I posted a quick note about creating dynamic sets in ProClarity for SSAS 2005 at An Uncommon Approach: Business Intelligence.

http://uncommonbi.blogspot.com/2009/04/dynamic-sets-in-ms-sql-server-2005.html 

-Pete 

I am excited to announce that as of tomorrow I am leaving Microsoft to work for one of our consulting partners. It has been a great pleasure and one of the defining periods of my career to work on this team both at Knosys/ProClarity and here at Microsoft. For the last nine years I have had the opportunity to work with some truly incredible people within this team and with our customers and partners. I look forward to continuing those relations into the future.

I will continue to post about Microsoft Business Intelligence with more regularity at my new blog site An Uncommon Approach: Business Intelligence. Please stay tuned, I hope to see you there.

Thank you, 

Peter Sprague

Last week during a long flight, I had an opportunity to finally read Michael Lewis's book, Moneyball: The Art of Winning an Unfair Game. Commonly considered a sports or sports economics book, it is in fact an entertaining and complete case study in performance management. 

Lewis carefully chronicles how the front office of the Oakland A's uses performance management for competitive advantage with clear and definable monetary results. The management team uses performance metrics to better manage the amateur draft, to better value trade prospects and to even rethink baseball tactics such as the sacrifice bunt. In a classic example of monetizing knowledge General Manager, Billy Beane, and his front office identify a serious, market inefficiency for professional baseball talent and leverage it. The result is a string of playoff appearances by an Oakland team with a payroll a fraction of that spent by their serious competitors.

This book should be a must read for anyone interested in BI, anyone interested in baseball, and absolutely everyone interested in getting mind space for a BI project among a management team more interested in sports metaphors than technology spending. 

Greg Tatko and I will be in Bellevue May 14, 2008. Location and time TBD. I will post an update as that information becomes available.

Again, my presentation will focus on creating high quality assisted evaluations to better demonstrate the MS BI stack to your customers using their data. I will also be reviewing the architecture and discussing SSAS and PPS best practices. Feel free to contact me if you have questions or are interested in attending.

I hope to see you there.

Pete Sprague

Michael Patterson and I will be in Denver May 21, 2008. Location and time TBD. I will post an update as that information becomes available.

My presentation will focus on creating high quality assisted evaluations to better demonstrate the MS BI stack to your customers using their data. I will also be reviewing the architecture and discussing SSAS and PPS best practices. Feel free to contact me if you have questions or are interested in attending.

I hope to see you there.

 Pete Sprague

This is a nice list of resources (blogs, books, and training) that I recieved from Michael Patterson. I have been asked for PerformancePoint Server and ProClarity links a number of times.I allways end up recreating an approximatation of this list as a one off each time. So here is Michael's good work for your benefit.

Training

PerformancePoint Server Training Classes and Materials

http://www.microsoft.com/business/performancepoint/resources/training.aspx

ProClarity Training Classes

http://www.microsoft.com/bi/products/proclarity/proclarity-training.aspx

Books on PerformancePoint Server 2007

The Rational Guide to Monitoring and Analyzing with…

The Rational Guide to Planning with…

Business Intelligence with Microsoft Office PerformancePoint Server…  (recommended for deeper levels of detail)

Pro PerformancePoint Server 2007 (not released yet)

Technical Resources

Technical Forum (PPS Monitor and Analytics)

http://forums.microsoft.com/TechNet/ShowForum.aspx?ForumID=1872&SiteID=17

Technical Forum (PPS Planning)

http://forums.microsoft.com/TechNet/ShowForum.aspx?ForumID=1871&SiteID=17

TechNet PerformancePoint Server Site (Includes links for eval downloads)

http://technet.microsoft.com/en-us/office/performancepoint/default.aspx

PerformancePoint Server Development Portal

http://msdn2.microsoft.com/en-us/office/bb660518.aspx

PerformancePoint Server Download Center

http://www.microsoft.com/downloads/Browse.aspx?displaylang=en&productID=F52B1E9C-E169-4654-9A83-14A58A51C275

Performance Tuning and Capacity Planning for PPS

http://technet.microsoft.com/en-us/library/bb660539.aspx

PerformancePoint Monitoring Server Page

http://technet.microsoft.com/en-us/library/bb794650.aspx  (details under this topic include Hardware and Software Reqs)

Hardware reqs:  http://technet.microsoft.com/en-us/library/bb838792.aspx

Software reqs:  http://technet.microsoft.com/en-us/library/bb838753.aspx

PPS Team Blog (Monitor, Analysis)

http://blogs.msdn.com/performancepoint/default.aspx

Kevin White’s PPS Blog (Planning)

http://blogs.msdn.com/kevinwhite/default.aspx

Later I will post a few articles to summarize what is available from the different training resources. I also intend to post some reviews of the PPS books as I read them. Give me some feedback if this is particularly interesting to you. If I get enough feedback I may push the summaries up the priority stack. For now I will settle with posting these links.

Pete Sprague

Here is the final location and time information for the Calgary presentation:

Date: Monday April 7

Time: 8:30-1:00pm

Location:

Altius Center

2nd floor Conference room

500 – 4th Avenue SW

Calgary, AB

Hope to see you there.

Pete Sprague

One common theme I see in Analysis Services cubes as I visit partners and customers is large time dimensions. In the last year alone I have seen at least 10 cubes that have had time dimensions that extended past 2010. None of these cubes had facts associated with that range. I often seen cubes with time dimensions that start as early as 1901 and one that started with the year 0. There are often facts that appear to be associated with this historical period. Why is this a concern? Two reasons stand out, User Experience Complexity and Data Correctness.

