For many years, we and our client, a German-speaking keynote speaker, dominated the search results with our pages and also search term combinations with the search intention of a campaign – “book keynote speaker”, “book speaker in Munich” etc., continuous traffic, regular booking inquiries. The SEO work paid off. Until the summer of 2024.
Within three months, visibility collapsed completely. Not the rankings for generic keywords – these remained stable. But for action keywords. Google simply stopped showing our pages in the search results. Instead: speaker platforms such as 5 star speakers, Athenas, Speakers Excellence. Comparison portals. AI-generated overviews. And most shocking of all: our own website became a knowledge database.
The pattern: From Sixt to Speaker – comparison platforms take over
This phenomenon is not new. Sixt has already experienced this in the past. Not because of poor SEO. But because users want to compare. Check24, Billiger Mietwagen and Happycar dominate the results.
What began with car rentals is systematically continuing: In more and more service sectors, platforms and aggregators are taking visible positions in the search results – often at the expense of the actual service providers.
The logic behind it: People want choice. A single provider page – no matter how well optimized – does not offer this choice. So Google systematically favors platforms that present multiple options.
What initially made sense for transactional searches such as rental cars or hotels now also applies to knowledge-intensive services: Lawyers, consultants, coaches, speakers. The same development everywhere. Individual providers are disappearing, platforms are dominating.
The AI factor: search behavior is changing fundamentally
Parallel to platform dominance, something more fundamental is happening: AI is changing how people search. Not gradually – fundamentally.
Our analysis of 50,000 search queries in the speaker segment between 2022 and 2024 shows that
- Classic action keywords (“searched”, “book”, “find”): -73% search volume
- Problem-describing search queries: +340% search volume
- Question-based search queries: +420% search volume
- Average search query length: from 3.2 to 8.7 words
In concrete terms, this means
| 2022: Keyword search | 2024: Problem description | Change |
|---|---|---|
| “Keynote Speaker Munich” | “Speaker for AI transformation in SMEs – who has practical experience” | From 3 to 13 words |
| “Book speaker digitalization” | “How do I digitalize my company without losing employees – need speaker for executive event” | From 3 to 17 words |
| “Motivation speaker sales” | “Sales team loses motivation in a difficult market environment – which approaches work” | From 2 to 13 words |
The reason is obvious: ChatGPT, Perplexity, Claude and other AI tools have taught millions of people to search differently. Not in keywords – in whole sentences. Not for providers – for solutions. Not generic – specific.
Many users formulate search queries much more conversationally today – they describe their problem as if they were talking to a person. This is a massive change for classic keyword SEO.
The brutal diagnosis: Our website was an SEO construct
When we carried out the analysis for our speaker client, the result was sobering. His website was perfectly optimized – for 2019, each page a keyword cluster:
- “Keynote Speaker Munich” – with all local variations
- “Speaker Digital Transformation” – with an industry focus
- “Motivation speaker sales Germany” – geographically optimized
Technically perfect. Rankings stable. But the content? Generic. Interchangeable. No real answers, no specific solutions, no demonstrable expertise. Only claims: “I am an expert in digitalization”, “My keynotes inspire”, “Individual presentations for your event”.
Google’s assessment was clear: knowledge database. The website was indexed, crawled, saved as a source – but no longer displayed. This was because it did not answer any of the new, specific questions that users were actually asking.
This was confirmed in the Search Console data: Impressions stable, clicks -82%. The pages were there – but nobody saw them anymore.
The comparison platform problem: We are fighting against the wrong competition
To make matters worse, the speaker platforms had long understood the change. Their strategy was simple and brutally effective:
5 star speakers: Lists 20-30 speakers with short profiles for each search query. Google loves this structure – users get choice, the platform ranks.
Athenas: Positions itself as “consulting” – helps event organizers with speaker selection. Offers exactly what Google considers “helpful content”.
Speakers Excellence: Combines speaker lists with editorial articles on event topics. Double added value from Google’s point of view.
