I especially look forward to TDWI’s World Conference on data warehousing and business intelligence when it is held in San Diego. This is one of my favorite towns. It has great sailing, great jazz,
great food, and great beaches. Add good friends and a stimulating program to the mix. What’s not to like?
Glenn Gutwillig’s morning session titled “Conceive It: Formulating an Enterprise Information Strategy” seemed to fit right in with my recent ruminations about data warehousing drift. Here we are
a decade plus into this business of data warehousing which got its start by taking a strategic view of decision support, and I still run into data warehousing programs whose foci have drifted into
the minutiae of technology. This doesn’t count the many efforts that companies tag with the name data warehouse that started out as tactical data marts, and still are. When you combine these
situations with executives that think that the new data warehouse should do everything, including butter their toast for them in the morning, we could do well to create an enterprise information
management (EIM) strategy that would put data warehousing into proper perspective. Gutwillig did that and more.
Gutwillig showed how an EIM strategy integrates information across business functions and geographic units. It puts the ERP, other operational systems, and the data warehouse in perspective for the
executive suite. The really big payoffs for EIM come from in the areas of the enhanced ability for the organization to respond to new revenue opportunities, meet regulatory compliance, identify
cost reduction opportunities, and manage risk, i.e. EIM includes information mappings, which makes adding increments to the data warehouse that much easier.
After explaining the value of EIM, Gutwillig spent the rest of the morning going over the EIM maturity model and an EIM architecture.
That afternoon, a old friend and I went up to Old Town and our favorite haunt, the Coyote Café. The weather was superb, and the food was excellent. We spent the time catching up on each
other’s recent data warehousing experiences. Afterwards we headed back to catch Anthony Politano’s peer networking session on “Talking with Business Executives on Their Terms.”
Dave Wells kicked off the week with a great keynote, “BI and Business/IT Relationships: Catalyst or Cataclysm.” He described how business and IT use different value chains. Each group starts at a
different point, business with business drivers and IT with data. Each group then proceeds to move towards producing value for the organization without regard to the other’s perspective. This is a
prescription for mismatches. He suggested that we need to integrate these value chains, but that just saying this would not make it happen.
Wells then showed how diverse areas in business and IT view business intelligence much differently. He went on to list the major business and IT roles that impact the delivery of BI and suggested
ways to integrate many of these roles into a single BI organization. He likened the business side of the house to the blue camp and the IT side as the red. However, the new organization that he
proposed would be made up of “purple people” who would be able to bridge business and IT perspectives and deliver a highly effective BI environment.
Wells made me think back to the most effective group of “purple people” that I had ever worked with. It was based on the old Information Center concept that IBM had been pushing over 10 years
ago. I also thought back on how many Information Centers had come to focus more on technology and less on business. However, this one group was different. It truly kept its eye on the challenges to
the business while effectively marshaling its technology to meet those challenges. We had many successes, including some netting nine figure revenue gains.
After the keynote I pressed on to Anthony Politano’s “CPM Bootcamp: Introduction to CPM.” Politano claimed that the overarching driver for BI is corporate performance management. He made his
case through showing how growing market acceptance reflects the linkage between BI and CPM. He then broke down the components of CPM and covered best practices on its delivery. He put metrics in
perspective and showed how the balanced scorecard is a vital component of CPM. Politano reminded us that financials, although essential, are still trailing indicators. After lunch, Politano took us
through several exercises to show us how we could use analysis chaining to work back from critical lagging indicators to find the important leading indicators and create an action matrix. This was
one cool session.
That evening, I sat in on Cristel Ryan’s night school session, “Strategies for Managing the ‘Accidental Warehouse’ Environment”. I don’t know, maybe it’s because I am increasingly called in
as the “fixer” to help out data warehousing programs that got started on the wrong foot, but I think that Cristel was really on to the way most data warehousing programs get started. She did a
good job of laying out the elements of a sound data warehousing program and suggested ways to move to that spot. It was all about data warehousing basics, but for companies which had not started
out with a clean slate.
