Eye on TDWI – Orlando – November 2005

See also: TDWI Orlando Exhibit Hall Highlights

Anticipation

I was wondering what to expect from the weather as I was planning my trip to TDWI’s World Conference in Florida. Last year, I flew into Orlando just after hurricane Charley had blown through. This
year, Rita and Wilma hammered the southern part of the state. However, fortune smiled and the flight to Orlando was as smooth as glass. As we landed, the sun was setting over a lush green landscape
dotted with hundreds of lakes, streams, and marshes. Kids and their parents eagerly looked forward to Disney World, and I looked forward to another great TDWI conference. I was also looking forward
to a kayak trip in central Florida, but more about that later.

Sunday

Technology updates are always high on my list of priorities for every conference that I attend, so I started off with Michael Gonzales’s “Technology Update Live!” TDWI does a good job of keeping
managers, architects, and consultants up to speed on changes in technology. While most conferences depend upon vendor shows for this function, TDWI goes one step further and schedules classes based
on product overviews and comparisons that don’t pull any punches. While I never rely on a single source for this sort of background, I find that the classes that TDWI offers are quite valuable as
a starting point. I especially like Gonzales’s classes because he frequently demonstrates a point by giving the class a short demo on the product he is talking about. His breadth of experience is
prodigious.

After giving us a run-down on recent acquisitions and progress in integrating acquired products, Gonzales pointed out some lower cost technologies that BI architects often overlook. He is
passionate about getting high returns on BI investments.

All of the major database vendors now feature in-database data mining. However, Gonzales often finds that data mining is done outside of the data warehouse. In one such case, he observed that the
only value that the data warehouse offered was a run of mailing labels for the resulting campaign. He noted that all of the major data mining tools can now use in-database capabilities through PMML
(Predictive Model Markup Language). In the long run, Gonzales sees data mining as becoming another BI tool and exhorted us to get more return on our investments by including in-database data mining
capabilities in our data warehouse plans.

Gonzales also demonstrated multi-dimensional visualization and spatial analysis technologies. In both cases, we were able to get a sense of solutions to business questions that would not have been
apparent with basic pie charts or list reports. Granted, in some instances, advanced visualization tools such as PolyVista, will be of more value to analysts than to the casual user. However, they
are great tools for solving vexing business questions. Monday

Wayne Eckerson started off day with his keynote, “Performance Dashboards: Measuring, Monitoring, and Managing Your Business.” Eckerson related a number of executive perceptions about data
warehousing and BI, i.e. too slow, too complex, too rigid, too expensive. He posed these perceptions next to what most of us recognize as common executive wants – system aligned with the business,
single version of the truth, delivers data that can make a difference, and delivers answers quickly. Eckerson then suggested that, in these situations, the performance dashboard was the logical
extension of BI.

Eckerson pointed out that performance dashboards were the result of evolutionary forces in our field. He suggested a BI architecture that includes monitoring (dashboards), analysis (dimensions,
advanced visualization), reporting (detail), and planning (plans, models, and forecasts). He also ran down some of the characteristics of good dashboards and outlined how they would be deployed
with different interfaces at the executive, mid-management, and operational levels. When properly deployed and supported, performance dashboards cut down on the “noise” and result in
organizations obtaining their goals more directly than they otherwise would.

After the keynote, I sat in on Colin White’s class, “Business Intelligence for the Rest of Us: Solutions for the Medium-Sized Business.” Since this was an area of focus for my consulting
practice, I was especially interested in what he had to say.

White started off by laying out a basic BI framework and giving us some common definitions of SMB (small to medium-sized business). IDC queues off of the number of employees: small less than 100,
and medium between 100 and 1000. Microsoft uses number of PCs: medium-sized businesses have between 25 and 500 PCs. Others define the SMB space in terms of revenue. White went on to say that the
SMB market is greater than 80% of US GNP and that SMBs collectively will spend about $360 billion on IT in 2005. Although I have heard some BI vendors say that their focus is on the Fortune 1000,
i.e., not the SMB market, this is a chunk of change to walk away from! However, Microsoft has the SMB squarely in its sights, and White appropriately went over its new offerings in some detail.

