• Per terabyte pricing
• Linear and predictable performance
• 1.6-32 terabytes capacity
• A very competitive $/terabyte offering starting at around $13K (32 TB capacity) and ending at $25K (4 TB capacity)

In addition the reference architecture consists of best practices in

• Windows OS configurations
• SQL Server startup options
• Database physical layout
• Table types
• Indexes
• Statistics
• Managing fragmentation
• Loading procedures

This is a so called Symmetric Multiprocessing (SMP) offering with a "Shared Everything" architecture. That means (amongst other things) that a single DB instance/Server/CPU shares memory. This is opposed to Massively Parallell Processing (MPP) where nothing is shared and each Server/CPU have their own dedicated resources. The key here is parallellism and this is what Madison/DataAlegro and petabyte scalability is all about (the unit of measure in those systems is 100's of terabytes).

Even the "mid level" data volume that the reference architecture supports is still pretty rare and can mostly be found within the C&HT industries (Telcos) and Finance. With the exponential growth in data volumes seen in virtually every vertical this will change and create a bigger market for the DWH appliance vendors. 

More information can be found @ http://www.microsoft.com/sqlserver/2008/en/us/fasttrack.aspx

In the latest quadrant Microsoft is still in the leader quadrant for BI platforms. Gartner highlights Microsoft’s attractive license model, growing market penetration and quality of development tools. On the negative side metadata management is still a concern and long release cycles slow down the rate at which innovation reaches customers compared to some competitors.

Compared to last year Microsoft is overtaken by IBM / Cognos on the “Ability to execute” axis which should worry Redmond, especially since “Office integration” is highlighted as very strong.

A summary of the report can be found here.

(These opinions and observations are my own and do not necessarily reflect those of my employer)

Why is Microsoft doing this?

-          This has to be seen in light of the general economic climate. Microsoft, like everyone else, needs to adapt and as part of this look over its products in terms of profitability.

-          The cost of sales for PerformancePoint has been very high. These products require a sales process and organization that MS did not have and needed to build. This has proven more costly than anticipated.

-          The partner channel has not worked as intended. Enterprise Planning requires a functional and vertical skill set that most of the partner ecosystem did not have.

What does it mean for Microsoft’s investment in the BI platform?

-          This does not mean that MS is backing out of the BI business. Microsoft can now sharpen its focus on its successful core BI platform based on SQL Server and the Office Platform (Office+SharePoint).

-          Microsoft will continue to offer data warehouse, OLAP, data integration and front-end tools and distribution mechanisms.

-          The core BI platform is unaffected by this. In fact, SharePoint got even more attractive as a distribution platform for BI by the inclusion of the monitoring and analysis capabilities from PerformancePoint.

-          Microsoft has accelerated its release schedule for SQL Server to a new release every 24-36 months. This means that the BI platforms capability and functionality will be improved upon and expanded more rapidly than ever.

-          The next release of SQL Server, codenamed “Kilimanjaro” will be a BI specific upgrade to the platform. It will include powerful self-service analysis functionality exposed through Excel.

-          Microsoft is also working on integrating their DataAllegro aquisisition into SQL Server for large scale data warehouse deployments in the hundreds of terabytes range.

How will it affect PPS customers?

-          Customers running the core BI platform from Microsoft will be unaffected by this.

-          Customers using the Monitoring and Analytics parts of PPS will be unaffected by this.

-          Customers with SharePoint licenses can look forward to even richer BI functionality to be included in their existing licenses.

-          Customers that have implemented planning will need to look over their requirements and see if the SP3 version of the product will support their needs for the future.

-          Historically Microsoft has always compensated customers who have bought licenses that have been discontinued.

How will it affect partners?

-          Small specialized partners who placed a bet on PPS will obviously struggle with the investments they have made in the platform.

-          The whole ecosystem has to adjust their BI strategy and market offerings around Enterprise Performance Management, specifically the planning component.

-     This is an opportunity for 3^rd parties to supply the needed planning functionality. There are a number of partners who offered this  functionality before PPS who now are in a good position to offer this to the market.

Update: The offical press release on this from Microsoft will come later this afternoon. As far as I know it will include guidance on what customers can do with their PPS licenses. Since the cat is out of the bag I will keep this blog post open even though the news is not technically public.

