Introduction to Business Intelligence Today
Business intelligence is defined as "mission critical" by many senior executives today. The emphasis and interest in BI, as we will often refer to it, has placed it in the forefront of the list of major corporate objectives. This adjective is quite valid because the value of unlocking critical information held in corporate and external data sources can be a significant game changer. At the enterprise level, BI is often just a stated goal with little actual practice other than perhaps setting a standard for a suite of tools. Having an enterprise goal and set of standards does not end with creating an approved vendor list—it is just the beginning. BI at the enterprise level suggests that there is a common vision and set of goals in the deployment and use of BI on a broad scale within the entire organization.
In my opinion, business intelligence is the application of end-user query, reporting, dashboards, and other non-programming technologies to provide information that is not available to the business using traditional programming methods and services. BI requires a clear direction at the enterprise level with the realistic expectation of the skills required to deliver BI output that is mission critical. It also requires a support infrastructure to ensure accuray of results produced and that the proper skills are in place.
Let's think about how you would proceed with a corporate-wide ERP or CRM system and the resources, dedication, and critical scrutiny you would apply in selecting, implementing, and supporting one of these major application solutions. Would you have the system installed, show a few people how to use it, tell everyone it's now the corporate standard, and then trust its acceptance to mere synergy? I certainly hope not! Yet, this is often the case when a BI solution has been chosen.
In this book, I have taken the approach of opening a frank, personal dialogue with you. It is an open discussion about enabling BI at the enterprise level. It rarely mentions any product, but rather addresses the requirements and thought processes necessary to succeed at the macro level of BI. It is intended to assist in forming, articulating, and defending a global BI strategy and vision. For the most part, the days of acquiring a set of independent BI tools and turning them loose in the enterprise are over. However, the majority of clients I talk to have an already-established set of BI tools in-house. They may have from three to a dozen different BI tools with overlapping functions. One of the first steps in establishing enterprise BI sanity is a bit of winnowing out of the less productive or dated ones. I will have much more to say about this later.
One of the first rules of thumb today regarding BI enablement is to totally avoid the "Fire! Ready! Aim!" approach. Uncoordinated, anarchistic BI has never been effective, and it can be costly. Your end users can easily populate a spreadsheet in a myriad of ways and run amok without much assistance. When you do not have a plan for BI, this is the most common form of analysis within any enterprise. End users will always find their own way if they are not led in a positive, orderly manner.
If you believe that a BI solution can change your corporate world, there must be an internal paradigm you adhere to. Typically, BI is thought to have the following characteristics, at a minimum:
- An effective set of tools for accessing data and delivering business information
- A means to gain insight into areas of the business not accessible with existing systems
- Advanced analytics that, if applied, can actually "discover" new information
- The capability to make people more productive and less reliant upon IT
- The capability to provide a different interpretation of critical information than we have today
The corporate BI quagmire becomes deep when a mismatch between desire and commitment becomes apparent. I often get engaged in BI conversations where a client will talk about his avid interest in BI and how he feels it can make a significant difference in his success. Then, as I probe a bit about the overall plan, it becomes apparent far too often that much of the "plan" is based upon assumptions about what BI solutions really do, along with the ease of use factors the client believes will be in play but that have not been proven.
In this chapter, we discuss overall BI scenarios today, the view of the CIO, the IT perspective, the end user perspective, and establishing a vision. Lewis Carroll wrote in Alice in Wonderland, "If you don't know where you are going, any road will take you there." I would also add: "How do you know when you've arrived?"
"I am not sure what BI really is these days, but our execs tell me we need it." This was quoted in a seminar on business intelligence by an experienced IT individual who had been forced to attend the event in mid-2009. You may be tempted to snicker at this naïveté in this day and age but, as the old saying goes, sometimes ignorance is bliss. When probed a little further regarding his inquiry, what he was really asking was: "Why is BI suddenly such a hot topic with our senior management team? We are already using several end-user tools and yet they want more!"
Having worked in this arena since 1981, I can think of countless customer engagements where this question arose in some manner or another. My answer in 2003 (Mike Biere, Business Intelligence for the Enterprise, Upper Saddle River, NJ: Prentice Hall PTR) was, "Business intelligence is a word problem!" What I meant was that BI transcends simple query and reporting. It eclipses dashboards and charts and portals. It is often applied to solve complex business problems and provide an answer heretofore unknown. It often requires complex logic to be applied. I also constantly ranted about the lack of BI skills that fell short of the desire to deliver BI analyses. There is a certain level of skill required for the various degrees of BI complexity being addressed that many end users ignore until they get in over their heads. There is a continuing gap between user groups where "power users" still produce the bulk of the output for consumers, regardless of how much easier to use many BI tools are touted to be.
BI skills are not easily mastered; nor are they acquired by those who do not have the proper technical skills to work with a tool that may require extensive manipulation of data. This text is not a rehash of the first edition but a guide on BI today. The world of BI today is dramatically different than a few years ago and must be examined in a new light.
The emerging tidal wave of BI interest was beginning to dramatically build in 2003 and, at that time, the emphasis was on making people aware that BI efforts needed to be properly supported, that skills had to be assessed realistically, and that we must not assume that just anyone in the enterprise would be able to use a tool effectively. The ongoing myth of ease-of-use and universal applicability of a BI tool being a trivial exercise had to be addressed. The transition toward self-service, on-demand BI was beginning to take place, and it deeply affected the marketplace and how many viewed BI in a new light.
BI should be considered a "potentially" powerful weapon in the hands of all employees within an enterprise. In today's world, it is best to think of BI as an integrated solution suite, where its power and functionality may be utilized by anyone who touches data within a particular context. It is all about equipping individuals with the proper functions based upon their needs and skills. It is far less about equipping everyone to be a BI hands-on tools "mechanic." The push today is to drive BI deployments as broadly and deeply into the organization as possible. It is also about providing BI functions that add tremendous value without the end user having any skills in the tools being used. This is referred to as "embedded BI." The age of the BI consumer is here.
The business intelligence market is heating up but with an entirely new suite of players, such as options available on the open source market. Well-established vendors are piling on to this enormous market by acquiring others to fill in portfolio gaps, and thus we see a series of mergers absorbing some of the longer-standing independents. This is wonderful news to a BI vendor but, for anyone involved in the acquisition process, it can be a nightmare. There are decision points and options not available in the past, but the options have also become far more complex in many ways.