Business Intelligence Realization: An Agile Answer

Originally published May 19, 2009


In my first article in this series, From Cave Paintings to CNN: the Quest for Business Intelligence , we establish that using visual metaphors is a pervasive form of communication we call “consumer intelligence.” In a consumer context, information is communicated in graphical and pictorial forms since they “speak” to people in a way dry content cannot. At the same time, graphical forms transcend language and even the knowledge of the consumer. Inspired by rapid advancements in consumer intelligence, we discovered the root cause of why the full potential of business intelligence (BI) has not been realized by enterprises despite huge investments of precious time and money. We examined the way BI solutions are built and found the problem starts in the requirements phase. Put simply, dry textual forms of setting forth business requirements are not expressive enough to allow business people to envision the solution and how they will use it to derive value. By revising the methods and methodology used to build BI solutions, we can exploit the lessons of consumer intelligence and visualization.

Three objectives for this enhanced methodology stand out:

  1. Deliver “business insight” using visual metaphors;

  2. Discover the requirements that will yield the “right” information; and

  3. Use visualization techniques and representations to discover and confirm what is “right.”
Ultimately, the new methodology must emphasize delivering high value to the business by employing what we know about visual communication. We should visually create a shared understanding of the context in which we construct a solution.

Agile is Everything These Days!

Agile software development is a time-boxed, iterative way of building software that attempts to solve many of the past trials and tribulations of the waterfall approach. According to Wikipedia, Agile development applies to an approach “where requirements and solutions evolve through collaboration between self-organizing cross-functional teams.” In other words, let’s work together and get results oriented! The methods incorporated into the Enterprise Intelligence Realization Methodology™ (EIRM) deeply embrace and embody the values of the Agile Manifesto. The value added in EIRM is the concentrated use of visualization as the path to enhanced communication.

It is important to be on the same page as we discuss methodology. To clarify, EIRM refers to a series of methods, practices, procedures and activities. This article provides not only the framework upon which the series of methods are built, but also focuses on the specific methods to be used for extracting business requirements. For now, let’s just use the visual model below as our Agile methodology.

 
alt
Figure 1: Agile Development Life Cycle

Designing the World the Way We Want It to Be!

Inspired and guided by the appeal of visualization in consumer intelligence and taking into consideration the pillars of Agile software development, four guiding principles are used in specifying EIRM methods.
  1. Engage the business

  2. Leverage best practice methods

  3. Use visualization to communicate

  4. Use deconstruction to simplify

Engage the Business

EIRM methods for gathering requirements should engage business participants by using a context that is relevant to their daily lives. By pinpointing the right context, we have a better chance of getting the requirements right and helping the participants envision how they will use information in their day-to-day operation.

Leverage Best Practice Methods

Over the past 50+ years, many millions of hours have been spent implementing software applications. The challenges, successes, failures and lessons of this work provide a fertile field of inspiration for business intelligence. Looking specifically at the years of packaged ERP, “best practices” spring forth assuring successful implementation and achievement of business benefits. Unbelievably, these best practices have not yet been applied to business intelligence!

So here are four best practices from ERP applied to business intelligence:

Business flows must be incorporated into driving and rationalizing requirements and understanding how BI information will be made actionable. The use of flow diagrams showing the processes and how they interact become our “cave painting.” This is the most effective way to translate the context of the BI solution into pictures for all participants to absorb and understand. Depending on the school of thought in your company, there are multiple variations of process flows. Among the variations are value streams, business scenarios, value chains and process decomposition.

Conference room pilots prove that to see is to believe. Visualization is the best way to confirm requirements with the business user. Real life use cases provide rapid and reliable validation of the solution throughout the project.

Organizational design must be understood as a critical part of making information actionable. All participants must know what actions should be taken and by whom. This is also the best way to make sure all the right people are participating in the project.

Change management must be used to help the company align with the goals and targets being measured. Too often, business intelligence value gets lost because people in the organization do not understand exactly how they can contribute to successful achievement of objectives.

Use Visualization to Communicate

What we are really attempting with business intelligence is to present an accessible picture of the business on screen or paper. Keeping in mind that metrics do not exist independently of the processes of the enterprise, we have to find a means to create a sensible picture of the model in the minds of the solution users. Dry text doesn’t work. Visualization is the most effective means for creating a common language that can be understood by all the parties required to create and evolve a solution that delivers high business value.

