How Saarbruecker Zeitung handles 3,800+ subscriber conversations with Paula, an AI agent built on Chatbase

At a glance

Saarbruecker Zeitung Medienhaus is a German regional newspaper publisher and digital media brand with a circulation of approximately 70,000 daily newspapers. The publisher serves print subscribers and a growing digital base across ePaper, mobile app, and the SZ+ subscription product. Its customer contact center handles thousands of inbound inquiries every month, ranging from subscription changes to digital product support. The team built Paula, a 24/7 AI agent powered by Chatbase, to give subscribers instant answers and free the contact center to focus on the conversations that need a human.

Long hotline wait times and a call center stretched too thin

The contact center was getting hit daily with calls and emails about subscriptions and delivery. When delivery failures hit, the lines piled up fast. The team tried the obvious fixes first. They restructured the call center, added more agents, brought in a phone system with partially automated complaint intake. The hold times eased a bit, but the underlying workload on the team did not. The volume kept coming, and a human always had to pick up.

Paula, the 24/7 AI agent for newspaper subscribers, live in 3 weeks

Paula is embedded as a widget on the Saarbruecker Zeitung website and handles inbound customer support around the clock. She is not used for lead capture or onboarding. She is a self-service support agent for existing subscribers and website visitors, focused on the highest-volume topics the contact center sees every day.

• Subscription and pricing questions, by far the most frequent topic
• Digital product support for ePaper, app, and SZ+ access
• Account issues including login problems, forgotten passwords, and customer ID lookups
• Contract changes including cancellations, address changes, email updates, and bank details
• Vacation and pause requests for print delivery
• General delivery and print edition inquiries

Built with no code, live in 3 weeks

The team trained Paula on the Saarbruecker Zeitung websites, internal OSC info pages, and FAQ pages, with more specific questions added through Chatbase's Q&A tool. Three things made Chatbase the right choice. No programming was required, so the team could ship without engineering support. Custom training and system prompts let them tune Paula's responses to Saarbruecker Zeitung's tone and support style. And GDPR compliance was non-negotiable for a German publisher operating under European data protection rules.

The results

Since launching in July 2025, Paula has handled nearly 3,800 conversations and over 18,000 individual messages. That is consistent traffic of around 400 conversations every month, reliably, around the clock. The headline result is capacity relief for the contact center. Every Paula conversation is one the team did not have to pick up. That time is being reinvested into higher-value work: welcome calls for new subscribers, win-back conversations with churned ones, and personalized support for customers adopting the digital products. The second result, more interesting than the volume, is engagement depth. With an average of nearly 5 messages per conversation, subscribers are not just testing Paula with a single question and leaving. They are working through real problems with her, end to end.

What's next

For now, Paula points subscribers to the right websites and forms when an action is needed beyond an answer. The team's next big project is integrations: connecting Paula directly into the systems that handle subscriptions, billing, address changes, and digital product access. Once that lands, Paula will not just guide subscribers to the right form. She will complete the action for them.