Big Payoffs in Las Vegas
Even considering the bright lights and lavish shows, this town always generates more energy than it consumes. This week, the confluence of Chinese New Year, the NBA All Star Game, the Magic
Wholesale Clothing Expo, and TDWI really generated the heat in ‘Vegas. With each conference, the program at TDWI manages to give newbies to BI and data warehousing a solid grounding, and the old
salts food for thought. Although I have been around long enough to experience most of subjects presented, I was looking forward to seeing new thinking about some of the consistently problematic
issues facing our discipline. I wasn’t disappointed. At the end of the week, I felt a deep appreciation for the efforts that some of the best of our thought-leaders have put in to their
presentations, and I felt re-energized.
Metadata has to be at or near the top of lists as the chief pain for most data warehousing teams. Jonathan Geiger’s session, “Let’s Stop Calling It Metadata: It’s about Managing Information,”
was spot-on in sorting through the issue. He started with the obvious, but usually overlooked perspective, that metadata, as data, was an asset. Although there might not be any good way to value
it, most would agree that the value of metadata is directly related to its organization, accessibility, and usefulness. Geiger then suggested that teams implement an asset management system for
metadata. He went on to describe several different approaches to building, or buying that system, depending upon the inclinations of the enterprise and whether it had a centralized or decentralized
approach towards data management. Although he used the Corporate Information Factory framework to describe the different sources and uses of metadata, the way in which he did it left room for those
using either a hub and spoke or a bus architecture for their data warehouse. When the session ended we had a step-by-step project plan that addressed intermediate targets, models, architecture,
strategy considerations, and resources. Geiger’s session alone was worth the trip.
That evening, TDWI and a number of sponsors hosted a reception. It was a good way to kick off the week and connect with old friends.
Jonathan Wu’s keynote, “Applying IT Governance to BI Solutions to Ensure Lasting Value,” took a hard-nosed business approach to managing BI. Given that we were in ‘Vegas, Wu pointed out that BI
success wasn’t just luck. He gave us a brief overview of how IT governance is about aligning “the priorities, process, projects, and people required to run IT like a business.” When he applied
this approach to BI, Wu stressed that organizations would end up getting a lot more bang for their BI buck. One of the points that he made about measuring BI usage is one that resonated with me
personally. I have seen basic monitoring build a very strong appreciation for BI when a nine figure business success was tied back to data warehouse use. Indeed, most of Wu’s slides appropriately
contained money-oriented graphics.
The next session, Claudia Imhoff’s, “Governance for Business Intelligence Programs,” turned out to be a great follow-on to Wu’s keynote. Imhoff co-authored “The Corporate Information Factory”
and used that framework to show the connections between enterprise data sources, data integration, and data use through structures and application that support BI. Imhoff also introduced us to an
expanded version of the Corporate Information Factory, CIFe, which included a BI governance wrapper with several different components, i.e. governance, infrastructure management, center of
excellence, quality management, application management, and metadata management. She stressed that BI governance should not be considered a subset of IT or Data Governance, but stand on its own.
This generated a lively discussion. My classmates were experienced, quite bright, and their comments added significantly to the session. However, Imhoff gave some persuasive arguments for her case.
One example that she gave was that best practices in BI have demonstrated the importance managing BI as a program instead of a series of disconnected projects. She went on to give us some sound
organizational strategies and, like Wu earlier in the day, stressed the importance of measuring BI use.
Given that I had been exposed to some great thinking over the last couple of days, I decided to take a break and take a day trip up to Zion National Park.
The territory north of Las Vegas is sagebrush and jack rabbit country. The land is flat with a few hills in the distance. However, this soon gives way to a rainbow of colors on higher hills and
snow in the mountains. The highway links up with the Virgin River, and I follow it into the canyons of Zion. I go nuts as a drive-by tourist. I snap photos everywhere I can in the short time I
have. The scenery is indescribable. I stop in Springdale for a late lunch, chat with some locals, and head back for the bright lights of the strip. What a trip!
I arrived back at TDWI in time to hit the SAS Hospitality Suite. SAS has brought in the Taikoza Drum Troupe to put on a show, and they do indeed put on a show. I am a fan of Taiko Drums and never
miss a chance to hear different groups. Taikoza is top notch. I have a cup of sake, chat with new friends, and call it a night. Thank you, SAS.
