I was fortunate to be given the task of “investigating” the newest version of the Microsoft BI stack at last week’s SharePoint 2009 conference in Las Vegas. It was a very interesting, productive, well done, and well-attended conference. As both a QlikView partner and a Microsoft Inner Circle partner, it is important that I stay up-to-date on the latest offerings from both vendors. If you are pressed for time, let me sum up my thoughts in a sentence…Microsoft has made significant strides in improving their BI stack, however, their own team recognizes that they are still playing catch-up when it comes to in-memory analytics.
For those of you who have time to keep reading, let me elaborate a bit. Microsoft’s BI stack is still comprised of many pieces and parts. You have SQL Server in most cases, SSAS cubes, SharePoint, SRS reports, PerformancePoint services, Excel 2010, Excel Services, “Slicers”, and the newly announced PowerPivot add-in for Excel 2010 (formerly known as Project Gemini). Just like SharePoint is often referred to as the “swiss army knife” of business applications, it seems like the Microsoft BI stack is also a swiss army knife of sorts. It can do a lot of things, but the question is…can it do them well?
My impression from the many BI-related sessions that I attended, is that it takes a lot of deep experience and time to execute a robust BI strategy when leveraging the Microsoft BI stack. This is in comparison to using QlikView to accomplish the same strategy. Understanding when to use which Microsoft BI “tool” in the right manner and then hooking them together takes more skill and time then using QlikView. For some companies who have made significant prior investments in base Microsoft technologies, like SSAS cubes, this may not be a bad thing. But for those companies who are still somewhat new with their BI strategy, or are on other BI platforms, they will likely find the move to QlikView to be easier to manage and will realize faster time to value.
Now, that being said, the Microsoft BI stack appears to be making significant improvements. I’ll go into more details in a subsequent blog post, but let me just scratch the surface here. The crown jewel in the new Microsoft BI world is PowerPivot, Microsoft’s answer to in-memory analytics (again, previously known as Project Gemini). When demonstrated to the conference crowd, PowerPivot got a lot a “oohs” and “ahhs”. For those of us familiar with QlikView, our reaction was a little less enthusiastic because we’ve simply seen it before. Similar to QlikView, PowerPivot let’s you pull data from a number of different data sources, link the data together, load it into memory, and then analyze the heck out of it. Each presenter of PowerPivot seemed to have a laptop running a 64-bit OS, probably with 8GB of memory and had a standard load of data with 101 million rows of data. Yes, you will no longer be limited to 65k rows in Excel…”which is nice”. (Almost as nice as receiving total consciousness on your deathbed…any Caddyshack fans out there?!)
PowerPivot seemed to work pretty well, with only a few crashes here and there. It leverages “slicers” as well, which are akin to list boxes in QlikView. Slicers use roughly the same model of color-coding to indicate which field values are “selected”, which are “associated” and which are “unrelated”. Different slicers can be similarly “associated” with each other and selections change the data being analyzed and/or visualized very quickly. This idea has been proven to add value in QlikView and I am glad to see Microsoft add this functionality. By the way, the slicers also work with the new Excel 2010 pivot charts. The key thing to note about PowerPivot is that it is an Excel powertoy. It is intended for individual analysis…at least at this point in time…and therefore must be used only within an Excel workbook. Their intended market is to go after 40 million Excel users, rather than 9 million business analysts. I see their point, but am not sure if I would have approached it the same way. Feel free to add your comments if you like.
Now, don’t get me wrong, PowerPivot is a really nice move for Microsoft, and many business analysts will love it. However, you cannot use PowerPivot to “power” an interactive, user-oriented dashboard. You are able to post the PowerPivot-enabled Excel workbook in the SharePoint document library for others to check-out and then use in their own version of Excel, but that is not as user-friendly as adding the object right into a dashboard. As is so often the case with life, the key to happiness is the set your expectations properly. For those users who have never used an in-memory, associative BI tool, PowerPivot appears to be a major advancement. For most QlikView users, PowerPivot signals a “shot across the bow” indicating that Microsoft is keenly aware of the fastest growing BI software company in the world.
That’s it for now. Check out my subsequent posts to learn more about my thoughts on the Microsoft BI stack. I’ll discuss some of the other BI stack components and will even talk about a few capabilities that surpass QlikView.
Until next time…Shawn