• What will this mean for 3rd party solutions like Profitbase that also offer "packaged" BI solutions for verticals?
• Enterprise Cube deployment and customization is supposedly a Microsoft Consulting Services responsebility. What will this mean for the MS BI partner network? Does MCS have the resources for this Globally?

With the release of PPS in the very near future I thought I'd share some of the thoughts I've had presenting, integrating and getting to know the product.

A new world

For Microsoft's partners within the Business Intelligence area this will present a couple of challenges. To be able to deliver technology within Corporate Performance Management one needs not only expertise on how to configure and integrate the technology, but also an understanding of the customer's vertical, functional expertise and knowledge of how CPM is implemented on the business side. Especially with the planning piece of PPS we are touching on the very heart and soul of corporations: How they model their business, how they set up their goals and how they plan to reach those goals. This requires that partners understand things like how to set up a budgeting and forecasting model, how to do financial consolidations to get an accurate view of enterprise performance and the whole process of planning for an enterprise with forecasting and budgeting cycles, approval processes, etc. Partners with a business consulting arm will definatly have a great advantage provided they can succesfully mate their technological know how with their business expertise.

Tip of the iceberg

Integration

Getting closer to BI Nirvana

Just a quick heads-up that there are now two more CTP programs on connect related to PerformancePoint. One is for Management Reporter, a tool that lets users create boardroom quality financial reports in Excel or RDL format (as far as I can understand). After skimming through the documentation it looks like this is somewhat similar to report builder but with functional features for reporting on financial data. There also seems to be out of the box Dynamics AX integration. Pretty exciting stuff if you ask me, and another hint of the fact that Microsoft is broadening into the functional space of BI in addition of being a platform provider.

Update: Looks like this is the consolidation piece of PPS where you do your eliminations and so on. Not sure why the have taken it out of the "normal" PPS CTP.

The other CTP is the Data Integration Toolkit which i dont know much about, but it looks more like an accelerator for integrating GL data than an application.

Here are my thoughts on various topics that caught my attention at the MS BI Conference in Seattle.

Firstly I am not all that thrilled about the speeding up of the release cycles for SQL Server. Of course we will have to see how this plays out in reality. But if it means having to deal with a new core product every two or so years, with all that this entails in terms of certifications, migrations, customer confusion etc., I am more than happy to wait four years for the next release. The need for MS licencing revenue aside, I would much rather have solid SPs and the ability to buy an addin like the OfficeWriter as an extension to the license than have to deal with an "all new" product this often.

I was a bit surprised there was no mention of a MDM strategy of any kind from Microsoft. Especially considering the surfacing of recent MSDN articles on MDM. I am pretty sure we will hear more of this, as it is one of the key areas within BI where Microsoft lacks any real offering (that is, if you consider MDM a part of BI).

I like the scalability focus of the next version of SQL Server, codenamed Katmai. With data heavy technologies like RFID hitting the mainstream we will probably begin to see a demand for MUCH larger datawarehouses in the near future. Also with a bigger capacity SQL Server will be able to support large datawarehouses that have been the domain of appliance pureplays such as Teradata and Netezza. I just hope they scale all the components of the stack to the same level so that we wont end up with a classic as-good-as-its-weakest-link scenario.

Other than the Katmai release date and the OfficeWriter licensing there were disappointingly few announcements on Technology investments from Microsoft. I would have thought they would milk the BI Conference momentum for all its worth.

The data visualization / portal  3rd party vendors are clearly struggling. Apart from StrategyCompanion (I love the new version of their Analyzer product) none of them could tell me exactly why I should use their product over Reporting Services / Excel / PerformancePoint monitoring. In addition most of them are still hopelessy lost in the whole lets-make-everything-look-like-dashboard-widgets mantra. Is the market really so immature that they dont see the utter pointlessness of emulating real life objects to display data in meaningful ways?  I'm sure Stephen Few would go berserk in that Exhibitor hall.

Lastly: Where are all the 3rd party plugins to SQL Server 2005? Where are all the cool datamining algorithms? Where are all the datasource adapters for SSIS? With such an extensible platform as SQL Server it is pretty surprising that there are not more quality components being produced for SQL Server. All hail to the free stuff floating around (I use much of it) but isnt there a viable market for these things out there?