Nicolas Bouliane’s travel blog All About Berlin lost 75% of its search traffic in 2024, a decline he directly attributes to the proliferation of AI-generated content saturating Google results. The case study, which sparked a 429-point Hacker News discussion, documents a precise, real-world economic impact on a creator whose detailed, first-hand guides are being buried by lower-quality, mass-produced AI articles.
Site: All About Berlin | Traffic Loss: 75% (2023-2024) | Core Issue: AI-generated content saturation
How AI Content Wins the SEO Game
Bouliane’s analysis shows AI-powered sites outcompete human writers through pure volume and refresh rate. A human can research, visit, and write a detailed guide to a Berlin museum once. An AI content farm can generate hundreds of slightly varied articles on the same topic, target thousands of long-tail keywords, and continuously re-spin content—a scale impossible for an individual.
The technical advantage is in automated optimization. These sites use AI not just for writing, but for keyword clustering, internal linking, and publishing at a frequency that Google’s algorithms initially interpret as authority and relevance. The result, as seen in Bouliane’s search console, is that his original pages lose rankings to newer, less accurate AI pages.
The Data Behind the Traffic Collapse
Bouliane provides concrete analytics. From 2023 to 2024:
- Total traffic fell by 75%, decimating ad revenue.
- Keyword rankings dropped for core terms like “Berlin travel tips.”
- Search Engine Results Pages (SERPs) for his topics are now dominated by domains with generic names (e.g., “bestoftravel.com”) filled with factually shallow content.
The HN thread added quantitative context: several commenters, who are themselves site owners or SEO professionals, reported similar traffic declines of 50-90% across niches like programming tutorials, product reviews, and legal advice.
| Feature | Human-Created (All About Berlin) | AI-Generated Competitor |
|---|---|---|
| Research Basis | First-hand visits, photography, local updates | Web scraping, other articles |
| Update Frequency | Months/years (as places change) | Minutes/hours (automated) |
| Scale | Hundreds of detailed guides | Hundreds of thousands of pages |
| Economic Model | Ads/affiliate (requires traffic) | Ads/affiliate (volume-driven) |
| Primary Advantage | Accuracy, depth, trust | Volume, cost, speed |
Who This Hurts (And Who It Doesn't)
This trend is not uniformly negative across all users or creators.
Impacted Negatively:
- Independent creators and small publishers: Lacking the domain authority of major media brands, they are most vulnerable to being displaced by AI spam.
- Information seekers: Users get lower-quality, potentially inaccurate information, often presented with confident, authoritative prose that obscures its lack of primary experience.
- Google’s core product: Search quality degrades, pushing users toward alternative platforms like Reddit or TikTok for “real” human experience.
Unaffected or Benefiting:
- Large, established brands (e.g., Lonely Planet, NYT Travel): Their domain authority and brand recognition currently insulate them from most AI spam.
- AI content tool vendors and SEO agencies: They profit from selling the infrastructure of content creation at scale.
- Affiliate marketers with low ethical barriers: They can launch thousands of pages at near-zero marginal cost.
Bottom line: The economics of AI content creation disproportionately punish independent experts while rewarding scalable mediocrity.
Practical Steps for Creators and Consumers
For creators facing this, the HN discussion and Bouliane’s experience point to specific, though difficult, adaptations.
For Content Creators:
- Double down on E-E-A-T: Google’s guidelines for Experience, Expertise, Authoritativeness, and Trustworthiness are now a survival manual. Showcase first-hand experience with original photos, videos, and receipts. Use author bios with verifiable credentials.
- Pivot to community and platform diversification: Build an email list, a dedicated social following, or a newsletter. Direct traffic is immune to SEO shifts. Platforms like Substack or YouTube may offer more stable algorithmic treatment.
- Technical SEO is not enough: While still necessary, traditional keyword and backlink strategies are now table stakes. The differentiating factor must be unique value AI cannot replicate.
For Information Consumers:
- Audit the source: Look for signs of human experience—specific anecdotes, original media, dated updates, and a transparent author profile.
- Use search modifiers: Try adding “reddit,” “forum,” or “review” to queries to bypass generic AI content farms.
- Prioritize known entities: For critical information (e.g., legal, medical, complex tutorials), default to established institutions or recognized expert communities.
"What is E-E-A-T?"
E-E-A-T is a core concept in Google's Search Quality Rater Guidelines. It stands for Experience, Expertise, Authoritativeness, and Trustworthiness. Raters use it to assess content quality. While not a direct ranking algorithm, it informs Google's systems. In the AI content era, demonstrating strong E-E-A-T through first-hand experience, author credentials, and cited sources is increasingly critical for ranking.
The Bottom Line: A Crisis of Information Integrity
The “All About Berlin” case is not an isolated SEO story; it’s a leading indicator for the integrity of the open web. When the economic incentives for creating accurate, experienced-based content are destroyed by the economics of generating plausible, low-cost text at scale, the public’s access to reliable information deteriorates. The solution, if there is one, will require significant adjustments from search engines to reward genuine expertise and from consumers to become more discerning. The alternative is a web that is larger, faster, and profoundly less useful.
The market will correct only when users consistently reject low-value AI content, or when search algorithms become sophisticated enough to detect and demote it at scale—a technical arms race that has only just begun.

Top comments (0)