Microsoft’s SQL Server 2005 has a number of components that fit within the BI framework, including, of course, its relational data warehouse engine, ETL, a multidimensional database, date mining,
managed reporting, ad hoc query and analysis, plus development and management tools. White pointed out that Office 12 will include extended BI functionality and that SharePoint, Microsoft’s
portal, will provide Reporting Services integration. Although this list is impressive, and nothing approaches it in the open market space, many of Microsoft’s reporting capabilities are relatively
new. This may give other, established BI vendors time to rethink their strategies for this market. However, they will need to sharpen their pencils and be quick about it.

That evening, Sid Adelman presented a short class on data strategy. I see it all the time. Data just happens. Sure there are minor skirmishes that focus on bits and pieces that we bring into our
lives. These skirmishes are called applications. Once in a while we have an all out integration war. However, those efforts don’t always go the way we would like, so Adelman’s session struck a
resonant chord with me.

In 90 minutes, Adelman covered components, motivations, the critical factors of success, and how to implement a successful data strategy. The component list covered a fair amount of ground – from
standards, database systems, modeling, data quality, security, sharing, BI, integration, applications, to service level agreements. Adelman suggested ways to sell the value of bringing the
management of these components under one tent, and then organizing to put your strategy in place.

Given the size of the topic, Adelman did an admirable job of tying everything together. However, I understand that he will offer this class in a longer daytime format in the future. It will be
worth it.

Tuesday

I decided to go for an early morning swim. The Omni Orlando Resort at Championsgate was first class in every way. The restaurants were great. The hotel was surrounded by two golf courses and
wetlands, plus we had our choice of three swimming areas. One was an “endless river” that flowed through the grounds. Given that it had a pleasant current and connected back to itself, swimming
it was pretty easy. Any time you wanted, you could quit kicking, lay back, and just go with the flow – kinda like life should be.

Later in the morning, I went down to the vendor show to see what was cooking. I wanted to spend some with quality time with a few vendors and get beyond the cursory fly-by that I think too many of
us often get in the habit of doing at these events. There were a couple of interesting finds.

Business Objects has been one of the leaders in the BI space for a number of years. They have built their business on a strong user interface, managed reporting capabilities, and very accessible
analytics. However, when Business Objects bought Crystal Reports, a lot of us were concerned about how Crystal’s management capabilities for enterprise reporting would be integrated with Business
Objects ease of use. I was impressed to find that in their new release, users can develop Business Objects reports and publish them to Crystal’s reporting environment – pretty cool.

A few years ago, anybody suggesting that you could buy a data warehouse was suspect. Custom operational systems don’t lend themselves to packaged data warehousing solutions. However, off-the-shelf
ERP systems have become ubiquitous. While their companion data warehousing solutions have not been as strong as they could be, DecisionPoint Software has forcefully moved into this space.
DecisionPoint offers pre-built, auto configured data warehouse components for ERP finance, manufacturing, distribution, and HR packages. These components include ETL and ERP specific sourcing
tools, 3rd normal form and star schema data models (your choice), and DecisionCast which is a ad hoc query and analysis tool. The promise of rapid data warehouse development is DecisionPoint’s
primary selling point. However, for those involved with mergers and acquisitions, DecisionPoint’s ability to integrate different financial systems is an added benefit.

That evening I stopped by the SAS hospitality suite. SAS really knows how to through a party. It was Star Wars all over. Episode III, which was being released that day on DVD, was showing on the
big screen – Darth Vader, Princess Leia, and various storm troopers roamed the floor – and everybody got a light saber. The food was great and games fun. There was no vendor pitch. However, their
booth in the vendor area was pretty busy the following day.

Wednesday

Since Wayne Eckerson had kicked off the week with his keynote on performance dashboards, I decided to follow up with Stephen Few’s session on “Dashboard Design for Immediate Insight.” Few is one
of the leading gurus on design for business communication and his classes are always enlightening.

