Blog: Claudia Imhoff Subscribe to this blog's RSS feed!

Claudia Imhoff

Welcome to my blog.

This is another means for me to communicate, educate and participate within the Business Intelligence industry. It is a perfect forum for airing opinions, thoughts, vendor and client updates, problems and questions. To maximize the blog's value, it must be a participative venue. This means I will look forward to hearing from you often, since your input is vital to the blog's success. All I ask is that you treat me, the blog, and everyone who uses it with respect.

So...check it out every week to see what is new and exciting in our ever changing BI world.

About the author >

A thought leader, visionary, and practitioner, Claudia Imhoff, Ph.D., is an internationally recognized expert on analytics, business intelligence, and the architectures to support these initiatives. Dr. Imhoff has co-authored five books on these subjects and writes articles (totaling more than 150) for technical and business magazines.

She is also the Founder of the Boulder BI Brain Trust, a consortium of independent analysts and consultants (www.BBBT.us). You can follow them on Twitter at #BBBT

Editor's Note:
More articles and resources are available in Claudia's BeyeNETWORK Expert Channel. Be sure to visit today!

June 2005 Archives

Yet another high tech "underdog" has decided to take on its Goliath. According to USA Today , Advanced Micro Devices filed an antiturst complaint against its longtime rival, Intel, claiming that Intel used its huge size to coerce customers into choosing Intel's chips over AMD's. This case may finally shed some light on the murky and sometimes less-than-ethical world of chip sales...

This is not new information -- AMD has long complained about the perks that Intel's offers customers to get them to sign up such as big discounts on large orders and the substantial "market development funds" for featuring Intel in ads (you know -- the "Intel Inside logo and earcon).

AMD says these are strong arm tactics giving Intel an unfair advantage. Intel says they are just smart business practices...

In the long run, some analysts think that the lawsuit will not harm Intel and, in fact, may actually give both companies a boost -- all this invaluable publicity for them, don't you know.

Of course, Intel is not taking this lying down. Their President and CEO, Paul Otellini, came back with his own shots, stating the company would not change its business practices and expects any court decision to be resolved in its favor.

"Intel has always respected the laws of the countries in which we operate," Otellini said in a statement. "We compete aggressively and fairly to deliver the best value to consumers. This will not change."

Meanwhile, AMD has not stopped its attacks. The company began running a full-page ad in newspapers today giving the reasons for its antitrust lawsuit and to issue a call to action. The ad, appearing in newspapers from The New York Times to Capitol Hill's Roll Call, expands AMD's legal fight into the realms of public relations and public policy.

Looks like it's going to get a lot nastier before this is all over. No trial date has been set yet.

Yours in BI Success,

CLaudia Imhoff


Posted June 30, 2005 9:23 AM
Permalink | No Comments |

There is one task in virtually every BI project that seems to cause the most surprises and to be the biggest schedule buster -- that is estimating how long the ETL or data integration process will take. While the reasons for this are numerous, it seems that a lack of business understanding leads the pack.

My friend, James Schardt, Chief Technologist for Advanced Concepts Center, sent me a suggestion for how he mitigates this serious problem in his DW projects.

From James Schardt:

"Here's the situation: A data model has been developed and the appropriate physical schemas for the DW have been designed. (This may or may not have been done using gathered documented requirements for the DW). The design team has some idea of which source system data records will be used to populate the DW. Now the design team must develop a set of transformations that must properly clean, transform, and/or merge the data into the right location in the DW. The team discovers they simply do not have an understanding of the business to correctly specify many of the more complex transformations. They make assumptions. They get it wrong. The users are not happy with what they get.

When I talk about gathering and documenting requirements for the DW, I mention the "Reverse Engineering" technique. During the Requirements and Analysis phases of DW development, have a couple of developers reverse engineer the important source systems. The result is a conceptual level ER or UML class diagram that forms the basis of bottom-up domain modeling. (This is then combined with top-down domain modeling to form the basis for subsequent data warehouse and ODS design).

While the development team is reverse engineering source systems, they will uncover problems with the data and the structure of the data. When they combine multiple reverse engineered models into an integrated bottom-up domain model, they will uncover many other problems with the operational data sources. But, these problems are discovered during analysis and before starting the design. The DW team has an opportunity to resolve these issues with users before they begin design work. Users have input and more buy-in to what the DW will provide them. We get the users to specify their requirements for clean and integrated data using business rules that make sense to them.

