From fragmented knowledge to automated customer service

Want to experience the ImmoScout24 case study for yourself? In a joint webinar, Felix Lessing and Sandra Schütt from ImmoScout24 will show you how the chatbot was developed in practice. Click here to watch the recording!

From query chaos to AI support

Around 18.6 million people use ImmoScout24 every month. As a digital ecosystem for the property sector, the platform connects estate agents, investors, property owners and those looking for a home. Users have a wide variety of queries regarding listings, memberships, application portfolios, technical issues and much more. This poses a major challenge for the support team, as everyone expects fast, reliable answers around the clock.

The problem with ImmoScout24 was that, although much of the company’s support knowledge was available, it was scattered across external FAQs, Confluence and Salesforce. There was no consistent structure and no clear lines of responsibility. Every change was processed as a ticket via the CRM team, whether it involved a new FAQ category or an additional service unit for a subsidiary. This made it virtually impossible to respond quickly to new customer enquiries. The team even said: “Every enquiry is unique; you can’t automate anything here.”

This starting point gave rise to an AI project with three clear objectives:

1. Faster response times, 24/7;
2. clearly categorising enquiries;
3. a customer experience that customers rate highly

Why moinAI?

For ImmoScout24, one thing was clear: without a solid database, generative AI simply cannot work. They were therefore looking for a solution that would put control of content and processes firmly in the hands of the service team, rather than having to wait for other departments to make every change. In a direct comparison with the previous solution, moinAI stood out in four key areas:

• Autonomy & control: The service team manages content, workflows and service logic directly within the moinAI Hub, without having to go through other departments.
• Transparent pricing model: Fixed prices rather than hidden costs arising from maintenance, dependencies and a lack of flexibility.
• Rapid integration: A short code snippet is all it takes. Everything else can be configured directly within the tool without the need for developers.
• Legal certainty: moinAI is GDPR-compliant and hosted in Germany. AI-generated responses are automatically labelled, meaning moinAI meets the transparency requirements of the EU AI Act.

Learning by Doing as the motto

ImmoScout24 completely overhauled its enquiry category system and deliberately began with simple, frequently encountered issues, such as login problems, billing queries or questions about navigation. Existing content could be carried over directly. After around four weeks, the AI assistant went live.

It became apparent that the majority of the content can be broken down into a manageable core of recurring themes.

Today, the automation rate stands at around 75 per cent; ImmoScout24 started out at around 20 per cent in mid-2023. Of the approximately 300,000 conversations in 2025, only 20 per cent were escalated to Customer Care. When matters become complex, staff take over directly via live chat.

This case study shows how ImmoScout24 has turned dispersed knowledge into a centralised knowledge base. Today, this forms the basis for a 24/7 service without the need to increase staff numbers, regardless of how much the volume of enquiries increases. New subsidiaries are thus up and running in just three days.

Discover this and more in the case study:

• Why fragmented knowledge was the real driving force behind the AI chatbot project – and not the chatbot itself!
• How ImmoScout24 used the enquiry categorisation system to make recurring queries visible and automatable

‍

• Why ImmoScout24 decided against a voicebot and what this means in terms of costs;
• how the automatic identification of knowledge gaps is gradually increasing the automation rate;
• how ImmoScout24 aims to raise the automation rate to over 90 per cent;
• what companies in a similar situation can learn from this project.