In today’s digital publishing landscape, visibility is crucial. For media houses striving to extend their reach beyond loyal local readers, platforms like Google Discover offer immense potential — if content is optimised correctly.

Südwest Presse, one of Germany’s leading regional newspapers, partnered with Retresco to transform its search strategy through AI-powered content tagging — achieving up to 300% more reach on Google Discover.

Südwest Presse reaches more than 1 million readers daily, primarily through strong local and regional journalism. However, the editorial team recognised a gap: Whilst local articles performed well, national content — such as sports, entertainment, and broader current events — struggled to gain traction on external platforms like Google Discover.

The goal was to make national content just as discoverable and engaging to audiences across Germany as their local reporting.

The AI solution

To address this, Südwest Presse embarked on a partnership with the media AI specialist Retresco. Together, we designed and implemented an AI-driven, entity-based content tagging system.

At the heart of the solution was a robust natural language processing (NLP) engine capable of:

• Analysing more than 250,000 indexed articles.
• Identifying key entities and topics.
• Applying semantically relevant tags aligned with Google’s taxonomy.

Leveraging the Google Knowledge Graph API, the system ensured that article metadata was accurate, standardised, and optimised for search visibility. This not only improved visibility but also helped Google better understand the content, increasing its chances of appearing in Google Discover feeds.

More than just metadata

This was more than a technical implementation; it was a strategic transformation.

For example, by adjusting tags from the generic “Champions League” to the Google Discover-aligned “UEFA Champions League,” Südwest Presse saw a notable improvement in article visibility.

This type of semantic precision became a cornerstone of the tagging process, ensuring that content was not only discoverable but also highly relevant to user interests.

Data-driven success

The results of this collaboration were both measurable and remarkable:

• 3.4x increase in visibility. An experimental group of articles tagged achieved an average of 1,305 pageviews via Google Discover — 3.4x more than the control group. This underscores the effectiveness of precise, AI-powered tagging in capturing audience attention.
• Over 1 million pageviews from 820 articles. A total of 820 AI-tagged articles generated more than 1 million pageviews during the test period. Despite representing just 1% of the total article volume, these pieces accounted for 3.6% of all Google Discover traffic.
• Streamlined editorial workflows. By automating the tagging process, the editorial team was able to eliminate manual tagging efforts, ensuring consistent metadata quality whilst freeing up time for other journalistic activities.

Bridging journalism and discoverability

In an era where digital visibility is shaped by algorithms, aligning editorial content semantically with platforms like Google Discover is essential.

This project demonstrates editorial quality alone is not enough; you also need intelligent systems that match your content with user intent. This is especially vital given the rise of AI-driven search experiences, such as AI Overviews, AI Mode, and tools like ChatGPT.

By optimising how stories are structured, categorised, and discovered, content tagging empowers publishers to ensure that high-quality journalism reaches wider, more relevant audiences. With platforms like Google Discover becoming central to content distribution, now is the time to embrace semantic tagging and intelligent automation.

The success of Südwest Presse underscores this potential: Through precise, entity-based tagging aligned with Google’s taxonomy, the publisher streamlined editorial workflows and extended its reach well beyond its regional core audiences.

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