Microsoft Power BI powers its way to GA

Excel- and cloud-based BI suite finally comes out of preview.
Written by Andrew Brust, Contributor

Microsoft's Power BI suite, a collection of Excel features and add-ins, as well as corresponding SharePoint and Windows 8 viewer clients, has been released to general availability (GA).  In preview since last summer, when it was announced at Microsoft's Worldwide Partner Conference in Houston, Power BI represents Microsoft's first mobile BI offering, and its first cloud-based BI service, save for the now-deprecated Windows Azure SQL Reporting service.

Once upon a time, in 2010
The Power BI story started when PowerPivot was released — in conjunction with SQL Server 2008 R2 — as an add-in for Excel 2010.  Much of PowerPivot's functionality was subsequently built-in to Excel 2013, but the plot further thickened, with the release of a Power View (formerly "Crescent") add-in and preview releases of Power Query (formerly "Data Explorer") and Power Map (formerly "GeoFlow").

Power BI combines these add-ins with SharePoint-based rendering for Power View (in both Silverlight and HTML 5 flavors) and a very innovative natural language processing front-end to it, called Q&A.  Unlike many search expression-based query front-ends, Q&A interprets detailed, natural language requests for specific data presented in specific formats, and doesn't just return the data, but can make intelligent decisions on how it ought to be visualized.  Here's an example (albeit one where the visualization type was explicitly requested):

The Q&A natural language data discovery interface

Membership has its privileges
Subscribers to Power BI are also able to connect the Windows 8 Power BI app to their data models in the cloud (something that cannot be done with just a SharePoint online account).  This app provides a touch-centric interface to both Excel content (i.e. cell ranges, charts and PivotTables) as well as Power View content.  Connectivity to on-premises data is offered as well.

Power BI customers also have search-based access to data set catalogs, including versions for public and corporate data.  This is built into Power Query, which can also query Hadoop's HDFS directly, as it can Facebook, Exchange, Active Directory and a number of more conventional relational databases and data warehouse platforms.

Cheap it ain't
Mary Jo Foley reported on Power BI's pricing, when it was announced last month.  In her post, she quoted my analysis of the monthly fees (which range from a promotional $20/month/user up to $52/month/user depending on what the user already subscribes to) as often close to, or higher than, the subscription pricing for Tableau Online.  

That said, Microsoft has case studies already written up on early adopter customers of the Power BI suite.  These include Revlon, Trek Bikes and Carnegie Mellon University.  I think Power BI pricing is too high for mass appeal, but for shops already on an Office 365 E3/E4 subscription, and the Microsoft stack in general, there is good value in Power BI.

But it ain't bad, either
Power BI's Power View data discovery component is very easy to use and quite productive.  Power Map, though only viewable from the desktop, provides very attractive 3D geographic data visualizations and is also quite easy to use.

Power Map
Power Map

The combination of Power Query and Power Pivot (yes, it's now spelled with a space) make for excellent data acquisition and modeling capabilities.  The HTML 5 and Windows 8 clients still have some maturing to do, but the SharePoint and Excel-based experiences are top-notch.

More than spreadsheets
Speaking of Excel, some users will hate that it's required to do the modeling and authoring work in Power BI.  Many, many others will love it, or learn to, though.  There's no shortage of people who like to hate on Excel as a BI tool, but the fact is that most professionals on this planet crunch their numbers there and so using Excel as the native authoring environment will, to many, be a victory for common sense.

Microsoft, for its part, sees Q&A and the data catalogs as strategically key.  Corporate Vice President Quentin Clark said as much when he pre-briefed ZDNet on Power BI on Friday.  The public data catalog now works with data from Wikipedia and Dun & Bradstreet, and Microsoft has a dedicated "external data curation" team working on expanding this selection.  I would hope the same iterative improvement approach will be applied to Q&A, because, along with the data catalog, it truly has the potential to deliver on the promise of self-service BI.

Fantasy futures
As a long-time user and implementer of Microsoft's BI technology myself, what I also hope to see is true integration of Power BI's capabilities into Office and SharePoint.  I'd prefer a single add-in to four of them, and a full-fledged, from-the-ground-up implementation of HTML 5 rendering.  

I'd also like to see lower pricing, especially for individuals, members of academia, or business users willing to accept a limit on data set sizes and types.  The best way to build platform momentum is to provide easy access to enthusiasts and to reap great revenue and margins from large production deployments by customers who will be making or saving much more money with the technology then they will be spending on it.

Of course, momentum cannot be built at all unless the technology is good.  Power BI's technology is good, combining foundational technology from SQL Server Analysis Services, powerful re-purposed modeling technology from Microsoft's former "Oslo" initiative, Windows Presentation Foundation 3D graphics and impressive natural language query capabilities.  All of this technology came from Satya Nadella's Enterprise & Cloud division.  May the whole company innovate in this spirit under his leadership as CEO.

Editorial standards