User experience is an issue that I am going to address often on this blog. There are some great tools in the MS BI but their value to the end user is driven to a great deal by the quality of the cube. Large time dimensions are a simple but common example of poor cube design effecting user experience. When navigating the time dimension the user will see a large number of years that have no purpose. If they select any of these they receive a resulting empty row or column. While there is nothing "wrong" about this it does add unneeded clutter and complexity to the user experience. The goal should be clarity, only including members that have some relation to the data.

Does this mean a cube designer should only include members that directly link to a fact as a best practice? No. It means the dimension should include only those members that give the user context when understanding the data. It is certainly reasonable to include periods when there was no data captured because there was no business activity. Including members for holidays or weekends is a classic example where we include members without facts. Likewise including members for 1990 in a sales cube for a software firm founded in 2002 isn’t meaningful.

Data correctness is a more critical problem. When a fact in links to a nonsense date and you expose that to your business users you are advertising that you have a data problem. Do we really expect our business users to make decisions based on a cube with obvious data errors even when those errors are trivial? I am always surprised when I visit a customer and I see data in 1907 or 1906. Invariably I am told that the data warehouse is wrong. The suggestion is that the business users are "ok" because they understand that 1906 probably means 2006. This is one of the reasons that cube designers add or allow large time dimensions. This allows the cube to be processed with dirty data. It is a much better practice to limit the dimension members to a reasonable range of dates and deal with the bad data before proceeding. Better yet, a limited range of dimension dates AND an SSIS package that checks for out of range data as part of the ETL process.

Limiting the time dimension to a range of dates reasonable for your business use is just one simple step towards better time dimensions but the end result is certainly worth it. Over the next few weeks I will have a number of additional posts to discuss more ways to better handle time in SSAS cubes.

Remember, in the Microsoft Analytics environment the SSAS cube is a business document for all intent and purposes. Time and any other dimensions should be prepared and formatted with the care you would reserve for any end user documentation. 

Pete Sprague

I have been seeing a lot of the cubes recently that look very similar despite being created independently by different customers and partners across the US and Canada. They always looks something like this:

• 5 or 6 dimensions with few or no natural hierarchies
• 70 or more (often many many more...) attributes hierarchies, including the ever popular attribute [Customer Fax]
• 4 or more user defined time hierarchies
• Time attributes that contain all dates between 1901 and 2060
• 35+ measures
• Multiple measure groups
• Little relation to a specific user problem

Why is this a concern? These cubes make it more difficult for the business user navigate the data, and learn the tools. The complexity also makes it more difficult to understand the data that is returned. The biggest concern is that the users may misunderstand the data and base decisions on that understanding. One of the reasons for this is analytic and monitoring tools (as opposed to reporting) directly surface the cube metadata as part of the user interface (Excel, ProClarity and PerformancePoint Monitoring). Also, multiple measure groups are dangerous. Let me be clear. All of the features in MSAS 2005 are useful. They just aren't useful all of the time.

These problem cubes rarely help users support a series of decisions. Seldom do they help users analyze their data to get to an actionable step.

Let me introduce Vilfredo Pareto, an Italian economist who in 1906 observed that 20% of the population in Italy owned 80% of that country's wealth. He also noted that this ratio held true for other scientific and economic distributions. A hundred years and four times that number of self help books later we have the Pareto Principle. 80% of the value come from 20% of the resources. This principle is true for cube design with the additional observation that the extra complexity from the 80% of the resources cost far greater than the 20% value that those resources provide.

Next time you create a cube, strongly consider what design set will result in 80% functionality and stop there. In my experience these cubes often look similar to this:

• 6 to 10 dimensions all with 1 strong natural hierarchy
• 2 or 3 exposed attribute hierarchies
• 1 measure group
• 1 time hierarchy that only contains dates relevant to the period the cube data
• All metadata expressed in business friendly terms
• Strong relation to a business problem

Remember our goal, a cube that helps our users understand the data and supports further business action or decision. Consider this the first rule of cube design, Less is More.

Pete Sprague

I am going to be a presenter at a partner event next month (April 2008) in Calgary.  Please contact me or your PAM if you can be in the area and are interested in attending. The focus of the technical portion will be on SSAS, PPS and ProClarity. Here are the details:

BI Overview Event Details

Date – April 7^th 8:30-1:00PM

Location – downtown Calgary (tentative Westin Hotel) based on number of attendees

Agenda

8:30 welcome

8:45-10:45 BI Sales Training

·       BI Product Overview

o   Microsoft BI Vision and Product Stack

o   What is PPS?

o   Why do customers want PPS?

·       Co-Engagement

o   Best practices for working with Microsoft BI sales team to present and deliver PPS solutions to customers

·       PPS Resources

o   Review of documentation, sites, training options, forums to support your BI Consulting Practice

10:45-11:00 break

11:00-1:00pm BI Technical Overview

·          Whiteboard Architecture overview

·          Demos

·          Solution delivery

o   Planning your BI solution

o   Pitfalls and traps, Best Practices

·          Discussion Q&A

1:00 PM session end

This blog is intended to be a discusion of some common challenges implementing PerformancePoint and other Microsoft BI that I and my coworkers encounter in the field. Feel free to send suggestions for topics that you may be interested in or comments in general.

Pete Sprague