5 reasons why we LOVE Microsoft Power BI

by Dave O'Donnell

Originally released by Microsoft in 2015, Power BI is a self-service data discovery, visualisation and reporting tool.  Microsoft has touted Power BI as a “Tableau killer” and although each product has its strengths and weaknesses, Microsoft has sent a clear signal to that it is committed to making Power BI the leading Business Intelligence platform in the market.

Originally released by Microsoft in 2015, Power BI is a self-service data discovery, visualisation and reporting tool.  Microsoft has touted Power BI as a “Tableau killer” and although each product has its strengths and weaknesses, Microsoft has sent a clear signal that it is committed to making Power BI the leading Business Intelligence platform in the market.

At Revenite we are strong advocates of Microsoft’s technology platform within the Data and Analytics space, and we are increasingly using Power BI to deliver value upfront in engagements by providing analytic insights and dashboard prototypes within a timely manner and at a low cost.  Power BI is both powerful and versatile, and we would like to share the top five things we love about it.

#1 Custom Visuals

Power BI comes bundled with all of the standard chart types including: bar, line, scatter, pie and table.  Each of these charts also come with a host of options that allow you to customise their behavioural and presentation aspects to suit.  Given Microsoft’s developer friendly heritage, they made the smart decision to “open up” the product, and allow users to extend the product by writing their own chart types (or Visuals as Microsoft likes to call them). 

PBI Custom Visuals.PNG

This has resulted in a wide range of additional visuals being made available to Power BI including spark lines, bullet graphs, maps, gauges of all descriptions, tickers – you name it!  The set of available visuals is growing rapidly supporting new uses that weren’t originally conceived at the product’s launch.  From a prospective customer point of view, it also means that if there is still a specific visual presentation that isn’t currently provided (or the available visuals don’t precisely meet your needs) then there is the option available of creating your own visual.

#2 Embedded Analytical Engine (SSAS)

Power BI contains an in memory analytical engine under the covers.  In fact, Power BI houses an embedded version of SQL Server Analysis Services (SSAS) Tabular which it uses to store all of its underlying data (or metadata in cases where the source data cannot be fit in memory).

PBI DataModel.PNG

Because Power BI uses a SSAS Tabular, data can be modelling in a relational fashion without the need to fully “flatten” the data.  SSAS Tabular uses a query language called Data Analysis Expressions (DAX) which was originally modelled on Excel but has a host of additional features.  DAX in Power BI provides you with an incredible amount of power and flexibility, giving you the ability to create complex formulas such as running sums, conditional sums, rolling sums and even entire forecasts based on historic data.  DAX calculations can also be dynamic, meaning that they can recalculate in response to the way you interact with a report or dashboard.

SSAS Tabular uses columnar compression which minimise the size of its models, and makes it incredibly fast to access.  While Excel will start to slow down when dealing with sheets of over 1 million records Power BI has no such limitation and is still blazingly fast dealing with tables in excess of 100 million records!

#3 Integrated with R for Advanced Analytics

Power BI allows users to leverage the power of R in their models.  For the uninitiated, R is an open source scripting language that includes over 5000 packages which cover a wide range of functions including statistical modelling, data manipulation and machine learning.  R also has the ability to produce complex visualisations of data and Power BI supports the ability to integrate these into a Power BI dashboard.

PBIRAnalytics.PNG

While Power BI (without R) can represent data relationships, hierarchies and measures; the inclusion of R extends Power BI to be able to incorporate Machine Learning (predictive) models which can be used to solve classification or regression problems, such as the likelihood of a customer leaving or taking up an offer.  This extends Power BI from being an OLAP tool (slice, drill-down, pivot) to an Advanced Analytics tool.

#4 Power BI embedded

Power BI is a comprehensive desktop data analysis tool able to handle duties of data integration, modelling, analysis, visualisation and reporting.  With powerbi.com (subscription required) reports and dashboards (and the underlying data) can be published and shared with other Power BI users and even with non-Power BI users (i.e. those without a subscription).

Power BI Embedded.PNG

Using Power BI embedded, Power BI visuals can even be presented to users within a webpage or application without a user having to log in to Power BI, or in fact even knowing that the visual is a Power BI visual.  This is a great option to incorporate analytic capability in a web or mobile application. 

#5 Power BI desktop is free!

While publishing Power BI dashboards to powerbi.com (to share with other users) requires a subscription, Power BI desktop is completely free.  There are some minor limitations such as a 30k row limit for data exports with the free edition (extended to 150k for pro), but otherwise all of the features of the product are available.

Microsoft provides monthly releases of Power BI desktop which ensures a steady stream of new features and product stability improvements.

In Summary

Microsoft Power BI desktop can be used as a standalone application to perform complex analytics and visualisations, and with its underlying data storage engine (SSAS Tabular) it becomes a “one stop shop” for building prototypes or to support once-off analysis pieces where the volume of data is too large for Excel to handle.  Combined with powerbi.com, Power BI desktop can also be used as an authoring and publishing platform for reports and dashboards which can be shared with other Power BI subscribers, or alternatively “white labelled” in your own applications using Power BI Embedded.

We’re very excited about Power BI, and we’re keen to see how far it can go in becoming the best Business Intelligence product in the market.