From time to time I have found myself in the role of mentor. I decided to check out Jennifer Hay’s session on “Building a BI Career: A Personal Growth Plan.” I had to see if it would be
something that I wanted to recommend to folks starting out in this discipline.
Hay knew how to break the ice. By way of introduction, she started out by asking what we wanted to be, then showed off her tutu (she wanted to be a ballerina early-on), and then donned her tiara
and feather boa (later, she wanted to become an actress). Her point was that we change paths as we progress through our careers. She said that we will probably change our career goals several times
in our lives and that for each change we should identify measures that will help us on our paths to achieving those goals. Hay led us through a BI Skills assessment exercise that included a Gap
analysis of our knowledge, experience, and skill level – versus our interest in various BI roles. After that, she showed us how to square those roles with our own personality types by reviewing
Meyers Briggs descriptions. Hay finished up by giving us a run-down on career development resources including TDWI On-Track and Certified Business Intelligence Professional programs, plus TDWI’s
Report on Salary, Roles, and Responsibilities. I’d recommend her class whether you are planning on working your way up through the BI discipline or you are in the position of mentoring others.
After Hay’s class, I dropped in on the vendor area and spent some time with the folks at Teleran and iWay. Teleran’s suite of products operates like an SQL database firewall and monitors user
interactions with the data warehouse. Their suite also includes a query manager that stops potentially runaway queries before they hit the data. iWay also has similar capabilities that plug into
their data integration software. However, given the focus on BI governance and usage measurement earlier in the week, I was particularly interested in checking out the data usage monitoring
capabilities of these vendors. I had built a very crude data monitor early in my career, and it turned out to pay off handsomely when I was able to tie data warehouse use to a multi-million dollar
business success. Without an objective way to link these successes to data warehouse usage, the data warehouse becomes just another cost center with hearsay benefits. Teleran and iWay are companies
that I would definitely look into if I were putting BI governance into practice.
That afternoon, I sat in on a class taught by Bill Belknap & Anthony Politano titled “Diving for Success: Using Data Warehousing as a Career Springboard”. They delightfully reminded us that,
as outsourcing grows, there will be fewer technical jobs available in IT. However, that trend will also include an increasing demand for managers of technology. They gave us some good tips on how
to move up through this hour-glass shaped model with technical inventory maximization and networking techniques plus tips on building follow-up plans. I was impressed.
Just about the time when you think you have a good handle on BI along comes a guy like Jonathan Koomey to disabuse you of your sense of comfort. Koomey’s keynote, “Sorry, Wrong Number: Real-Life
Lessons for the Responsible Use of Data and Analysis in Decision Making,” showed how we can really screw up our decision processes by careless use of statistics. He cited a Forbes article about
the internet being responsible for 8% of all electricity use and its projected growth would run us out of house and home in the next 10-20 years. He traced the numbers from this article through
dozens of mainstream press articles. However, the numbers turned out to be wrong. Koomey stressed that there were lessons here for BI. He suggested that we focus too much on the data and not enough
on how the data is used. He urged us to understand the decision making process in our organizations, think critically, and assess the validity of inferences and arguments.
I walked out of the keynote thinking that the BI/IT partnership has to go beyond the building stage. It needs to be on-going when questions arise, models are developed, and the data is put to use.
I also thought back to the biggest data warehousing successes that I had been a part of, and I realized that Koomey was right on.
The relationship between IT and business is perhaps no more problematic than in the development of performance management. I have seen that relationship work well, and I have seen it not work at
all. When it is bad it seems to be horrid. Nancy Williams tackled the subject in her class “Balanced Scorecards in the Business Centric BI Architecture”, so I dropped in.
Williams reviewed performance management and specifically Balanced Scorecards from a business and then an IT perspective. There were conflicts, with each side thinking that it knew the subject best
and that the other side needed to just get on-board and follow. It was clear that a combination of views and good cooperation would likely be more successful. She suggested an action plan with a
Strategic Feedback Loop (Learning, Communicating, Changing) for iterative development. She also cited several critical success factors for the Balanced Scorecard Program including – securing
executive sponsorship and involvement (i.e. setting and reviewing KPIs for enterprise), treating it as an on-going program – not a standalone project, developing an integrated data architecture,
ensuring that consistent definitions be used throughout the program, and gaining support throughout the organizational.
That night a group of us went to Rao’s for dinner. The food was great, and the company was excellent. We drank a toast to another successful week and planned on meeting in Boston in May for the
next TDWI conference. See you there.