That morning, I felt the awesome weight of two days full of great content. I needed to step back and decompress. Fortunately Seaforth Boat Rentals was only a few short steps from the hotel, and San
Diego Bay was calling out to me. I, in turn, called up a buddy that I had met at a didjeridu workshop earlier this past summer and we went sailing. The weather was superb. The wind was outstanding.
We sailed the afternoon away talking about the didjeridu, native flutes, and life.
Fresh from a great day on the bay, I was ready to take on the “Unique Challenges in Financial Services” session with Jonathan Geiger that next morning. Geiger did an exquisite job of bringing
data warehousing strategies home to this particular sector. Wherever he could, Geiger, dressed in purple, took technical concepts and gave them financial services applications. It was a true
demonstration of how to bridge the IT/business divide that Dave Wells had driven home Monday morning.
I dropped by the vendor show during lunch and talked with a couple of companies that are on the rise.
Datallegro markets two database machines that they refer to as data warehouse appliances. The name is a good one. They are about the size of a small refrigerator and come in two basic
configurations. The P3 is a 3 TB machine designed for high performance, and the C25 is a high capacity, 25 TB box. Both are designed with high availability in mind. These appliances can be strung
together to provide high performance for relatively current data and the capacity to serve the need for deep history. Both are targeted at mid sized companies.
Datallegro’s sales strategy was refreshingly straight-forward. Their pitch was easy to understand, and prices for their machines ($450,000 ea) were posted at their booth. They clearly want to
focus on those customers that need exactly what Datallegro offers. I think that this will be a company to watch.
I also stopped at the e2e booth to talk to an up and coming consulting firm. There are a growing number of such firms who specialize in data warehousing and related areas such as balanced scorecard
and knowledge management. E2e is out in front with their focus on BINI or BI network infrastructure which focuses on the infrastructure to span operational and decision support sources of
That evening I caught a night school session on “Performance Management: Keys to Successful Solutions” that was developed by Chip Goodman and Bennett Indart. The session dealt with leveraging CPM
work through analytical applications. It was a good follow-on to Anthony Politano’s CPM Bootcamp.
After night school a few of us went out to catch some sounds at Croce’s Jazz Bar. Jorge Camberos and his quartet gave us a nice mix of Latin and straight-ahead jazz. The food was great. What a
The conference had shaped up to be one of the best, and I decided to finish with James Thomann’s session, “TDWI Data Analysis and Design Basics for IT Professionals.” Right off the bat, he
addressed one of my pet peeves. My formal training was in economics, and data analysis, to me, was always about testing a hypothesis by creating a model and applying it to relevant sets of data. Of
course, when I joined IT, data analysis meant data modeling. The words were similar, but the intent was something else again. Thomann suggested that we sort the term into two applications, business
data analysis and information systems data analysis. Good move. Thomann also suggested some other strategies that I thought were fairly unique, but genuinely helpful.
Thomann linked logical data models for applications, an integrated enterprise logical model, business metrics models, logical dimensional models, the conformed bus matrix, mining data models, and
analytical designs all to the same logical level of a variant of the Zachman Framework. He then went on to describe a structural level that fit between logical models and physical database designs.
Thomann put the star schema into this category. This artifice allows the information systems data analyst to create a dimensional model with normalized dimensions (snowflake) to insure that the
model she/he is creating is complete and flexible. That analyst could then take the dimensional model, a logical construct, to the structural level by collapsing the dimensions into a star schema,
as appropriate. The star schema focuses on ease of user understanding and performance. This also allows the analyst some wiggle room to implement the star schema, if that’s what’s indicated, in
the physical design notation of the database of choice. Got that? Well maybe you had to be there, but trust me, this course was pretty slick, and TDWI is planning to offer it on a regular basis.
After class I hopped on a plane back to the Emerald City. It was a great conference and a mellow flight. See you in Orlando in a couple of months.