By performing reverse engineering early in the process the developers should be able to perform the ETL development without a lot of surprises."

Please add your comments to James' suggestion or make your own suggestions to this universal BI project schedule killer.

Yours in BI success!

Claudia


Posted June 24, 2005 1:29 PM
Permalink | 2 Comments |

With the emphasis on IT's critical role in compliance monitoring and reporting, as well as in supporting such initiatives as CRM and Business Performance Monitoring (BPM), perhaps it is finally time to recognize that the CIO must report directly to the CEO - not the CFO as is traditional. The importance of this direct reporting role cannot be stressed enough but, unfortunately, many CIOs have still not made that case in their organizations. A recent article in ComputerWorld lists 8 reasons why a CIO should report to the CEO. Here are some of these reasons.

1. Most companies claim that their IT systems are what differentiate them from their competitors. Since the CEO is the chief strategist (at least, he or she should be), then that person should directly oversee and instruct the individual responsible for aligning the technologies behind that strategy.
2. IT is thought to be an enterprise resource. If anyone, other than the CEO, is in charge of the CIO, then it can be assumed that the technological agenda will be biased or influenced by the goals and objectives of that particular executive. For example, if the CIO reports to the CFO, it’s a sure bet that financial systems will get more attention than, say, marketing or HR ones. It is mandatory that IT develop business applications that are critical to the overall well-being of the enterprise, not those favored by one senior executive.
3. Since most strategic IT implementations like CRM, BI and BPM take long periods of time before their full benefits can be felt, it is necessary that the CIO have an orientation toward the future. He or she must maintain a long-term vision of these future benefits and this can only come from the CEO.
4. Then there are the rising costs of IT. Without direct access to the CEO, the CIO is in a weak position to defend the needed budgetary funding. Without direct input from the CIO, the CEO won’t understand that IT is an investment, not just a cost center to be mitigated.
5. We have long preached that BI and other sweeping initiatives like data quality, compliance, CRM, etc., should be enterprise initiatives. Unless IT is recognized as the strategic driver behind the enterprise, it will be very difficult to get input from all parts of the enterprise in terms of IT’s priorities. If IT reports directly to the CEO, then this becomes very clear to other C-level executives.
6. The IT landscape is littered with failed projects, missed expectations, deadlines, and budgets, and critical shortages in personnel and resources. This is exacerbated by the constant turnover of CIO’s (according to the article, the average CIO tenure is 18 to 36 months). Unless the CIO has the ear of the CEO, he or she will always be a convenient scapegoat when things go badly for IT.

So, given these reasons for the CIO to report directly to the CEO, why are we seeing a rise in CIOs reporting to CFOs? I suppose we could come up with just as many reasons for this trend but it seems to me that the biggest impediment to the CIO becoming one of the CEO’s “inner circle” stems from the CIO’s inability to speak in terms that the business understands. CIOs must stop the technobabble they are so fond of and become business-oriented. Then the CEO will enjoy me


Posted June 21, 2005 7:20 PM
Permalink | No Comments |

I received an email the other day from someone working for a large international financial institution. I thought his comments were worth publishing with the hopes that you have some ideas on how a global company can truly get to visionary "single version of the truth" when it comes to global customers. Read on for some enlightening and thought-provoking global issues. And please -- if you have any thoughts on the topics, enter them in the comment area.

As always - yours in BI success!

Claudia

Dear Mrs. Imhoff,

I have been reading about your Corporate Information Factory (CIF) in many different white papers and articles on the internet. They are all very interesting and well written with lots of valuable information and ideas. First, thank you for this treasure. Second, why don't you also write something about the challenge a global company is up against trying to give the organization the "single view" of their clients? This is always left out, not only in your analysis, but in everyone's research. It seems to me that this is something like a forgotten topic.

Let me try to explain what I mean in more detail: a global company, like [our financial institution], is storing customer information in disparate systems all over the globe. This is a result of numerous mergers and acquisitions and the "silo'd thinking" in many of our business divisions and group.

We have a corporate system which is collecting (or harvesting) the data, matching them to one another to create a global client view (much like the D-U-N-S number) and adding connections between these global clients to supply the bank with the "household view". We have been doing this for credit risk purposes for more than 10 years. The global identifier is assigned once and not changed unless corporate actions is of such a type that we need to. This gives us a very good grasp of our history as well. But unfortunately this is done only for credit risk purposes.

