AI Trust AuditBETA
    Back to overview
    Studies & benchmarks

    Strengths and Weaknesses: Study on AI Visibility of 10,000 Websites

    GEO Expert22. June 20267 min read
    Strengths and Weaknesses: Study on AI Visibility of 10,000 Websites

    E-E-A-T and Generative Engine Optimization: Status Quo of AI Visibility (June 2026)

    In the age of generative Artificial Intelligence, the way information is consumed on the web has fundamentally changed. It is no longer enough to be listed in classic search engines. The decisive question for modern companies is: Is my company perceived and cited as a trustworthy source by Large Language Models (LLMs) like GPT-4, Claude, or specialized Search Generative Experiences (SGE)?

    In our large-scale study in June 2026, we examined over 10,000 websites regarding their AI visibility and trustworthiness. The results show a clear discrepancy between technical accessibility and content authority.

    Core Thesis: In 2026, 94 percent of websites are technically accessible to AI—but only 39 percent are credible. This trust gap will determine visibility in generative search systems in the future.

    Key Takeaways

    • Average Score 64 Percent: Technical accessibility is resolved, but content quality varies greatly.
    • Trust Gap: Technical accessibility of 94 percent stands against only 39 percent of proven authority (Source Authority).
    • Only 9 out of 100 Websites are AI-Excellent (Score ≥ 80 percent), while around 6 percent fall into the critical range with under 40 percent.
    • Biggest Lever: Structured data, reputation signals, and trust marks—not more technology.

    Overview of Core Results

    The average score across all examined websites is 64 percent. While this seems solid at first glance, the distribution of the data reveals significant differences in the maturity of the examined online presences.

    Distribution of trust scores across over 10,000 examined websites

    It is concerning that around 6 percent of websites achieve a critical score below 40 percent. These companies run the risk of simply not appearing in AI-based answers or being classified as qualitatively insufficient. The majority of websites move in the midfield between 40 and 79 percent, while only a small portion (927 out of over 10,000 websites) can be classified as excellent (over 80 percent).

    Category Analysis: Great Tech, Poor Reputation?

    The analysis of the different audit areas shows a clear picture. Websites today are technically excellently positioned but often neglect the signals that are essential for the evaluation of authority and reputation by AI models.

    Ranking of examined categories by average score

    Crawling Permissions: The Foundation is Laid

    With a top value of 94 percent in the "Crawling Permissions" category, the data shows that most websites explicitly allow access for AI bots or at least do not hinder them. This is the first necessary step for visibility in Generative Engines.

    The Challenge: Source Authority and External Reputation

    At the lower end of the scale, we find the categories "Source Authority" (39 percent) and "External Reputation" (44 percent). This is where the greatest potential for optimization lies. AI models evaluate the trustworthiness of a website heavily based on external signals and the proven expertise of the authors. If this evidence is missing, the likelihood of a recommendation by the AI decreases drastically.

    Top Performers: Where Websites Already Shine Today

    The technical standard on the web is at an impressive level in June 2026. Many basic metrics of web performance and accessibility are met almost across the board.

    Top 10 individual checks with the highest success rates

    Particular highlights include:

    • HTTPS Protocol: With a 99.6% success rate, encryption has become the absolute standard.
    • AI Bot Approval: Access permissions for OAI-SearchBot, Cohere-AI, ChatGPT-User, and PerplexityBot are consistently around 96 percent. This proves that companies have recognized the relevance of these new traffic sources.
    • Mobile Optimization & Speed: With 96 percent each, the technical homework in the area of user experience is largely completed.

    Curious which online shops and websites perform best in our analysis? Our Top-50 ranking of the highest-rated websites highlights the front-runners in AI visibility and trust.

    Critical Deficits: The "Flop Checks" as Growth Inhibitors

    Despite the strong technical basis, the "flop checks" reveal serious gaps in the machine-readable communication of trust and structure.

    The 10 most common deficits in the AI Trust Audit

    Missing Structured Data (Schema Markup)

    A central problem is the lack of use of Schema markups. Only 21 percent of websites use FAQ or Product Schema correctly. These data points, however, are the "native language" of AIs. Without them, models find it difficult to precisely extract product features, prices, or frequently asked questions and integrate them into answers. Organization Schema (40%) and Breadcrumb Schema (34%) are also used too infrequently.

    E-E-A-T and Reputation: The Untapped Potential

    The study shows that many websites do not optimally prepare important trust signals:

    • Only 36 percent have a high-quality "About Us" page that conveys expertise and history.
    • The publication date of content (Publication Date) is clearly defined for only 25 percent, which complicates the evaluation of topicality by AIs.
    • Trust Marks as Trust Anchors: An important signal for reputation in E-commerce is the presence of recognized certifications. The Trusted Shops trust mark represents a particularly strong, machine-readable trust signal here. It not only signals security to human users but also helps AI algorithms classify a website as a reputable and validated player in the market. Currently, however, only 46 percent of the examined sites possess this important signal.

    Conclusion: From Accessibility to Authority

    The data from the June 2026 report makes it clear: Mere accessibility for AI crawlers is resolved. The new challenge for companies is to make the quality and trustworthiness of their content machine-readable (E-E-A-T: Experience, Expertise, Authoritativeness, Trustworthiness).

    Anyone who wants to take place in the search results of the future must imperatively invest in structured data and actively and technically cleanly document their own reputation—for example, through strong partners like Trusted Shops.

    The Three Most Important Levers

    1. Expand Structured Data: Provide Product, FAQ, and Organization Schema in machine-readable format.
    2. Prove Reputation: Technically document recognized trust marks like Trusted Shops correctly.
    3. Make Expertise Visible: Establish authors, publication dates, and a strong "About Us" page as E-E-A-T signals.

    Methodology: This study is based on the automated analysis of over 10,000 websites during the period of June 2026. Over 50 individual criteria from the areas of technology, content, E-E-A-T, and AI visibility were collected and aggregated.