ImmoScout24
end-to-end AI layer over a marketplace, pushed into higher-tier subscriptions to raise revenue per professional client
ImmoScout24 launched in 2025 the AI search assistant HeyImmo, an image search that accepts natural language queries, and AI features in its professional CRM Propstack, and became one of the first European real estate platforms to launch an app in ChatGPT, on 2025 revenue of 649.6 million euros up 14.7% that the company does not attribute to AI.
Key points
- ImmoScout24 launched HeyImmo, an AI property search assistant, and a natural language image search.
- The platform is one of the first European real estate marketplaces to launch an app in ChatGPT.
- The professional CRM Propstack received AI features: voice-dictated listings, automatic floor plan generation.
- 2025 revenue of 649.6 million euros (+14.7%) and professional ARPU +9.5%, with no attribution to AI by the company.
Objective
Answer the threat that generative search poses to classifieds marketplaces by becoming the intelligence layer itself rather than a catalog queried by others. On the demand side, the assistant and image search aim to keep the property seeker on the platform. On the revenue side, the AI features feed higher-tier professional subscriptions: monetization does not come from selling a standalone AI product but from moving professional clients toward more expensive offers.
The deployment
Scout24 spent 2025 laying an AI layer across the whole of ImmoScout24, on both sides of the market. On the property seeker side, the platform launched HeyImmo, an AI assistant that personalizes the search experience, and built an image search that accepts natural language queries into its classic filter-based search. It also became one of the first European real estate platforms to launch an app in ChatGPT, which amounts to accepting that the entry point for search can be a third-party agent. On the professional side, the Propstack CRM was enriched in 2025 with AI features: voice-dictated listing creation and automatic floor plan generation. The whole is presented as an agentic operating system for real estate, built on the immo.ai layer. The monetization logic is indirect: these features feed higher-tier professional subscriptions. The 2025 figures published on February 26, 2026 show revenue of 649.6 million euros, up 14.7%, professional revenue of 470.5 million, up 14.8%, professional ARPU up 9.5%, consumer ARPU up 4.2%, and an average of 26,027 professional clients. A necessary caveat: Scout24 publishes no isolated revenue for its AI products and does not attribute the ARPU increase to AI. The setup is established; its commercial effect is not.
Results Proof B
The anchor source S1 is Scout24's official publication of preliminary 2025 results (February 26, 2026), a primary document that records the AI products launched (HeyImmo, natural language image search, app in ChatGPT, Propstack AI features) and provides the year's financial figures: revenue of 649.6 million euros up 14.7%, professional ARPU up 9.5%. The Online Marketplaces coverage (S2, June 22, 2026) confirms the agentic positioning. This is not level A because Scout24 isolates no revenue for these AI products and does not attribute the ARPU increase to AI: the specific commercial result is not proven. It is above a plain level C because the deployment and the figures come from the company's primary results document, not from the press. Presenting ARPU as an effect of the setup would be an inference dressed up as fact.
How it works
Inferred typical approachThe internal detail is not public. Here is a proven approach that leads to the same result, to adapt to your stack.
The stack in detail
- outil HeyImmo AI assistant that personalizes the property search experience on the consumer side, advanced as part of the 2025 AI strategy.
- outil Recherche par image Visual search that accepts a natural language query, integrated into the classic filter-based search.
- plateforme Application dans ChatGPT ImmoScout24 is one of the first European real estate platforms to launch an app in ChatGPT, which exposes its catalog to a third-party conversational agent.
- plateforme Propstack CRM aimed at real estate professionals, enriched in 2025 with AI features including voice-based listing creation and automatic floor plan generation.
- infra immo.ai AI layer presented as the foundation of an agentic operating system for real estate.
How it runs, concretely
For ops teams-
1Listing submission client
The professional creates the listing in Propstack, with voice dictation and automatic floor plan generation.
-
2Catalog structuring AI
The enriched listings feed the catalog that the assistants can query.
-
3Search on the consumer side AI
The property seeker queries HeyImmo in conversation or launches a natural language image search.