It can be observed in many service sectors: platforms are gaining a disproportionate amount of organic presence – while individual providers are more likely to lose visibility, even though their technical foundations are correct.
The hard truth: fighting platforms as individual providers is futile. We cannot win this battle. But we can play a different game.
The solution: content for AI systems and humans
After months of analysis and testing, we developed a new strategy. Not “more SEO” – but fundamentally different content.
Step 1: From keywords to specific problems
We eliminated all “wanted/booking” pages. Completely. Instead, we analyzed 200 past booking conversations: What specific problems did the organizers have? What questions did they keep asking?
This resulted in new content:
Old: “Speaker Digital Transformation – Keynotes on digitalization and innovation”
New: “AI implementation in SMEs: Why 83% of projects fail – analysis of 200 failed transformations”
Old: “Motivational speaker sales – successful speeches for sales teams”
New: “Sales motivation is dead: what really drives high performers (spoiler: no motivational speeches)”
The difference: each new article answered a specific question. With concrete examples. With a clear stance. With demonstrable expertise.
Step 2: Optimize content for AI search
ChatGPT, Perplexity, and Google’s AI Overviews in search – these systems look for specific content patterns:
Structure with clear answers: AI systems prefer content that answers questions directly. Not “could”, “possibly”, “under certain circumstances” – but clear statements with reasons.
Expertise signals: concrete figures, studies, our own data. We integrate results from 500+ keynotes, anonymized client cases and our own surveys.
References: Every central statement with a reference. Content with traceable references is easier to verify and cite.
Comprehensible language: no buzzwords, no marketing speak. Concrete, comprehensible argumentation.
One example: The article “Why 83% of all AI projects in SMEs fail” contained:
- Own analysis of 200 transformation projects (concrete database)
- 12 specific sources of error with examples (clear structure)
- References to scientific studies (expertise signals)
- 3 detailed case studies of failed projects (practical relevance)
- Specific recommendations for action for each phase (added value)
The result: In tests with recurring prompt sets, the article was regularly used as a source in AI response systems – significantly more often than classic “book/search” landing pages.
Step 3: Specialization as a strategic focus
The hardest step: Complete refocusing. From “Speaker for Innovation, Digital Transformation, Motivation” to “Expert for AI-driven business model transformation in B2B SMEs”.
This specialization had three consequences:
1. dramatically reduced search market: the target group became significantly smaller and more specific. Instead of a broad mass, we now reached a highly specialized niche.
2. significantly higher relevance: Every inquiry came from exactly the right target group. No more “looking for speakers for our team celebration” requests. Only qualified event planners with specific needs.
3. Completely changed fee structure: Generic digitalization keynotes: €3,500-5,000. Specialized AI transformation keynotes: €12,000-18,000.
The results: 8 months later
The transformation took 8 months. From summer 2024 to spring 2025, we reworked the complete content strategy. The numbers:
Traffic & Visibility:
- Organic traffic: -42% (from 8,200 to 4,750 visitors/month)
- Keyword rankings top 10: -38% (from 340 to 210 keywords)
- Visibility index (Sistrix): -35%
Classic SEO metrics: Catastrophic. But then:
Quality & conversion:
- Booking inquiries: +135% (from 18 to 42 per quarter)
- Qualified inquiries (target group fits): +290%
- Requests via “I have read your article”: from 12% to 78%
- Average fee: +110% (from €4,200 to €8,800)
- Booking rate: from 22% to 58
- Follow-up bookings: from 15% to 41
AI presence:
- Mentions in AI answers: regularly for relevant queries (internal tests)
- References in AI tools: significantly more often than with classic landing pages (internal tests)
- Display in Google AI Overviews: observed in individual cases for specific questions (internal observation)
The brutal truth: less traffic does not mean less success. It means the right traffic.
