Power BI : Lesson 1: An Introduction to Power BI

#

One may be forgiven for being confused at the number of ‘Power’ products that Microsoft have rolled out in the last few years. These were generally integrated into the Office products, mainly as Excel-Addins. So to recap, here they are:

Power Product Name My Description
Power Pivot

An in memory analytics tool to vastly expand the capabilities of modelling data in Excel. Closely aligned with the new Tabular model of SQL Server Analysis Services (SSAS).

Use Case: When your Excel files have a large number of rows and Excel is struggling, e.g. > 1 million rows. When you want to combine data from a number of sources in one data model for analysis in Excel.

Power View

An early attempt to deliver in-memory reporting capabilities. Nice idea but needed much more functionality and development. More an analytical tool than an operational reporting tool.

Use Case: A good POC tool but not really an Enterprise reporting tool and not comparable to SSRS reporting functionality.

Power Map

Introducing map capabilities and some nice 3D maps – also in-memory and generally pretty fast.

Use Case: Another nice POC demo tool, I haven’t seen it being used much in a production analysis way.

Power Query

An excellent tool utilising the ‘M’ language which introduces ways for the power (sorry) user to import external data sources, e.g. Wikipedia tables, and perform basic ETL while importing into a data model.

Use Case : When data resides outside the company but would be useful to bring into analysis. When one needs to do simple ETL on imported data.

 

These tools have now been decoupled from their reliance on Microsoft Office as, one of the reasons I presume, was to allow greater up-take especially as not all clients use Office.

So even though these Power products are still integrated into Excel if one chooses to use them, they are now products in their own right. So let me introduce you to the 3 Power BI products

Power Product Name My Description
Power BI Desktop

All 4 of the above products reside in this desktop report authoring application. Multi-page reports can be quickly built in here and from many (and growing) number of data sources. The reports and charts are built from data sets either connected directly to the data (more on this in further chapters) or imported data and then they are optionally published to a web portal.

Use Case: I tend to author reports in this tool as it provides a mechanism for version control, in that a physical file (pbix extension) is created. As you will see one could author directly in the PowerBI service but that way does not create a file you can save.

Power BI Service

A web portal on which authored reports can be published. Administered in the cloud by Microsoft – there is no (at time of writing) way of hosting this portal on your own environment. Authentication is managed through Office 365 / Azure Active Directory.

The portal includes Dashboards which contain pinned visuals from the authored reports (note the desktop does not include Dashboards). As well as this it includes cool functionality utilising Cortana (more on that later) which allows English type queries on your data and the tool automatically produces suggested visuals.

Visuals are rendered in d3js.

Use Case: When one needs dashboards linked to reports for others to see. A portal one could have their authenticated users accessing and sharing.

Power BI mobile

An HTML5 compliant mobile application linked to the Power BI service.

Use Case: When you require mobile and tablet sharing of your interactive dashboards.

 

So, this has been a condensed introduction to the toolset and I will take you through the details – with demos in upcoming posts.

Want to see more? Part Two here!

Latest from this author