In recent years, the internet has come to play an important role in searching for houses. This has led to novel user-generated data that can be used to investigate housing search behaviour. This project uses data from the largest digital housing platform in the Netherlands: Funda. The user-generated data include, for instance, mouse clicks and user queries. 

The novel data provide information on housing searches that until recently remained unobserved. Therefore, it is possible to analyse real search behaviour of potential house buyers in more detail than ever before. The research project is led by Joep Steegmans, Assistant Professor in the Department of Economics of Leiden University. 
 

Digital housing platform data make it possible to analyse housing search in much more detail than ever before.

Joep Steegmans, Leiden University

Processing 10 terabytes of data

The novelty of the project is that it analyses housing platform data at the lowest level possible; that is, interrelated mouse clicks made by the housing platform users. In order to study the dynamics of the buyer search processes 10 terabytes of data need to be processed and analysed. This creates important computational challenges. Setting-up an in-house IT infrastructure that can live up to the challenge is costly and time-consuming. 

Infrastructure as a Service and Software as a Service

The computational challenges, and storage issues, have been overcome by using a commercial cloud service. The Google Cloud Platform has been chosen due to it being a flexible cloud solution. The main cloud services are related to storage and advanced analytics; that is, Google Cloud Storage (IaaS) and Google BigQuery (SaaS). The cloud service provides the opportunity to do a large-scale research project with limited resources, in terms of both personnel and funding. 

The commercial cloud service has saved Leiden University from having to set up its own IT infrastructure and made it possible to do a complex big data project with modest means. Furthermore, the cloud solution is highly scalable; allowing for changing needs over time. The project demonstrates that commercial cloud services can be valuable, for instance, in economic research
 

The commercial cloud has provided us the opportunity to do a large-scale research project with limited resources.

Joep Steegmans, Leiden University

Providing new insights into buyer search dynamics and decision making

The OCRE Cloud Voucher Funding presents the opportunity to push the frontiers in housing market research by providing new insights into buyer search dynamics and decision making. The project demonstrates the potential of commercial cloud services in Economics, a field in which these cloud services are still rarely used. 

The project shows that big (platform) data can be processed and analysed with commercial cloud services. The project can thus stimulate the use of these novel data for both academics and policy makers.