“As of today, we can finally seal the marriage between corporate business intelligence and self-service BI using Power BI.” – Alberto Ferrari from https://www.youtube.com/watch?v=g-nRxDVt3To&feature=youtu.be (Unboxing new Power BI composite models)
Consumers are now empowered to truly “self-serve” business intelligence analytics without polluting their existing ecosystem of data; no longer are they forced to run analytics in excel just because the information they need is not available in the central corporate dataset. Composite models enable loading that ad-hoc spreadsheet containing one-off metrics into to an existing central dataset, add calculations, or bringing in another enterprise-level source of data altogether.
By connecting to an existing dataset and then creating a local model, users are now able to add their own data sources and enrich existing tables with calculated columns and measures. The new objects are only stored in the local model (even if they are connecting to the centrally published dataset). “In other words, you work as if you had a local copy of the model and full rights to modify and expand it, even though you are not duplicating any data that is stored on the server.” – Marco Russo from SQLBI (New composite models in Power BI: a milestone in Business Intelligence)
The new update to composite models allow for further optimisation of datasets, by allowing for data “siloing”. Think about it, central datasets these days are built on a one-size-fits-all scale, with some models encompassing 50 entities or more…. With composite models, you can keep your centrally managed finance data separate from your HR data separate from your population growth data – only bringing them together in one composite model for the reports that require information from all three sources.
So, go ahead, take the reign on your data exploration and harness the Power (BI) of DirectQuery for Power BI datasets and Azure Analysis Services.