The future of search: What insurers need to do now
Websites are becoming more important, not less important
At the last AMC Espresso, we discussed a topic that is currently shaking up the entire industry: zero-click searches and the future of digital visibility. Fabian Kirchberger, Product Owner Web Analytics at Allianz, and Jonas Strübig, Data Analytics Consultant at webalyse, were there.
Zero-click searches are not a marginal phenomenon. They are now at the heart of digital search behavior. Over 65% of global Google searches end without a click, and on mobile devices, the figure is even higher at over 75%1. This means that users expect answers directly on the search results page without having to fight their way through websites. The click is no longer the goal! It's about appearing in the relevant answers.
In this article, we take a look at the most important trends in zero-click searches and the crucial steps insurance companies need to take to remain relevant.
And for anyone who's interested, here's a short interview in which Sven Köhler and I revisit the AMC Espresso on the topic of "The Future of Search".
Changing search behavior
The real paradigm shift is evident among younger generations. Whereas previously the central impulse was to enter a keyword and sift through search results for information, the use of AI tools such as ChatGPT or similar AI tools now dominates2. They ask specific questions and expect a precise, immediate answer – from search engine to answer engine3 4.
The trust issue in the insurance context
Insurance is incredibly intimate – it's about basic protection for health, family, and home. The level of trust is crucial here.
The danger: if the younger generation communicates with AI assistants on a daily basis and receives reliable answers from them over a long period of time, their trust shifts away from the unknown insurance agent and toward the digital companion that they use every day anyway.
The three pillars of modern search behavior
Before we dive into specific strategies, it's important to understand the key terms and their differences:
AIO (AI Overviews)
AI Overviews are AI-generated summaries that appear directly in Google search results, among other places. They aim to provide an instant synthesis of relevant content from the web. Unlike featured snippets, which extract one or two sentences from a single source, AI Overviews attempt to consolidate knowledge from multiple sources. The data shows a clear trend: AIOs now appear in 30% of all US desktop keywords. Similar figures are expected for Europe.
Zero-click searches
Zero-click searches are search queries where users get their answers directly on the search results page without having to visit a website. In 2024, around 65% of global Google searches did not result in a click; on mobile devices, this figure was even higher at over 75%5.
Long-tail search vs. short-tail search
Long-tail search refers to longer, more specific search queries (usually 3+ words) that often have less search volume but reveal a stronger, clearer user intent and thus have higher conversion rates. These detailed search queries are particularly relevant for AI Overviews: The data shows that AI Overviews mainly appear for long-tail, informational search queries with low difficulty—not for high-volume, transactional short-tail keywords6.
In contrast, short-tail searches are short, general terms (1-2 words) with high search volume that tend to have commercial or transactional intent and currently trigger AI overviews less frequently.
Understanding the impact
The paradox of zero-click development in long-tail searches:
It is interesting to note that AI Overviews generally have higher zero-click rates. A detailed analysis of the same long-tail keywords before and after the introduction of AI overviews showed that the zero-click rate for long-tail searches fell slightly (from 38.1% to 36.2%)7. This suggests that AI overviews do not automatically lead to fewer clicks for long-tail searches.
Long-tail optimization becomes crucial
88.1% of search queries that trigger an AI Overview are informational. This is precisely the area where long-tail keywords dominate. This shift makes optimization for long-tail searches more important than ever, as they represent the main area where AI Overviews are currently active.
What insurers should do NOW
Here are the specific measures. Some of them should have been implemented yesterday:
1. Implement WCAG
The European Accessibility Act requires the implementation of WCAG 2.1 Level AA standards from 2025 onwards8. WCAG is essential and should be implemented now at the latest. It not only makes your website accessible, but also more machine-readable – a clear advantage9.
2. Structured data is a must
Structured data is not becoming less important; it is becoming more important than ever. According to Google, structured data plays a central role in search engine visibility because it helps crawlers better understand the content of a website.10 11 In our DX projects, we have long relied on the consistent use of structured and semantic web.
Tip for implementation:
Implement JSON-LD markup for FAQs, product reviews, or article overviews.
Example FAQ markup:
{
"@context": "https://schema.org",
"@type": "FAQPage",
"mainEntity": [
{
"@type": "Question",
"name": "Wie finde ich eine ideale Versicherung?",
"acceptedAnswer": {
"@type": "Answer",
"text": "Vergleiche verschiedene Anbieter, um ein gutes Preis-Leistungs-Verhältnis zu finden. Achte auf Bewertungen und spezifische Leistungen."
}
}
]
3. Update meta tags
Optimize meta titles and meta descriptions by directly integrating current keywords. An appealing description increases the click-through rate and helps the algorithm understand that the page is relevant.
