Pricing Recommendations Built on AI Forecasts
We trained models on a Swiss real estate company's data that predict footfall and spend at individual locations, and built a pricing recommendation for spaces on top of them.
- Client
- Swiss real estate company
- Status
- Completed
At a glance
A clean data foundation
Existing data flows through reproducible pipelines into training and forecasting.
Forecasts per location
The models predict footfall and spend for individual locations and time periods.
Concrete price recommendations
The forecasts turn into a price recommendation for each location and period.
The challenge
Many real estate companies hold large amounts of data on their locations but rarely use it systematically for pricing. Prices are often set from experience and comparable values rather than from reliable forecasts.
Our solution
We turned the existing data into clean, reproducible pipelines and trained forecasting models for footfall and spend on it. A pricing recommendation builds on them and suggests a suitable price for a location and period. Training and operation run in the client's cloud environment, designed for very large amounts of data.