Companies of this type typically operate in widely divergent areas, in independent companies, often without actual interaction with each other. Data are not harmonised and have different technical formats; duplicates do exist. Just as in Replication, the Senior Management will need to be able to view selected aggregated data across companies, whereas individual companies will typically only require their own data. Since the source systems have diverse data architecture, the harmonisation task will be of a corresponding magnitude. In the context of BI, one possible idea would be to merely reflect the “silo” organisation that exists for accommodating company-specific reporting, i.e. maintain separate data architecture per company. A consolidated data architecture can be built on top of this where only necessary data, as required by the Senior Management, are harmonised at an aggregated level. Controversial with regard to the EDW mindset? Yes, certainly, but also an approach that slashes costs tremendously, while adding approximately the same business value. Data governance in this type of companies typically enjoys little attention, which is why the BI function should perhaps accept this and instead assume responsibility for in-house harmonisation, with a focus on what data are required for the overall reporting.