This is the last episode in the Microsoft BI tooling 101 series where I cover the Microsoft BI platform from the perspective of a Developer / DBA. I recently had to do a deep dive into this subject and therefore decided to put some of my findings into writing. As the title suggests, this part will cover how to visualize data.
Data Visualization – Rich user experience
Within the first two articles of this Microsoft business intelligence tooling series I’ve briefly covered some of the basics on how to gather and transform data including analyses, all from within Excel. Now once you (or whoever performance the analysis) have transformed and analyzed the data, it’s likely that this will need to be visualized in one way or another.
I’m targeting IT centric people here, mostly developers, and therefore won’t be covering all the steps in great detail. I just want to provide some insights on what’s already in the box when it comes to data visualization. My personal experience (Software developer’s standpoint) with reporting and Data Visualization in the past has been somewhat disappointing. Most standard tools work fine for static reports, however nowadays users are demanding a rich user experience. Being able to alter properties, apply filters, drill-down visually and in addition should work on a mobile device.
Before you start looking for chart ‘controls’ and bootstrap dashboard templates, I would highly recommend looking into Power View and the Power BI. This will not only allow non-technical users to build report/visualizations (self-service BI), but also covers a lot of advanced plumbing you most likely don’t want to build on your own.
Power View – Building data visualizations
Building reports within Power View is very simple once you have the data gathered beforehand. Power View’s report builder allows you to add columns from predefined tables on the reporting section. After pacing the table on the canvas, it’s possible to select the desired visualization style.
The interesting part is that the reports are interactive by default and therefore allowing users to explore data ad-hoc. This is a very useful feature especially during presentations and working on business questions or problems, eliminating the need of creating an endless stream of reports for the end-users. This is all possible because of the metadata in the underlying Data Model, Power View uses the relationships between the different tables and fields in the model to make items on a sheet or view highlight and filter each other.
Power View has many different ways to slice and dice the data. Filters remove all but the data I want to focus on. To filter data in Power View, I can use Filter Pane filters, slicers, and cross-filters. Highlighting is not filtering since it does not remove data but highlights a subset of the viewable data; the un-highlighted data remains visible but dimmed. Apart from this, Power View has some very solid reporting capabilities as expected.
Microsoft Power BI for Office 365
When it comes to sharing data with a larger audience, it might be a good idea to look at Office 365 or SharePoint. Here users can view and interact with the Power View sheets for a given workbook in either location.
I’ve never been a big fan of SharePoint from both an end user and developer’s perspective due to various reasons however, I got interested in what Microsoft Power BI has to offer and its roadmap. At the moment Microsoft Power includes a variety of different tools for data discovery, analysis and visualization, a mobile BI app, and in addition to this supports HTML 5 which allows users to view and interact with Power View reports on any device supporting HTML5. New features are added on a monthly basis and all is looking very promising.
The following video provides an overview of the features and tools available for Microsoft Power BI: https://www.youtube.com/watch?t=75&v=3WCJsCFQGOw
Power BI features available for preview:
Unfortunately it’s not possible to license it without an Office 365 subscription, which is a big holding back some organizations.
I hope this series helped with getting a grasp of the basic building block within Microsoft’s BI stack. The focus on self-service is a very interesting aspect. There is still a lot of work that needs to be performed by DBA’s, Developers and Data scientists when it comes to exposing data and security. But I think we all getting rid of the building reports for John Doe Finance and Jane Doe Marketing is a very good thing.