Helping Business Evolution Along

Improved computer modelling tools can analyze each facet of a company's operations, allowing it to better adapt to new trends

By Paul Horn

Anthropologists trace human social and economic evolution through distinct phases, typically tied to the creation and use of tools. Tools provided new means for subsistence and commerce, which in turn led to human migrations all over the world. Newfound abilities allowed humans to roam farther and make sociological advances.

These historical trends persist in today's business environment. New tools permit different ways of doing things that can lead to development and growth -- if businesses have enough foresight to see the possibilities. Just as with early man, some lines won't make this transition and will die off in the face of competition from those who did.


  Not too long ago, these business tools often took the form of a computer system or piece of software that automated a company's back-office operations. Over time, these tools became the infrastructure on which nearly all business operations rely. Today, the real opportunity lies not in the infrastructure itself but in novel things that can be done with it to analyze and improve business models and find new opportunities.

The analysis of a business model can be complex, demanding a rigorous, disciplined approach, not unlike that of other scientific or engineering challenges. A business can be viewed as a system, with certain inputs and outputs and a great deal of churn in between. The detailed analysis of those in-between processes is where greater efficiencies can be gained, productivity improved, and new opportunities identified.

Just as with mechanical or electrical systems, developing a better understanding of how a business works can help one target improvements where they will have the most impact. Simple automation is not always the answer -- as the old adage says, automate a bad process, and all you have is an automated bad process.


  Most analyses of mechanical or electrical systems are done through modeling -- simulating the system, often in software, so that problems can be identified and enhancements tested without having to build a whole new system. Modeling techniques can be quite complex and involve detailed mathematical calculations that require large computational capabilities, storage capacity, and data. Those requirements become huge when you're trying to model an entire business, with all of its variables and pieces.

And yet, tremendous advances in information technologies and a better understanding of business processes have brought us to a point where we can now take on modeling challenges like this. To reap the benefits of improved tools, these modeling concepts have become a focus of much research and development activity.

We can now identify and classify the elements of a business, breaking it down into its most fundamental processes for in-depth analysis. Using this "componentized" approach, or "component business model," we can establish standards by which these processes can plug together and interoperate, even with those of customers, suppliers, and others outside the enterprise.


  In this model, the components are viewed simply as services on which the business relies, rather than as defining the business itself. The integration of these processes and the flow of necessary information between them can be accomplished by following a predefined set of rules known as a service-oriented architecture (SOA).

Complicated? Not necessarily, and the rewards can be significant. SOA is the next step up the evolutionary ladder for businesses, giving them the tools to go beyond entrenched processes and silos of activity to a more dynamic, adaptable model that can adjust as conditions warrant.

This is the nature of evolution: Exploit new tools and new thinking in order to ascend to the next level. Just as in human development, without continual improvement, the life of a business might be -- to borrow from writer Thomas Hobbes -- short and brutish.

Horn is senior vice-president of IBM Research

    Before it's here, it's on the Bloomberg Terminal. LEARN MORE