Under the Copernic contract, 18 Facility Management services are defined and subjected to a contract through SLAs. These services concern 47 French Thales sites with a surface area above 2000 square metres. Each service is quoted separately by Vinci Facilities, the service provider.
The purpose of this study is to isolate the most relevant determinants of this price, in order to better understand the formation of the prices as well as to provide information making it possible to better judge Vinci's offers. The standardised contractual framework and the data collection carried out by the General Services (both at the central and local levels) allow us to have a sufficiently complete database to consider statistical analysis methods. From this database, various variables have been identified as being of interest: the tertiary, industrial, data centre, clean room surface area, the number of occupants, the level of agreement between the VF and Thales managers of a site, the age of the site, the number of equipment per SLA and per site, and the location of a site in Paris or Province.
After defining and presenting the indicators which we will use to analyse our results, we will proceed with unsupervised statistical learning using clustering methods, then we will attempt to predict SLA prices using a conventional linear regression method. Finally, we will conduct several impact studies where the goal will be to determine whether the variables of agreement, age and location in Paris or Province have an impact on the price of SLA.