Although collecting information is a critical point in steering processes, a company doesn't always have absolute control over it.
Missing data can prevent certain analyses from being carried out, or even result in decisions being made about a misleading context. However, it can be complex and/or costly to attempt to correct past data. Analytical methods can be used to overcome these constraints by reconstructing the past and to predict future periods.
We helped our client to reconstruct the revenue generated by its commercial tenants.
Our role consisted of:
Automatically collecting and cleaning the current data (identification of reported but irregular revenue)
Producing a prediction engine for associating a context (period, environment) with revenue
These developments contributed to:
Better knowledge of the past and in particular an improved estimate of revenue generated in commercial complexes that indicated significant discrepancies
Accelerating the process of estimating recent periods starting from the first samples collected
Giving our clients the ability to simulate the location of chains