What Google has really changed: The three shifts
Our analysis of over 200 websites in different service areas shows three fundamental shifts in Google’s algorithm:
Shift 1: Selection via individual providers
Google systematically prioritizes content that offers users choices. Comparison and overview formats are more likely to match a search intention that aims at orientation and selection – while individual provider pages often do not meet this expectation.
Examples from various industries:
- Rental cars: Check24, Cheap Rental Cars dominate
- Speakers: Platforms like 5-star speakers dominate
- Lawyers: anwalt.de, advocado dominate
- Craftsmen: MyHammer, Check24 dominate
The consequence: fighting against platforms as individual providers is futile. We need a different strategy.
Shift 2: Problem solution via provider search
Google has repeatedly emphasized in recent years that content created primarily for search engines instead of people will perform worse in the long term – and that helpful, problem-solving content is preferred.
An analysis of the Google Search Quality Rater Guidelines shows explicit evaluation criteria for “helpful content”. The most important:
- Does the content fully answer a specific question?
- Does the content offer expertise that goes beyond generally available knowledge?
- Would a person rate this content as helpful for their problem?
- Is the content created primarily for people or for search engines?
Classic “Keynote Speaker Munich wanted” sites fail all four criteria.
Shift 3: Expertise signals via location signals
Perhaps the most drastic change: For knowledge-intensive services, proximity and location signals are often less decisive than demonstrable expertise. Users are no longer just looking for “in my city”, but for “fits my specific problem”.
Google’s algorithm reacts: Expertise signals (specialist articles, publications, citations, content quality) are often weighted more heavily in the ranking calculation than pure location signals (NAP, GBP, local backlinks) – especially if the service does not necessarily have to be provided locally.
Practical implementation: your roadmap
Based on our experience with over 40 service providers in transformation:
Phase 1: Analysis (week 1-2)
- Identify traffic sources: Where are your current inquiries really coming from? Not traffic – inquiries. Analyze the last 50 booking calls: Which pages did the interested parties see?
- Check keyword reality: Which keywords do you still rank for, but are no longer displayed? Search Console shows: Impressions vs. clicks. A ratio worse than 1:20 is a warning signal.
- Evaluate platform competition: Which platforms dominate your top keywords? If 7 out of 10 top positions belong to platforms, you have a problem.
Phase 2: Problem catalog (week 3-4)
- Analyze customer conversations: Go through your last 100 initial conversations. Which 20 questions/problems come up again and again?
- Simulate AI search: Ask ChatGPT, Perplexity, Claude about your topic area. What questions do these systems ask? What content do they recommend? Does your website come up?
- Define specialization: What can you demonstrably do better than 95% of your competitors? Don’t “believe” – prove it. With data, cases, results.
Phase 3: Content transformation (month 2-4)
- Restructure old pages: Systematically rebuild outdated “book/search” content into information pages that provide real booking information.
- Create problem content: For each of the 20 identified problems: A well-founded article, minimum 2,000 words, with concrete answers, sources, practical examples.
- Demonstrate expertise: Integrate your own data. Client cases (anonymized). Studies. Figures. Proof of your expertise.
- AI optimization: Clear structure, direct answers, references, understandable language. Test every article: Is it picked up as a source in AI answers?
Phase 4: Positioning (month 5-8)
- Communicate specialization: Everywhere. Website header, LinkedIn profile, email signature, first meetings. No longer “I do everything” – but “I am THE expert for X”.
- Remove local limits: If your service does not necessarily have to be local – why limit yourself geographically?
- Adjust fees: Specialists can (and must) charge more. Recalculate.
- Sharpen qualifications: No longer accept every job. Only those where your specialization is really in demand.
The most common mistakes (and how to avoid them)
Mistake 1: “I need to do more SEO”
No. More of the wrong thing doesn’t help. We saw a speaker who created 40 new “keyword-optimized” pages. Result: zero additional visibility. The problem is not quantity – it’s the wrong approach.
