Strong collaborative visualisation, though sometimes challenging administration

I think core features are generally easy to use, and most things function very consistently. It is easy to get up and running quickly when building reports. However, in my opinion, some things are made *too* flexible, especially for end-users, often with an all-or-nothing approach (either the end-user has access to all these features, or none of them). New users can easily get lost when trying to consume reports, since the app and web interface cannot be customised / unused features cannot be hidden. Good efforts recently to provide more granular access to enable/disable certain individual features or interactions, but, in my experience, these settings are inconsistent, are not always easy to find, and sometimes must be done in counterintuitive places, e.g. online settings after publication to the Microsoft Power BI service.

I think Microsoft Power BI is perfectly suited when seeking to visualise an existing data model that requires little to no adjustment. It offers good semi-self-service capabilities within the context of a dashboard, so users can interrogate well-modelled data themselves while having a baseline report provided to them.

In my opinion, Microsoft Power BI is not the best option when data needs significant enhancement or re-modelling. Although it is possible and can be fairly straightforward to import data from different sources, in my experience, transformation steps are inefficient and incremental loads are not possible -- you will be pushed to Microsoft Fabric at a significantly higher cost when looking for enhanced data modelling.