Few started off with showing us a number of dashboard samples. Most were from vendor Web sites. Theses samples turned out to be pretty good setups for Few’s “13 common mistakes in dashboard
design.” Few’s list:

  1. Exceeding the boundaries of a single screen
  2. Supplying inadequate context for the data
  3. Displaying excessive detail or precision (oh yeah!)
  4. Expressing a measure indirectly
  5. Choosing inappropriate display mechanisms
  6. Introducing meaningless variety
  7. Using poorly designed display mechanisms
  8. Encoding quantitative data inaccurately
  9. Arranging the data poorly
  10. Ineffectively highlighting what’s important
  11. Cluttering it with useless decoration
  12. Misusing or overusing color
  13. Designing an unappealing visual display

Few went on to give us the basic characteristics and some examples of good design. Although I felt that his list of mistakes was spot-on, and his suggestions for good design were sound, in the end,
there will likely be design elements that are dictated by the boss. At least if one of Few’s students needs to dress up the “pig,” he or she can do it with taste.

The night school schedule included an intriguing topic, “Operational Business Intelligence: It’s About Time.” Awhile back, if you were to mention BI and operational data in the same breath, you
would immediately spawn a fight with the vestal virgins of the data center. Although that is still true in some shops, Michael Corcoran gave us a number of reasons why the time has come for
operational BI. Machines are faster and BI software is scaling up. Information integration capabilities are also growing into the task. (However, I wouldn’t assume that all BI software features
these capabilities. Ask your vendor.) He continued with several case studies with eye-popping numbers, i.e. ten’s of thousands of users of BI applications going against real-time, operational
data. With the right kind of integration with the data warehouse, operational BI may address some of the information needs that real-time data warehousing is trying to address – with less cost and
development time.

Thursday

Philip Russom started off the morning with his keynote session: “Business Intelligence Trends in 2006.” Russom asserted that the mega trends in BI would center on size, speed, interoperability,
and economics. The fact that there weren’t any real surprises here indicated that BI has arrived and now the boss wants it to do unnatural acts. Some points to ponder: is batch ETL dead? Will it
be replaced or at least significantly supplemented with EII (Enterprise Information Integration) and messaging technology? As mergers and acquisitions continue to pose interoperability challenges
when marrying two or more mature data warehouses, will EII get an additional boost? Will hand-coded ETL finally be recognized for the true cost that it poses? Will lower cost ETL products may
finally take over hand-coding in many shops? Could be.

After the keynote, I dropped in on Rajeev Rawat’s class, “Build Compliance, Deliver Competitive Advantage.” Rawat has made a specialty of studying the implications of laws such as Sarbanes-Oxley
which carry extremely stiff penalties for a manager’s ignorance or misrepresentation of the material position of his/her company. The focus of the class was on building systems that will not only
meet compliance reporting requirements, but be used as the basis of an agile infrastructure that can be used to enhance the business. In fact, many of the requirements for providing sound
information systems for compliance overlap good data warehousing practices. The key is being intentional about building for both of these ends.

Friday

It was an intense, but rewarding week. However, I was looking forward to getting away early for the weekend.

Several friends and I had planned a three day kayaking trip to central Florida. We hit the Silver River on the first day. The weather was great and we saw a couple of ‘gators right off. Next we
came across several monkeys that had taken up residence in the swamp after their engagement as extras in “Tarzan” had ended. The next day we paddled the Chassahowitzka and explored a number of
remote side channels and exquisite springs. We wrapped up with a tour of the Crystal and found the manatee that we went looking for. Central Florida may often get overlooked, but it has some great
rivers and exotic scenery. This was an outstanding finish to a great week.

See you in Vegas.

Share this post

Tim Feetham

Tim Feetham

Tim is an independent consultant who specializes in data warehousing for small to medium sized businesses. He has worked in sectors ranging from travel, health care, finance and software, to higher education. He helped design the Data Resource Management Certificate at the University of Washington and has taught in that program for more than 10 years. Feetham is also a former senior research analyst for TDWI. He continues to contribute to TDWI publications and events.

scroll to top