While many media companies rush to block generative AI tools from accessing their content, emerging data suggests this could be a strategic mistake.

My analysis at NZZ, along with broader industry observations and traffic patterns from my personal blog, reveals a significant emerging trend: Though traffic from generative AI currently represents only 1% to 4% of overall visits, it’s steadily growing.

The shifting landscape of content discovery

Traditional SEO has long been an important part of a digital publishing strategy. However, the rise of AI-powered search and content discovery tools like ChatGPT, Claude, and Perplexity is creating new pathways for audience engagement.

This shift introduces two critical concepts publishers need to understand:

• Generative engine optimisation (GEO): Optimising content for AI-powered search queries.
• Generative AI optimisation (GAIO): Ensuring visibility in AI-generated results and recommendations.

Why media companies should pay attention

Our traffic analysis reveals several key insights.

AI is making up a growing traffic share. Both personal blog data and NZZ metrics show an upward trend in visitors from GenAI sources.

AI-driven visitors often engage with content differently, seeking deeper context and comprehensive information. Specifically for quality publications, this might be a very important source of traffic.

Additionally, there is a first-mover advantage. Publishers optimising for GenAI now have an opportunity to establish strong positions in this emerging channel.

Getting started with GenAI: strategic approaches for media companies

The following insights come from my initial research and experimentation. Because generative search evolves rapidly, some of these recommendations may change over time.

• Content optimisation strategies: Structure articles with clear hierarchical information, and include relevant statistics and cited sources. Use industry-standard terminology. Additionally, implement a schema mark-up for better AI comprehension.
• Technical considerations: It’s important to balance access control with discovery potential. Consider selective AI access through licensing agreements, then monitor and analyse AI-driven traffic patterns. Implement structured data formats as well.

Experimental observations

Current data suggests AI traffic exhibits unique characteristics. This includes higher engagement with evergreen content and an increased interest in contextual information. There are also different temporal patterns compared to traditional search traffic.

It’s important to note these observations are based on current trends and may evolve as the technology landscape changes. Regular strategy adjustment is recommended based on ongoing performance analysis.

Looking ahead

While this field remains experimental and evolves daily, early adoption of GenAI optimisation strategies could provide significant competitive advantages.

To adapt accordingly, publishers should develop flexible AI access policies, experiment with different content structuring approaches, monitor traffic patterns from various AI sources, and build capabilities for AI-optimised content creation.

The key is finding the right balance between protecting content assets and leveraging new distribution channels. As this technology develops, staying adaptable while maintaining a strategic presence on AI-driven discovery platforms will be crucial for publishers’ digital success.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .