After spending some time exploring MS Fabric from the perspective of a mid-size nonprofit, here are my initial observations. It is a well-put-together and powerful tool that can serve well to the right organization. On the other hand, it is not cheap and cost can increase very quickly. For organizations that are still managing Excel spreadsheets, GA, and smaller volumes of data, it can allow lightning-speed modernization of the data eco-system. However, it might make sense to take full advantage of the trial period and partner program to fully understand data needs, and better understand the tool and costs.
Pros:
- The tool is meant to tie together various data journeys from ingestion of the data to clean up and transformation to building a model and visualization. Having all those functions with one company and in one area can be very beneficial.
- The data warehouse and data lakehouse architecture allows handling structured and unstructured data much easier than before. Data warehouse make updating data easier, but lakehouse are more versatile, especially when dealing with unstructured data.
- Build-in analytics (the notes) allow using Python or R to model data and then write it back to warehouse. In the past, data would be extracted from warehouse, analyzed and then written back, which created delays and inefficiencies.
- Delta-parquet compression allows for handling very large volumes of data.
- Microsoft handles security, storage, and scaling.
Cons:
- The tool is expensive. There is a charge for computing capacities and storage. For smaller computing capacities and for the development of visualizations, the organization would need to pay for additional licenses.
- Less of computing capacities can cause delays in processing and slowdowns.
- The organization would need to dedicate at least 2 FTE to manage data integration and visualizations.