This is all fine for credit risk but already we see a problem: we are not allowed to harvest all client data from all databases in all countries in order to support other business areas like marketing. This means that what we call the Global Client is not really guaranteed to be truly global since we are in danger of missing out on one or more representations of a client in a database from which we receive no data. What do we do then? Basically there is nothing we can do since the legislation in these countries does not allow us to receive, store and work with client data from these countries.

I always argue that companies might be globalized but that the world is not. Meaning that the world must go global first in order for "global organizations" to become truly global.

Would you have any views to share on this "global issue"?


Posted June 20, 2005 12:03 PM
Permalink | No Comments |

And it is about time! For years, I and others have been touting the need for business intelligence to move from simple data gathering and reporting to having an active role in business process improvement. Well, we are finally starting to see this progression as a survey of 300 business-technology executives shows.

The survey was performed by InformationWeek Research and reported in an Optimize Magazine article. Certainly some of the old standards are still be chosen as key business objectives for a BI environment -- for example, increased revenues, faster up cycle times, and reports of current sales -- but these actually got the lowest number of respondents. The objectives that got the highest number of votes were monitoring performance metrics, improving business planning, and managing internal performance.

Colin White, my good friend and colleague, has long stated that BI has been way too data-focused and must become more process-centric. He will be pleased to see that this is exactly what is occurring. This will drive the enterprise to achieve what Keith Giles, an analyst at Forrester Research says is a "single version of the business truth". I like that modification to the popular "single version of the truth" saw.

Another interesting question in the survey dealt with the barriers that enterprises still have in adopting BI tools enterprise-wide. I would have guessed that data quality would have been the number one choice. I would have been wrong. It was the number three choice. What were one and two? Training (or a lack of technical expertise among the BI users) was the runaway winner with about 60% of the respondents and technological integration/compatibility coming in second place with a 50% respondent rate. Data quality and no clear ROI were in the low 40's percentage-wise.

Sounds like we need to spend a great deal more time working with the business communities to ensure that they are comfortable with all the fancy new BI tools they just received. We in the IT area tend to ignore this most crucial aspect of any BI implementation -- to our detriment.

Or perhaps we need to focus on building better dashboards or other forms of interfaces to mask the technological wizardry behind the curtains from our business communities...

The very active M & A or consolidation activity we have witnessed recently will be a great help in terms of the technological incompatibility -- or you could use technologies like Microsoft or SAS who grew their own.

Finally one last bit of good news from the survey -- BI spending will increase as companies integrate their BI applications into their business processes. Forty-seven percent of respondents said their budgets would increase and 43% said theirs would same at the current level.


Posted June 16, 2005 3:14 PM
Permalink | 1 Comment |

For those of us in the Denver area, we have long known about the infamous baggage handling system problems. For more than 10 years, Denver International Airport (DIA) has been trying to get its ultra-modern, $230 million computerized system to work. As one of its victims (my bag was shredded), I know I am happy to hear that United Airlines is giving up on trying to make it work.

"It never worked up to its potential" said United spokesperson, Jeff Green. I'll say. Its abilities to handle baggage are a far cry form what was hyped 11 years ago. We were promised a state-of-the-art system that was to use PCs and thousands of remote-controlled baggage carts to operate on a 21 mile long track. The carts were to carry the baggage from counter check in to sorting areas then directly to your airplane.

For many of us, United's "pulling the plug" on this deficient system has come many years too late. According to one spokesperson for a consulting company that works with troubled IT projects, " The first lesson is that the best way to build a large, complex system is to evolve it from a small systems that works."

There is a lesson here for those of you building a large and complex BI environment. Those of us who have done this for years always recommend building a working prototype before going after the full-blown production application. Let DIA's woes be a lesson for all of us.

United stated that "We're going to go back to the old way," -- that is, using manual labor to hump the bags from one place to another... Sure hope they read the tags correctly.


Posted June 14, 2005 2:58 PM
Permalink | 1 Comment |

There is a new acronym for everyone -- BPO or Business Process Outsourcing. According to Yankee Group, this is the integration of multiple business services in your enterprise with a single provider. "The whole delivery mechanism for ERP is changing" because of this trend, according to Phillip Fersht of the Yankee Group. "Suddenly ERP is no longer the core strategy of the enterprise". This spells big trouble for SAP and other large ERP vendors unless they can adapt.

The trend in outsourcing an enterprise's business processes is not new. It started years ago when companies figured out that HR management was not a core competency. Now, Fersht says, the idea is spreading to other organizational processes due to the almost instant benefits and cost savings. Yankee Group predicts that the fastest growth will occur in North America with a stellar 43% growth occurring between 2005 and 2006.

