Business Intelligence Network - The Vision for BI and Beyond

ADVERTISEMENT

Business Intelligence is Dead – Long Live the Highly Evolved Business, Part 1

What Does a Highly Evolved Business Look Like?

Published: 10 January 2007

This and following articles will explore why and how business intelligence and HEB differ, why a HEB represents a significant step forward and how this new approach is finally becoming a reality.

It’s probably not a good idea to start one’s first article on a business intelligence (BI) site with the words “Business intelligence is dead”! But, I was trying to catch your interest. And, in a sense, I do mean it. Business intelligence as it has been known for many years is dying; and, in some ways it’s about time, given that it has been around since the late ‘80s in this form. Business Intelligence needs to make way for a more comprehensive intelligence in business (Mike Ferguson of Intelligent Business Strategies has been promoting the concept of an intelligent business for a few years now), which I propose is a characteristic of a highly evolved business (HEB). In this and the following articles, I’ll explore why and how business intelligence and HEB differ, why a HEB represents a significant step forward and how this new approach is finally becoming a reality. 

A highly evolved business, like a well-integrated person, displays three behaviours – senses, responses and the intelligence to link one to the other. Together, these three aspects enable an organism, be it animal, person or organisation, to be adaptive to its environment. A well-functioning business therefore senses its environment and any changes occurring therein and responds to the environment and its changes in an appropriate manner. Determining an appropriate response is what is meant by intelligence. And a response (intelligent or otherwise!) causes changes in the environment which should be further sensed – leading to a classic feedback loop. The simple model in Figure 1 represents this.

Figure 1 

As in the animal world, not all sensory inputs require intelligence to generate an appropriate response. If you’re hungry, you eat; thirst invokes drinking and so on. These are instinctual reactions. In a business, if someone wants to buy a product, you sell it. At the most basic level, it is such instinctual responses that have traditionally been incorporated into operational applications. 

However, most sensory inputs demand a choice as to what is an appropriate response. In the face of a physical threat, an animal chooses between fight and flight based on an evaluation of the possible outcomes and its instinctual urgings. In business, a customer enquiry may lead to a choice about which particular product most closely satisfies the need. In many cases, such basic decision-making can also be codified into operational applications. For example, if the customer is high value client, the product offered should be the high-end version. On other occasions, a simple question to the customer or agent can determine the correct choice. 

From whence, therefore, does the intelligence come? If a person is physically involved in the decision process, it clearly comes from a person. But even in the case where the basis for the choice has been codified in an application, human knowledge is just as clearly the ultimate source of the decision-making. And in both cases, the basis on which the decision is finally taken is information about the real world. So, we must add people and information into our model, as shown in Figure 3.  

Note that in the more traditional business environment, the feedback loop in Figure 1 can be rather tenuous. A bit like the animal heading back to the waterhole where lions hang out again and again! In today’s on demand business environment, this feedback loop is vital, but let’s ignore it as we build this model and then add it back in at the end. 

 

Figure 2 

This model may be thought self-evident, but by exploring the concept of a highly evolved business in this way, we can come to an interesting conclusion. If we replace the word “intelligence” in Figure 2 with “business intelligence,” it becomes immediately clear that what we call business intelligence is simply a very restricted and focused example of what goes on in a business generally. Business Intelligence, as currently defined, is that subset of decision-making which requires extensive, cross-enterprise or historical information as a basis for complex and perhaps less urgent decisions. 

Figure 3 also clearly shows the three key dimensions of a modern enterprise environment – people, process and information.

 

Figure 3 

In the past, business intelligence has tended to focus only on the information and people layers of this model. It was assumed that little or no process applied to the type of decisions supported by business intelligence and further that any small amount needed would be known and understood by the end users themselves. As a result, business intelligence was less closely linked to the sense and response aspects of the model than most executives would like and was sometimes seen as being an intellectual exercise rather than driving the business. 

These assumptions are becoming less valid. Real time business intelligence, for example, needs well-defined processes that are relatively complex and require more control and structure than can be easily contained in users’ heads. Process is thus becoming increasingly important for business intelligence as the decisions being made become more tightly integrated in the day-to-day business and participate in closed feedback loops between different parts of the business. 

Finally, the model provides an interesting perspective on yet another triple view of the IT environment –operational, informational and collaborative.

 

Figure 4 

As previously described, the operational environment is focused on the day-to-day transactional operations and aligns with the sense – intelligence – response axis of the model; and thus is most closely focused on the process layer. The informational environment, focused on decision support, spans the three layers, but traditionally has been largely dissociated from the day-to-day operations. The collaborative environment, including e-mail, messaging, document sharing and general “office” work involving interaction with colleagues, is probably the most pervasive of the three environments, and at the same time more loosely connected to the other two – especially the operational environment. Paradoxically, it is this last environment that will bind the other two together in a new and innovative way, as we shall see later in this series. 

In the past especially, and still to some extent today, these three environments exist on different hardware and software platforms. This has traditionally made linking between them arduous. Data transfer from the operational environment, often mainframe-based, to the informational environment where the data warehouse might be on UNIX and the marts on Microsoft Windows has always been one of the more difficult aspects of running a BI environment. The collaborative environment, coming largely from the Microsoft Windows world, has tended to be seen as a “nice-to-have” and linking information to and from the other environments has typically been left to the users themselves.

Wouldn’t it be nice to have a fully integrated business, where all these different environments and layers are seamlessly interconnected and working together? This precisely what I mean by the term highly evolved business. The issues just discussed mean it has proven difficult to achieve this in the past. This, however, is about to change. 

 

What’s about to change it all is the service-oriented architecture (SOA) approach. In my next article, I’ll give an in-depth view of what SOA is all about. But, the key characteristic we need to see here is the fact that SOA allows, and in fact, demands a business-driven transformation of the entire IT environment into a flexible, integrated service-oriented infrastructure that spans every aspect of the business. And, for sure, “every aspect” includes everything we call BI, as well as operational and collaborative aspects.

As a result, BI as a standalone discipline is likely to disappear, or at least undergo a substantial and irreversible transformation, over the coming years as SOA takes off. Significant changes will be seen in extract-transform-load (ETL) processes, metadata and user interfaces in order to integrate the BI environment into the larger HEB. Much of what we’ve learned as we built our data warehouses will be of high value to SOA, especially in the areas of modelling, information engineering and cross-enterprise project initiation and management. And many of the tools and techniques delivered by SOA will be adopted in the BI world.

But the separate and independent BI world looks like it’s coming to an end as and when SOA achieves widespread traction and the highly evolved business begins to emerge. In the meantime, enjoy it while it lasts! 

If you found this article helpful and would like to receive the latest insights each month from Mike Ferguson and other experts featured on the Business Intelligence Network, please subscribe to the UK Business Intelligence Network Newsletter.