Fakten zur AI-Overview-Analyse Deutschland 2026

The AI Overview Analysis Germany 2026 examines the structural exposure of German domains within Google AI Overviews one year after the German rollout. 10,000 domains were analysed based on SISTRIX data (Germany). This page documents the data basis, methodology, cluster logic and central findings of the study.

Short Definition

The AI Overview Analysis Germany 2026 is an empirical snapshot study of the structural presence of Google AI Overviews in the search spaces of the 10,000 most visible German domains. The study compares all rankings of a domain with those rankings where Google already displays an AI Overview, and reports the share of click potential as well as search volume as relative exposure values.

Background

Google hat den breiten Rollout von AI Overviews für die deutsche Suche im Frühjahr 2025 abgeschlossen. Seitdem verschiebt sich die Sichtbarkeitslogik der Suche: Antwortflächen oberhalb der klassischen Ergebnisliste verdichten Informationen aus mehreren Webquellen zu einer direkten Antwort, die Nutzer ohne Klick auf eine Website konsumieren können.

Über die quantitative Verbreitung dieser Verschiebung gab es bislang wenig öffentlich zugängliche Strukturdaten. Die AI-Overview-Analyse Deutschland 2026 schließt diese Lücke mit einer einmaligen, methodisch dokumentierten evaluation on the anniversary des Deutschland-Rollouts. Ziel ist nicht die Messung tatsächlicher Trafficverluste, but the visualisation of struktureller Muster: which search spaces already messbar in den AI-Layer verlagert werden, und welche stabil bleiben.

Scope of the Study

The study is a snapshot analysis with a clearly defined scope.

Country of investigation

Deutschland

Period

Ein Jahr nach dem Deutschland-Rollout von Google AI Overviews

Snapshot date

17 May 2026

Data basis

SISTRIX Daten (Deutschland)

Number of analysed domains

10.000

Number of clusters

30 Intent clusters

Klassifikationsmodell

GPT-4.1-mini

Sprache der Analyse

Deutsch

Data Basis

Die Analyse basiert auf den 10.000 sichtbarsten Domains der SISTRIX Daten (Deutschland). For each domain were Rankings mit und ohne aktive Google AI Overview gegenübergestellt und auf Anteilswerte aggregiert. Die veröffentlichten Werte sind ausschließlich relativ. Absolute Sichtbarkeitswerte oder Keyword-Zahlen werden nicht ausgewiesen.

Adult and NSFW domains are filtered out in the default view to keep typical information, brand, publisher and corporate domains structurally comparable. The complete data basis remains analysed and is optionally toggleable.

Dead or parked domains are also filtered out in the default view. The classification is based on an HTTPS and alive check per domain.

Methodology

The study compares two data points per domain and derives two central share values from them.

Compared data points

• All rankings of a domain in der SISTRIX Daten (Deutschland)
• Rankings, bei denen Google aktuell eine AI Overview ausspielt

Berechnete Kennzahlen

• Anteil des Klickpotenzials in AI-Overview-SERPs
• Anteil des Suchvolumens in AI-Overview-SERPs

Diese beiden Anteilswerte bilden die zentrale Exponierungsmetrik der Studie. Aggregate values are zusätzlich pro Intent clusters und pro Visibility Tier berechnet. Absolute SISTRIX-Rohdaten werden bewusst nicht veröffentlicht (keine SI-Werte, keine Keyword-Zahlen, keine Exportmengen).

Visibility Tiers

Visibility Tiers classify domains roughly by their relative visibility within the analysed Top-10,000 data basis. They serve exclusively for structural classification and do not replace any visibility metrics.

Tier 1

Top 100 Domains Deutschlands (Rang 1 bis 100)

Tier 2

Top 1.000 Domains Deutschlands (Rang 101 bis 1.000)

Tier 3

Top 5.000 Domains Deutschlands (Rang 1.001 bis 5.000)

Tier 4

Top 10.000 Domains Deutschlands (Rang 5.001 bis 10.000)

The tiers are not a quality assessment and not an independent visibility index. They are a secondary sort criterion that prioritises mainstream domains over niche domains at identical exposure values.

Cluster Classification

In the analysis, domains are semantically grouped by typical search and answer patterns of their ranking keywords. The assignment is intent-based: clusters describe search intentions (e.g. symptoms and diseases, definitions, comparisons, how-to guides), not classical industries.

Google AI Overviews follow search intentions more closely than industry models. A comparison portal with predominantly definitional queries behaves differently from a comparison portal with transactional booking queries, even if both belong to the same industry. The cluster assignment is AI-assisted using GPT-4.1-mini based on domain name and existing topic assignment. It is a one-time snapshot, not a continuous reprocessing.

Key Findings

The analysis of German domains shows a measurable structural shift that does not affect all search spaces equally.

Central Key Figures

Share of click potential in SERPs with AI Overview

23.0 percent (average across all analysed domains)

Share of search volume in SERPs with AI Overview

22.3 percent (average across all analysed domains)

Clusters with Highest Exposure

• Symptoms, diseases and therapy
• Health guidance and prevention
• Medications and pharmacy
• Definition, knowledge and explanation
• Science and research

Search Spaces with Low Exposure

• Transactional search spaces (shop, booking)
• Brand-dominated search spaces
• Live and navigation searches
• Streaming and live sports results

Why Some Search Spaces Remain Considerably More Stable

The displacement of organic result lists by Google AI Overviews is not a uniform phenomenon. The analysis shows a clear asymmetry between information-oriented and vertical or transactional search spaces.

