We have struggled to find a best practice in regards to reusing standard data entities across our company. I expect that our use cases are commonplace. I am hoping that others have found solutions to what I understand are purposeful limitations within Power BI dataset reuse.
We have a standard dataset that contains information about stores. It has basic store info (address, type, size), contact info (phone numbers and people), alignment (district, area, region) and other related asset information regarding the stores. This dataset is the official store list and information that we want to certify for the company. All report writers should use this source.
A typical report writer will connect to the data sources within SQL Server to recreate this dataset, or copy an existing PBIX file that already contains this dataset. They will then add their custom data that they are relating to the stores, such as Sales, Demographics or whatever source data they have. They then publish their report and dataset to the service, and a refresh is scheduled for the store info, along with the other data sources.
We have 50 different report writers all writing reports utilizing store info. They are all copying or recreating store info datasets to tie to their other data. Each of these is publishing a copy of the same dataset and scheduling refreshes independently.
We can publish a certified dataset for store info, but if a report writer uses it then they cannot add any additional data. It is not practical to try to add all possible reporting data into a few models. We can get a large quantity of data covered this way, but analysts will always have their own datasets they want to combine. We do not have a way to centraly manage common data without losing the real flexibility of Power BI, which is to allow people to add a variety of data for reporting.
We continue to copy premade models and publish them with reports. This feels wrong to me. Any advice?
Thanks,
Raymond