Solution: Fewer pages, but better content. A well-founded 3,000-word article on a specific problem is more valuable than 10 generic keyword pages.
Mistake 2: “I lose customers when I specialize”
The biggest fear of all service providers. Our analysis of 40 transformation cases shows: Most service providers had more qualified inquiries after specialization – not less.
Solution: You lose the wrong customers (who you didn’t want anyway) and gain the right ones (who are willing to pay for expertise).
Mistake 3: “Locality is my advantage”
It was. Until 2022. Today it is a limitation. A speaker from Munich no longer competes with speakers from Munich – but with all German-speaking speakers on his or her topic.
Solution: Rethink your market area. Digital, nationwide, international. Expertise knows no city limits.
Error 4: “AI search is not yet relevant”
AI features in Google Search (AI Overviews) have been heavily rolled out in recent years: since October 2024 in 100+ countries, since May 2025 in 200+ countries and territories (Google). At the same time, AI tools such as ChatGPT have achieved widespread use.
Solution: Optimize your content for AI systems NOW. In 12 months it will be too late.
What this means for different sectors
Lawyers:
Instead of “Fachanwalt Arbeitsrecht München” → “Kündigung erhalten: Your rights in the first 48 hours (checklist for employees)”
Consultant:
Instead of “Management consultancy Hamburg” → “How medium-sized production companies can make their supply chains crisis-proof (10 specific measures)”
Coaches:
Instead of “Business Coach Frankfurt” → “Why 90% of all scale-ups fail because of the management team (not the product)”
Tax consultant:
Instead of “Steuerberater Berlin” → “Kryptowährungen versteuern 2025: Der komplette Leitfaden (inkl. Beispielrechnungen)”
The pattern is always the same: from generic provider to specialized problem solver.
Time frame: Realistic expectations
This transformation is not a quick fix. Our experience shows:
- Months 1-3: Traffic decreases. Inquiries drop. It feels wrong. This is normal.
- Months 4-6: First qualified inquiries about new content. Not many yet, but significantly better qualified.
- Months 7-9: Tipping point. New inquiries exceed old inquiries – with better quality.
- Months 10-12: Stabilization at a new level. Less traffic, fewer inquiries, but significantly higher conversion and fees.
Allow 6-12 months for measurable success. If you don’t have this patience, you shouldn’t start.
The inconvenient truth
Classic keyword SEO for service providers no longer works.
The numbers are brutally clear: “search/book/find” keywords have noticeably lost search volume in many markets since 2022. Google often plays these pages less – even with perfect rankings. AI systems usually ignore them completely.
If you don’t change now, you will lose. Not someday – already now. Visibility is falling month by month, while platforms are taking over search results and AI systems are redefining the rules of the game.
The good news: the alternative works. Problem-focused content, clear specialization, proof of expertise – this approach generates measurably better results. Not more traffic – better traffic. Not more inquiries – better inquiries.
The question is no longer whether you will switch. It’s how quickly you can do it before your competitors do.
“The website is no longer a business card. It is your demonstration of expertise. And if you have no expertise to demonstrate – then you have a problem that no SEO in the world can solve.”
List of sources
- Google (2024). “AI Overviews in Search are coming to more places around the world.” Google Blog (Oct 28, 2024).
- Google (2025). “AI Overviews expand to over 200 countries and territories.” Google Blog (May 20, 2025).
- Google (continuously updated). “Search Quality Evaluator Guidelines.” Raterhub PDF.
- Google (2024). “March 2024 core update & spam updates” (Search Central Blog / Developer Blog).
- Reuters (2024). “OpenAI says ChatGPT’s weekly users have grown to 200 million” (Aug 29, 2024).
- Own analysis (2022–2024). Evaluation of 50,000 search queries in the speaker segment; plus evaluation of booking discussions and content tests.
Note: The case study presented is based on real project experience, but is abstracted and merged. Industries, numbers and time periods have been adjusted to exclude conclusions about individual customers.