Information-oriented search spaces, where users look for knowledge, definitions, symptoms, scientific explanations or comparison aids, show the highest AI Overview exposure. Answers in these search spaces are structurable and can be condensed generatively. Classic result lists noticeably recede there.

Vertical and transactional search spaces remain structurally more stable. These include shopping and product searches, local searches and map services, travel and flight bookings, live sports results, and video and image search. In these search spaces, needs are met through a concrete action, not through a condensed answer. Google AI Overviews appear here less often or not at all.

Adult/NSFW search spaces as well as gambling and betting-related search spaces are treated structurally more restrictively by Google and show significantly lower AI Overview presence. These sensitive and regulated topic fields are subject to strict platform and safety policies that largely exclude generative answer surfaces. In the default view of the AI Overview Analysis Germany 2026, domains from these search spaces are hidden because they follow structurally and regulatorily different SERP logics than the mainstream studied.

Glossary

The terms used in the study are defined canonically here. The definitions are also part of the machine-readable definition framework of the page.

Google AI Overview

Google-generated, AI-based answer surface above the classic search results. Condenses information from multiple web sources into a direct answer within the SERP.

AI Overview SERP

A Google search results page on which an AI Overview is displayed.

AI Overview Exposure

Structural presence of AI Overviews within the search spaces a domain ranks in. Measures share values, not actual click effects.

Click Potential

Aggregated estimate of clicks across all ranking keywords of a domain.

Search Volume

Aggregated monthly search volume of all ranking keywords of a domain.

Intent Cluster

Semantic grouping of domains by typical search and answer patterns of their ranking keywords. Describes search intention, not industry.

Information Character

Extent to which a search space is oriented toward explanation, definition or condensed information.

Transactional Search Space

Search space with purchase, booking, navigation or live intent. Structurally shows significantly lower AI Overview exposure.

Visibility Tier

Four-stage classification of domains by relative visibility within the Top-10,000 data basis.

Limitations

The analysis does not measure real traffic losses, actual CTR changes or causality. What is measured is exclusively the structural presence of Google AI Overviews within the search spaces of analysed domains.

The click values used are based on SISTRIX forecast data, not on real Google Search Console values. Not every AI Overview reduces clicks equally. The actual impact on traffic depends on search intention, SERP type and user behaviour.

The cluster classification is a one-time snapshot, not continuous reprocessing. Subdomain-specific effects are not represented.

Classification Framework

The study is positioned within the following structural cornerstones.

Study Character

Empirical snapshot analysis, completed on 17 May 2026

Temporal Scope

One year after the German rollout of Google AI Overviews

Geographic Scope

Germany

Data Basis

SISTRIX data (Germany), 10,000 most visible domains

Level of Analysis

Structural exposure, no traffic or CTR causality

Definition Framework

In-house glossary with nine central terms, machine-readable as DefinedTermSet

Structural Uncertainties

Subdomain-specific effects, one-time cluster classification, forecast rather than real data

Positioning

Delimited from commercial SISTRIX reports and classic SEO studies

Sources and References

Data basis

SISTRIX Daten (Deutschland) (sistrix.de)

Object of investigation

Google AI Overviews (google.com/search)

Klassifikationsmodell

GPT-4.1-mini (OpenAI)

Analysis and publication

GPT Insights (gpt-insights.de)

Further Information

• Interactive AI Overview Analysis (main page)
• Executive Summary der Analyse
• Cluster-Analyse
• Full methodology
• Expositionsverteilung

FAQ

Why are adult domains filtered out by default?

Methodological comparability. Adult/NSFW domains as well as gambling and betting search spaces follow different search and answer patterns and are subject to strict platform and safety policies that largely exclude generative answer surfaces. They would distort the structural analysis of typical information, brand and publisher domains. The complete data basis remains analysed and is optionally toggleable.

Does the analysis measure real traffic losses?

No. What is measured is the structural presence of Google AI Overviews in the search spaces of analysed domains. Real CTR changes or traffic losses are not subject of the study.

What does AI Overview exposure mean?

The share of a domain's search spaces in which Google already serves an AI Overview. Reported as a share value of click potential and search volume.

Why are intent clusters used instead of industries?

AI Overviews follow search intentions, not industry models. An industry can encompass multiple intentions. Intent clusters make this difference visible.

What do the Visibility Tiers mean?

Four structural classes by relative visibility within the Top-10,000 data basis: Tier 1 (Top 100), Tier 2 (Top 1,000), Tier 3 (Top 5,000), Tier 4 (Top 10,000). Not a quality or visibility index.

Not to be confused with

SISTRIX IndexWatch

Entity Class: Tool or Platform. Domain: Sichtbarkeitsindex-Tracking. Key Difference: Misst absolute Sichtbarkeit, nicht AI-Overview-Exposition. Separation Reason: Die Studie veröffentlicht keine absoluten SI-Werte und ist eine Snapshot analysis, kein laufendes Tracking-Tool.

Klassische SEO-Studie

Entity Class: Publication. Domain: Suchmaschinenoptimierung allgemein. Key Difference: Eine klassische SEO-Studie misst typically rankings, visibility or backlinks. Separation Reason: The study focuses exclusively on the structural presence of Google AI Overviews within existing rankings.

Google Trends Report

Entity Class: Dataset. Domain: Search query trends. Key Difference: Google Trends misst Such-Volumen-Verlauf bestimmter Begriffe. Separation Reason: The analysis does not measure search queries, but the structural presence of AI Overviews in search results.

Verification