYS 76 TO 90
Prove
Measure again, document the delta, scale the two most effective levers. That is how you check whether the AI now describes you the way you want to be described.
The model knowledge of 2027 is being formed right now.
Starting at 20 % AI traffic means competing against two years of citation history. As impossible to catch up on as missed SEO in 2010.
/ 08 · What is inside
Twelve chapters from the situation to the roadmap
56 pages, in German, with anonymized primary data from the AI monitoring and ten numbered sources in the appendix.
- 017 figures
The seven figures you need to know
The situation on two pages, every figure with its source.
- 02Buying process
The quiet revolution in B2B buying
Buyer journey, buying center and why DACH is shifting faster than expected.
- 032 mechanisms
How AI systems form recommendations
Model knowledge and live retrieval as the two mechanisms you can deliberately influence.
- 04SEO vs. GEO
The visibility paradox
Good Google ranking and still invisible. Four differences in a system comparison.
- 056 findings
Live data: 3,971 AI answers in detail
Methodology, prompt set and six findings from 30 days of daily measurement.
- 065.1×
What AI traffic is worth
Conversion, reasoning and a sample calculation for a technical B2B company.
- 07Industry
Industry focus: technical B2B
Three amplifiers, your structural advantage and the path from PDF archive to AI answer source.
- 086 levers
The anatomy of an AI-recommended website
Six levers, a before-and-after example and ready-made JSON-LD and llms.txt snippets.
- 099 mistakes
The nine most expensive mistakes
Sorted by cost, from locked-out bots to the most expensive mistake of all: waiting.
- 1010 prompts
Self-test: how visible are you?
Ten prompts, four systems, a scoring scheme with four levels.
- 115 KPIs
The 90-day roadmap
Four phases with a result per stage, plus the five KPIs of AI visibility.
- 127 objections
FAQ for managing directors
The seven objections that come up in the meeting, with evidenced answers.
- +Appendix
Appendix: glossary, sources and methodology
Eleven terms explained concisely, ten numbered studies and the fully disclosed methodology.
/ 09 · Is this for you?
Who this report was written for
Honestly: not for everyone. So you can skip the download if it will not help you.
Good fit
The report is for you if …
- you run or are responsible for a technical B2B company in the DACH region
- your buying cycles are long and your order values high
- you have good Google rankings and still feel you are asked less often
- your substance sits in datasheets, PDFs and configurators
- you need to show in the next meeting why AI visibility is a management matter
- you want to know what invisibility actually costs you
Not a fit
Skip the download if …
- you run transactional B2C volume business
- you already run AI monitoring across several systems
- you are looking for a finished agency quote rather than a toolkit
- you want to read general AI trends rather than measurement data
/ 10 · Who did the measuring
Two founders from marketing and technology
LILOX continuously monitors the AI visibility of its own clients, with exactly the monitoring whose data is in this report. The infrastructure was tested on our own brand first.
Nadine Fraiss · Marketing
15 years in digital marketing, several of them at CMO level.
Between traditional Mittelstand, startups and scale-ups. She has sat on both sides: as the client who had to justify agency budgets, and as the person expected to deliver pipeline at the end of the quarter.
Christoph Fraiss · Technology
Engineer, structural engineer on major construction sites, managing director of a manufacturing company.
He knows the moment when an offer fails on price. And the moment when it fails on a missing explanation. For you, he translates between technology and market.
/ 11 · Customer voices
What clients say about working with LILOX
Approved quotes from ongoing and completed engagements. More will follow as each approval comes in.
5 out of 5
A thorough analysis, an overview of current tools, and effective actions to start with. Exactly what you need in order to become visible digitally.
ManfredHead of Sales, Cloud ERP5 out of 5
Advice between equals, no fluff, always focused.
Prof. (FH) Dr. Uwe HeilEntrepreneur, Dr. Heil KG Consulting & Coaching for Executives5 out of 5
Nadine and her team were a stroke of luck for us. They helped us position our brand and product portfolio without forcing us to completely rethink our original vision or resources — they polished our rough diamond, so to speak. Thanks to a lead-generation strategy that actually fit our budget, we've been able to market our business model to exactly the right audience. A clear recommendation for SMEs like us.
