How FAZ uses an AI chatbot for service requests (German only)

Two use cases, one goal

New digital tools must offer genuine added value to FAZ staff and readers if the company is to adopt them. Two areas where readers frequently get stuck were cited as reasons for introducing an AI assistant:

The paywall: Anyone wishing to read an F+ article can sign up for a free trial subscription in three steps. However, many left the site after just one second. The reasons: fear of the ‘subscription trap’ and the question of why the article costs anything at all.

The Subscription Shop: Print or digital, daily or Sunday paper: the range on offer is vast and raises many recurring questions. Readers need help deciding which subscription is right for them.

This led to clear objectives: the bounce rate at the paywall should fall and the conversion rate should rise. In the subscription shop, the process should be simplified and the customer advice improved.

Why moinAI?

A traditional FAQ page wasn’t an option, as readers often can’t find the answer straight away and have to leave the page they’re currently on. A call centre or customer portal would have been time-consuming and too expensive. moinAI came up with a compelling alternative:

• Targeted service at just the right moment: the AI assistant appears exactly where questions arise and provides the right answers straight away.
• Fast and resource-efficient: just a few lines of code are needed for integration, and only a small team is required to manage the AI assistant.
• Experience in the publishing industry: moinAI brings with it a strong track record in the sector, a friendly, down-to-earth team and AI made in Germany.

First the paywall, then the subscription shop

Before the AI assistant went live, the project team tested the solution themselves to ensure everyone felt confident about the launch. It started in the paywall with 15 to 20 topics such as ‘How do I cancel my subscription?’, ‘Why do I have to pay for this article?’ or ‘How do I log in?’. Following the positive results, the FAZ then extended the AI assistant to the subscription shop. Thanks to automatic topic suggestions via Dreaming, the pool of topics grew to more than 40.

Today, the AI assistant automatically responds to 71 per cent of enquiries, and fewer readers are abandoning the subscription process at the paywall.

‍

This case study shows how the FAZ uses an AI assistant to integrate customer service and sales: frequently asked questions are answered automatically and readers are given personalised advice.

Find out, amongst other things:

• Why the FAZ decided against an FAQ page, a call centre and a customer portal;
• which questions cause readers to abandon the paywall and how the AI assistant steps in to help

‍

• how the FAZ increased click-through rates in the chat function to 18 per cent;
• how the FAZ introduced the AI assistant step by step and got the whole team on board;
• what publishers can learn from this project.
‍