Tip for implementation:
Example of an optimized meta title:
“Top health insurance plans for self-employed individuals | Expert guide 2025”
Meta description:
“Find the best health insurance plans for self-employed individuals in 2025. With reviews, price-performance comparisons, and tips for choosing your insurance.”
4. Less lead affinity
Excessive CTAs can impair the user experience and distract from informative content. For maximum visibility in AI responses, the focus should be on fact-rich, quotable content12.
5. Improved article structure
A clear outline helps both readers and search engines to quickly understand and use the content.
Tip for implementation:
Use clear headings (H1, H2, H3), short paragraphs, and visual elements such as lists or infographics.
Example outline for a blog article:
H1: How do I find the best insurance?
H2: What are the most important criteria?
H3: Price-performance ratio
H3: Customer recommendations
H2: Step-by-step guide to choosing
H2: Conclusion and helpful tips
6. Backlinks are back
Being linked to is becoming important again. Advertorials, backlinks, and networked content are important for increasing your visibility and trust index13 14.
7. Maintain trust signals from third-party sources
Maintaining Google reviews is crucial. Negative reviews without comment significantly weaken your brand's trust index. Studies have shown that interacting with reviews significantly influences brand perception and trust. 15 16
8. Include long-tail keywords
Use long-tail keywords, as these often have a more specific search intent and therefore offer more conversion potential. Optimize your content specifically for long-tail searches, as these appear particularly in AI overviews.
Tip for implementation:
Identify questions or phrases that your target audience is actually searching for (e.g., “How do I find the best health insurance for self-employed people?”).
Incorporate these keywords into headings, the first paragraph, and longer subheadings.
Model Context Protocol (MCP) – The Technical Revolution
The Model Context Protocol (MCP) is a new open standard introduced by Anthropic in November 2024 and already adopted by leading companies such as OpenAI, Microsoft, and Google. MCP solves a key problem: Until now, AI models worked in isolation and couldn't access external data sources or enterprise systems in real time18.
What makes MCP so revolutionary?
Instead of programming separate connections for each data source, MCP creates a unified “language” that all AI systems can speak. The protocol enables AI agents to securely access databases, APIs, file systems, and other external resources—in real time.
Architecture overview:
MCP Client: Manages connections and translates between different systems
MCP Server: Provides specific functions and data sources
Uniform communication: Uses standardized JSON-RPC 2.0 messages for secure data transmission
Specific application examples:
Companies can now use AI agents that automatically update tickets in incident response platforms, query live data from SIEM systems, or interact with customer relationship management systems—all via a single, standardized interface.
The connection to digital experience:
At eggs unimedia, we are already seeing how MCP simplifies integration between Adobe Experience Cloud solutions and AI-powered systems in our DX projects. When it comes to optimizing customer journeys in particular, MCP enables a seamless connection between analytics data, content management systems, and automated personalization tools—without the complex individual integrations for each data channel that were previously necessary.
At the same time, MCP has revolutionized development work at eggs: Tools such as Amazon Developer Q19 can access project-specific contexts directly through MCP integration, enabling them to generate customized code for Adobe Experience Manager. Instead of isolated code suggestions, this results in development solutions that take into account the entire technical stack and specific DX requirements.
Things get particularly exciting in the content supply chain: MCP enables AI agents to automatically orchestrate between content creation, asset management, translation services, and publication systems. A concrete example from our practice at eggs unimedia: We have developed an MCP-supported workflow in which an AI agent automatically analyzes and gradually optimizes an incoming campaign briefing. The agent checks the briefing against our defined brand standards, compliance guidelines, and Adobe Experience Cloud requirements. If the briefing does not yet meet the standards, the agent iterates independently—it adds missing information, suggests improvements, and refines the specifications until all criteria are met. Only then is the optimized briefing forwarded to the creative team for final approval. This process runs entirely via standardized MCP interfaces between our project management system, Adobe tools, and our internal quality standards.
This solution demonstrates how MCP not only connects individual tools, but also automates entire quality assurance processes—from the initial briefing input to the standard-compliant output.
The future does not belong to isolated AI tools, but to intelligent, networked systems that can communicate in real time with a company's entire digital infrastructure. From development to content delivery, MCP becomes the invisible nervous system that connects all systems.
Conclusion: Content management is becoming more important, not less important
A lot is changing. Technically structured websites and intelligent content management are becoming more important than ever before. You need to manage your content in such a way that the right content is found at the right time. The goal is no longer just to get someone to click, but to appear in the agent's responses.
In short, there is a shift from “being found” to “being quoted.”
Christoph Behounek, November 2025