Use Deconstruction to Simplify

Deconstruction results in a framework that breaks a complex thing into parts. It establishes the relationship with other parts and gives a common visual reference model so everyone knows where they fit. It is the basis for a common understanding between all the participants. We can use it over and over to put things in the right context and create order from chaos. Employing a visual framework allows us to address, align and assign technologies, issues and problems, responsibilities, costs, time, techniques, methods, and tests among other things.

Think Ecosystem!

In our biological ecosystem, nothing stands alone. All pieces and parts are symbiotic. The same applies to business intelligence within the enterprise. The entirety is inextricably intertwined. Thus, beyond the Agile development life cycle, we again need to articulate a visual framework to contextualize our BI solutions. At the core of this framework is the notion that BI capability exists in the context of all other systems and business models in operation within an enterprise. Here is a schematic of the framework.

alt

Figure 2: Enterprise Intelligence Realization™ Methodology Ecosystem Framework

As we consider this framework and its symbiotic implications, it is important to truly understand why “business intelligence” does not exist as an island. When business people think about the “business,” they have a mental model for the whole of the business rather than just an individual process. When we think about our personal finances, the balance in the checking account is not the sum total of our financial position. We consider the next paycheck, savings accounts, assets, retirement funds, and other sources of income. Precisely the same holistic model exists in the enterprise. Accounts receivable does not exist alone. It is part of a stream of processes that have upstream and downstream value and implications. It contributes to the operation of the enterprise as a whole. Too frequently, ERP systems and the corresponding business processes are designed and implemented independent of the need for business metrics, reporting and a unitary enterprise view. Worse, standard operating procedure for ERP projects is to defer reporting and business intelligence to future phases. When this happens, not only does the effort and requisite budget get deferred but the architecture for the ERP solution leaves out vital components needed for providing reporting and analytic capability: business intelligence!

Pan for BI Gold!

On-point business requirements form the currency of achieving BI value. Striking BI gold requires the right tools and methods. First, determine the right information – key performance indicators (KPIs). Draw a picture to visualize the context for the business questions to be answered. The goal is to get everyone on the same page and make sure all the right people are participating. As in best practices from ERP, use process models, value streams or business flows. They provide the relevant context for understanding the operational part of the business being measured as well as some of the possible sources of the data. Some examples are order to cash or procure to pay.

Second, go with the flow. Once the process context is agreed upon, identify the required metric content. It is useful to list the required KPIs on the flow diagram to reinforce the nexus between metrics and business processes. Ultimately, these visual elements take the form of annotated business process flows supported by definitions, descriptions and the myriad details normally collected.

Third, make sure the KPIs are actionable. The action should bring about an improvement or resolution of the problem. It is vital to know who in the company will take the action. “Use case analysis” should be used for this step. It combines a diagnostic trigger with the organizational model and can be easily diagrammed. Again, we are using pictures to communicate. “Use case analysis” is regularly the source of an “aha” moment when one discovers they have been measuring things that are not truly actionable and will not improve the business.

Finally, build the prototype scripted by the use cases, the first round of simulation. So that we know what we will measure, what actions should be taken and exactly who will take them, we need to use BI visualization to test our hypotheses. Critical to establishing buy-in from the business, build prototypes of the solution so the business user can see and experience how to interact with the information as they would in real life. New questions will emerge and the prototype will be tested to see if the questions can be answered. This step is the equivalent of the Conference Room Pilot best practice from ERP and should not be skipped. As the saying goes, “This is where the rubber meets the road!” But remember, we have not “built” a production system yet. We are still “panning for gold.” More work must be done to complete the picture. In my next article, I will focus on the data necessary to fulfill the requirements.

Visualization as a means to communicate, simulate, discover and inspire is not new. Cave men used it! Now CNN uses it! Rethink what can be accomplished through business intelligence and what should be enhanced to deliver on the promise. Business intelligence is not simply a technology or a product or a process. It is a method of communication requiring an understanding of how the human mind works and the power of pictures!

  • Steve PalmerSteve Palmer
    Steve is President of FactFusion, Inc., a consulting and advisory company focused on guiding companies toward realizing the benefits of business intelligence and enterprise information management. Steve has pioneered in data warehousing, business intelligence and information management since the mid 1980s. The FactFusion mission, Enterprise Intelligence Realization, is born of extensive experience as practitioners and consultants with some of the world’s largest companies. Steve can be reached at steve.palmer@factfusion.com or at (661) 251-9400.

Recent articles by Steve Palmer



 

Comments

Want to post a comment? Login or become a member today!

Posted May 20, 2009 by Anonymous

Thisw is great Steve.  A very comprehensive view.  Fantastic first article. 

Is this comment inappropriate? Click here to flag this comment.