-
4Entry through a third-party agent AI
Some searches go through the ImmoScout24 app in ChatGPT.
-
5Subscription upgrade marketing
The AI features feed the higher professional offers, which carry revenue per client.
The listings catalog itself, its freshness and the richness of its attributes. A search assistant is only as good as the structuring of the properties it queries: photos, floor plans, features. This is also why the listing-creation features on the professional side matter; they improve the raw material.
How your customers perceive this type of use
Sourced studiesLes consommateurs n'acceptent pas les chatbots par defaut : 64% prefereraient que les entreprises n'utilisent pas d'IA dans leur service client (Gartner, 2024) et pres d'un utilisateur sur cinq du service client par IA n'en retire aucun benefice (Qualtrics, 2025). L'acceptation se construit sur trois conditions mesurees par Salesforce : savoir qu'on parle a une IA, pouvoir escalader vers un humain, comprendre la logique de l'agent.
Acceptance conditions
- Etre informe qu'on parle a une IA et non a un humain (pres de 75% le demandent, Salesforce 2024)
- Un chemin d'escalade clair vers un agent humain (45% plus enclins a utiliser l'agent IA, Salesforce 2024)
- Une logique de l'agent clairement expliquee (44% plus enclins, Salesforce 2024)
Red lines
- Rendre l'humain injoignable : c'est la premiere inquietude des consommateurs sur l'IA dans le service client (Gartner 2024) et 50% craignent que l'IA les coupe du contact humain (Qualtrics 2025)
- Remplacer le service client par l'IA sans alternative : 53% envisageraient de partir chez un concurrent (Gartner 2024)
Sources: Salesforce 2024 · Gartner 2024 · Qualtrics 2025
How to replicate
Inference, not sourcedData prerequisites
- A structured and fresh listings catalog, with attributes usable by an assistant
- Visuals of sufficient quality to support an image search
- A search history that makes it possible to understand the intents actually expressed
- A link between the use of the AI features and the subscribed subscription tier, without which monetization will stay invisible
Org prerequisites
- Control of the catalog and of the listing-submission tool, on the professional side
- A clear policy on exposing the catalog to third-party agents, which shifts part of the customer relationship
- A range of professional subscriptions able to absorb the new features as an upgrade argument
- AI Act compliance for the generated recommendations
Possible stack
- Conversational assistant connected to the catalog
- Visual search with a natural language query
- Content generation features in the professional tool (dictation, floor plans)
- App or connector to a consumer conversational agent
The plan, step by step
- Step 1Tool the listing-submission process on the professional side to enrich the catalog automatically (dictation, floor plan generation).Deliverable: A more structured catalog, with no extra effort for the agency.
- Step 2Connect a conversational assistant to this catalog, on a limited search scope.Deliverable: Search assistant in production on one segment.
- Step 3Add visual search for the intents that text expresses poorly.Deliverable: Natural language image search.
- Step 4Decide explicitly on exposure to a third-party agent, weighing the traffic gained against the customer relationship shifted.Deliverable: Documented position on presence in consumer agents.
- Step 5Link the AI features to the subscription range and measure the actual upgrade.Deliverable: Measurement of revenue per professional client attributable to the new features.
First step: Improve the raw material before the interface. A search assistant fails first on poor listings, not on its model. The first useful piece of work is to tool the listing-submission process on the professional side so that each property arrives structured, with usable floor plans and attributes.
The same pattern, proven in other industries
Cars.com
Marketplace AI building blocks monetized on both sides: conversational search on the buyer side for conversion, and generation and AI tools packaged and sold to subscribing dealers.
Grab
AI assistant for merchant advice and activation
Ping An
AI conversational agent for service and sales support on the customer base
Sources
- S1 Scout24 publishes preliminary results for 2025 and provides guidance for 2026 Primary archive pending
- S2 Scout24 Lifts Shareholder Returns to €455M and Bets Big on Agentic AI Secondary archive pending
An error, newer info, a source?
This page lives on its accuracy. If a figure has moved, if the deployment has changed, or if you have a higher-quality source, tell us. Every sourced correction is verified before publication.