What does this means to ERP vendors like SAP? They will have to shift their business model from selling ERP packages to licensing their software to their BPO partners. So far, that is exactly what SAP is doing. They have signed up 10 partners whose licensing revenues hopefully will replace the earnings that are disappearing from the sales of their core ERP suites.

From an BI perspective, this may also free up these large vendors to focus more attention on the added value they can bring to customers from better analytics and BI applications offerings. It will be interesting to see what happens to these companies in the next 5 to 10 years.

So what are the benefits of BPO? Reduced administrative and transactional costs, transferance of management responsibilities to a third party, reduced software and hardware costs, and so on.

OK sounds good -- but a cautious note is needed here too. This is a new area. Granted a lot of the big companies like IBM, Accenture, EDS and CSC are poised to take over all of your operational processes but still, this is a mighty big leap for a lot of companies. If you are considering this, I recommend that you understand exactly what these companies are responsible for and what your company's responsibilities are. Determine which operational processes are amenable to outsourcing and which ones do still offer your company a competitive advantage. The outsourcers do not do everything well.

As in any outsourcing situation, there are a number of pitfalls as well as benefits. Your job is to mitigate the former and enhance the latter.


Posted June 9, 2005 2:16 PM
Permalink | 6 Comments |

Dave Koch was a good man. He was an even better friend. He disappeared while hiking Grouse Mountain just outside of Vancouver, BC on May 25th. After an intense search and rescue effort, his body was found today. I know I speak for many of his friends at DM Review and around the world as well as his family when I say that he was a joy to be around, a fun-loving and caring individual. I will miss him terribly.

My thoughts and prayers are with his family on this sad day.


Posted June 7, 2005 4:28 PM
Permalink | 2 Comments |

Sun announced today that it is acquiring StorageTek for $4.1 billion (all in cash) in a deal Sun hopes will restore its financial wealth. It's an enormous gamble and the last one of this size that Sun can do given its financial war chest. Is this a sound gamble?

Sun has long been trying to reestablish itself in the server business with limited success. It has also tried its hand at selling software with its well-publicized open source push. It seems strange then that the company would turn around and buy a long-established storage company but the talk on the street is that Sun believes that storage is "where it is at" -- that is, where they will be able to make their financial fortunes.

It also means that Sun wants to transform itself into a total systems company -- basically one stop shopping for all your hardware needs a la IBM. And it means that StorageTek may get the needed push to move it out of the languishing position it has been in for years.

But not everyone is cheering the deal. A Prudential analyst, Steve Fortuna, states that the deal "reduces Sun's cash hoard by 40% and does nothing to reignite revenue growth or profitability. We would rather have seen the company buy back a billion shares and fire 10,000 people". Ouch! That is some serious criticism at a time when Sun could use some positive news.

One can only hope that this acquisition fares better than several of its past ones like Cobalt Network, Tarantella and Procom. And one can only hope that the addition of StorageTek's 1000 person sales force will help the overall sales for the combined companies... I guess all that remains to be seen.


Posted June 2, 2005 3:56 PM
Permalink | No Comments |

In case you haven't heard, we Americans love our emails. So much so that we check them in the bathroom, in church, and even while driving... So says a new survey sponsored by one of the biggest email providers, America Online.

The article continues that the average emailer in the US has 2 or 3 email accounts and spends at lease an hour every day reading, sending and replying to emails. For 41% of the respondents, they are so hooked on email that they check their email boxes as soon as they get out of bed in the morning. I don't know about you but that is usually not the first thing I have to do in the morning!

So here is a test to see if you are addicted: See how long you can go -- hours, days, weeks? -- without checking your email box. For those surveyed, a large percentage couldn't wait 2 or 3 days without checking -- and that included even during vacations! Well, are you addicted?

Another interesting snippet in the article was that all that email has lead to regrets, Almost half of those surveyed said that they would have like to have the ability to "unsend" a message they sent by mistake. I count myself in that category.

For those of you determined to stop the addiction, here are some tips from the article:

1. Resolve NOT to check email after a certain hour of night.
2. Stop the interminable back and forth by picking up the phone and call the person.
3. Keep track of how many messages you send out in a day.
4. Go without email one day a week.

OK -- those are at least 4 steps in your 12 step program. For me, I gotta go. I have a zillion emails that came in while I was writing this...


Posted June 1, 2005 2:54 PM
Permalink | No Comments |