Sabrina WinischhoferCo-Founder, Vienna Etiquette Society5 out of 5
For my company, AnimatedReal GmbH, I was looking for a modern, contemporary, and engaging website. Nadine took the time to really understand what I wanted. I'm extremely happy with the result — the new site hits exactly the tone and look I had in mind. What I valued most was Nadine's reliability, her openness to feedback, and her ability to perfectly balance design and functionality. I can wholeheartedly recommend Nadine and the Lilox team!
Fabian OlbrechtManaging Director, AnimatedReal5 out of 5
As a complete newcomer to marketing (especially LinkedIn/Google Ads), working with Nadine Fraiss and Lilox has been not only highly professional but also incredibly personable and empathetic. I'd wager Nadine can help marketing pros just as well — marketing beginners are certainly in extremely good hands. 100% recommended!
Alex SchubertManaging Director, netwiss OG5 out of 5
A top-notch execution. So glad I got to know Nadine. Very professional and friendly!
Svetlana GradwohlManaging Director, Gradwohl Creations/ 12 · Frequently asked questions
The objections that will come up in the meeting
“AI traffic is under 2 % for us.”
Those 2 % convert 5 to 8 times better. Measured in inquiries, that is 10 to 15 % of your pipeline effect. Add +357 % growth per year and the inertia of model knowledge: whoever wants to be visible in 2027 publishes in 2026.
“Was our SEO investment wasted then?”
No. Good SEO is the foundation, because AI Overviews and Copilot draw on the same indexes. On its own it is not enough though: 93 % are AI-invisible despite Google rankings. GEO builds on SEO.
“Can we not simply buy our way in?”
No. There are no paid placements in organic AI answers. So far the channel is not decided by budgets but by substance. That is exactly the opportunity for mid-sized companies.
“How quickly do we see results?”
Retrieval effects after four to eight weeks, clear gains on purchase-relevant prompts after three to six months, anchoring in model knowledge in six to 18 months. The 90-day plan in the report targets the fast effects.
“Is a quarterly test in-house enough?”
As a start yes, for steering no. ±30 % weekly variation, four systems, dozens of prompts and the competitive comparison cannot be done manually. Monitoring relates to a spot check like a CRM relates to a spreadsheet.
“What does the report cost and what happens to my data?”
The report is free. Processing under GDPR, servers in the EU, no sharing with third parties. No call without your invitation. You can unsubscribe from the newsletter at any time with one click.
Download
Get the report as a PDF
56 pages, in German, with anonymized primary data from the AI monitoring and ten numbered sources in the appendix.
What is in it
Twelve chapters, a self-test and a roadmap. 56 pages in total.
- All six findings from 3,971 AI answers
- The nine most expensive mistakes, sorted by cost
- Self-test with ten prompts and a scoring scheme
- JSON-LD and llms.txt snippets to copy
- 90-day roadmap with five KPIs
- Glossary, methodology and ten numbered sources
/ 13 · Data basis
What this report is based on
- 01 · Emporix, “The Autonomous Commerce Shift 2026”: 94 % AI usage, 5.1× conversion
- 02 · G2 study 03/2026: 51 % start their research with an AI chatbot
- 03 · GEO study DACH 2026 via Big Data Insider: 93 % AI-invisible
- 04 · AI Overview click analyses 2025/26 via Omnibound: −58 % organic clicks
- 05 · Gartner Sales Survey 03/2026: 67 % prefer buying without sales contact
- 06 · Gartner Survey 05/2026, 645 buyers: 69 % validate, 7 information sources on average
- 07 · Omnibound AI Search Statistics and SearchSignal Benchmark 2026: +357 % AI traffic
- 08 · Similarweb Gen-AI stats 2026: ChatGPT with over 2.5 billion queries per day
- P · Primary data: LILOX AI visibility monitoring via Peec AI, 51 prompts, 4 systems, 3,971 answers, 16 providers, 02.06. to 02.07.2026
The complete, numbered source list and the methodology are in the appendix of the report.
/ 14 · Next step
You do not just want to read, you want to know your own score?
In the free AI visibility check we measure your visibility and that of three competitors across four AI systems. You get the findings as a document and we go through them together.
