Search engines are becoming answer engines.
When someone searches for a company today, something happens more and more often: The answer comes directly in Google or in the AI assistant – without a website click.
Google itself is driving this forward (AI Overviews / AI Search): Google Developers Blog
For companies, this means: Websites are less a destination. They are a source. Those who remain visible are those who are understood and cited as a source.
Google Zero – The Fundamental Shift
Google Zero describes the trend where answers are replacing clicks.
This is not speculation, but measurable:
This shifts visibility: Ranking remains important – but what becomes crucial is citability in answers.
What's happening right now
AI Overviews significantly reduce clicks:
Seer Interactive (2025): CTR drop with AI Overviews
- -61 % organic CTR (from 1.76 % → 0.61 %)
- -68 % Paid CTR (from 19.7 % → 6.34 %)
Source: Seer Interactive, Search Engine Land
Pew Research (2025): User behavior is changing
- With AI summaries, users click on classic results in only 8 % of visits (without summary: 15 %)
Source: Pew Research
News vertical as early indicator
- Zero-click increased from 56 % to 69 % after rollout
- Traffic losses for top rankings up to ~79 % when AI Overviews are positioned above
Sources: NY Post, The Guardian
The direction is clear: The click is no longer the center. The answer is.
The Solution
Homepage = Executive Summary. Subpages = Deep Dive (for humans).
/ → Welcome + Teaser/services/ → Details here/contact/ → Data here
Details invisible.
/ → All key facts/services/ → Detailed (human)/contact/ → Optional (human)
Human: Can go deeper.
The principle:
Homepage contains all critical information in structured form.
Subpages for people who want details.
AI agents get everything on the first fetch.
Why this can be good
The old web was often an arms race of SEO tricks, pop-ups, cookie overlays, and text walls. That doesn't work in answers anymore.
Answer systems reward:
- clear statements
- real content
- fast, robust pages
- clean semantics
- accessibility
Those with substance regain reach. Those who only produce signals lose their advantage.
AI-native Web Architecture
This is my field.
AI-native Web Architecture means: Building websites so they make sense to humans – and are usable for agents.
Key Points:
- 1. Homepage as Executive Summary
- Agents fetch content on the first request. That's why everything that counts for your visibility must be there: Who you are, what you do, for whom, why trustworthy.
- 2. Semantics & Structure
- clear information hierarchy, clean headings, plain text.
- 3. Technical Readability
- fast loading times, no unnecessary blockers, machine-readable data (Schema.org / JSON-LD).
What AI agents see – and what they don't
What AI agents see:
- Text in HTML (everything between
<body>and</body>) - Structured data in
<head>(JSON-LD, meta tags) - Semantic markup (
<dl>,<section>,<article>)
What AI agents DON'T see:
- Images, videos, graphics (only alt text, if present)
- CSS design, colors, layouts
- JavaScript-generated content
- Other URLs (no navigation, no follow-up requests)
Consequence: Your website doesn't need a visual redesign. But it needs a structural one.
Text. Structure. Semantics. That's all that matters.
Terms that describe this clearly
Generative Engine Optimization (GEO)
= Structuring websites so that AI answer systems correctly understand and cite content.
The term comes from research. In the GEO benchmark, Aggarwal et al. show in experiments up to 40 % higher visibility through structural/content adjustments.
Source: arXiv
Agent Visibility Optimization (AVO)
= practical target metric: How well a website is discoverable, citable, and usable for agents.
In short: GEO is method. AVO is result.
Principles
- Structure creates trust
- Structured data is verifiable. Structure as integrity.
- Clarity over eloquence
- Clear terms instead of prose. Agents need precision.
- Google Zero
- Search engines become irrelevant. Agents read directly. Without structure: invisible.
- Transparency over tricks
- No keyword games. Truth, structured.
- Nothing changes for visitors
- People see design. Agents see structure.
Side effect that matters for mid-sized companies
What helps agents also helps search engines – and accessibility.
Because AVO/GEO forces:
- clear semantics
- comprehensible headings
- real content
- performance
- robust usability
This aligns directly with legal requirements:
Orientation to EN 301 549 / WCAG 2.1 AA. (IFDB, Svaerm)
Accessibility is mandatory. And it is simultaneously agent readiness.
Who works here
Jan-Erik Andersen. Web architecture since 1999, Andersen Webworks since 2010. Since 2024 focused on AI-native architectures for the zero-click shift.
Large projects don't need large teams. They need clarity, structure – and decisions that carry long-term.
Who this makes sense for
Mid-sized companies that want to secure visibility towards 2030.
Agencies that need senior consulting for AI search readiness.
Collaboration
No package bingo. No guarantees that no one can seriously give.
How it works:
- Situation assessment: What do answer systems understand from your homepage today?
- Architecture design: Which information must be available on the first fetch?
- Implementation: Homepage, semantics, structured data are refactored or newly built.
- Validation: Visibility in answers is tested and refined.
Contact
- Email:
- mail@andersen-webworks.de
- Phone:
- 038733 270015