A New Analytic PerspectiveDecision makers throughout your enterprise need an IT architecture that serves their needs — rather than the other way around. Here is a view of emerging enterprise analytic systems that uses BI and analytic requirements as the point of origin
by Kemal A. Delic and Umeshwar Dayal In the first 50 years of the history of computing, we've seen the deep and comprehensive infusion of computing systems into various business domains. Today, computing systems represent an indispensable infrastructure, with which we run, manage, and coordinate business operations. We envision that the next decade and beyond will bring a new era of ubiquitous computing systems. We believe that this progress will lead to the better use of interoperating enterprise analytic systems: in other words, a vision of CRM, ERP, enterprise application integration, and enterprise knowledge management applications all interwoven through analytics. We predict the morphing of contemporary enterprises into more intelligent enterprises, which will operate in various business markets that mimic natural ecosystems, evolving and adapting based on their quest for survival. Today's enterprises function nonstop under an endless and variable stream of events that end with either the closure of a business transaction or a reaction to originating events. Data and information repositories, logs, and traces capture transactions and events, thereby providing the foundation for higher-level applications — namely enterprise analytics. Large enterprises operate today with the reality of huge data volumes and vivid business dynamics. They transfer terabits of data daily while accumulating terabytes of data and information and processing them with teraflops of computing power. To function smoothly, enterprises must build their architectures carefully to govern distinct enterprise layers for event management, transaction processing, and the rendering of enterprise analytics. Today, most approaches to enterprise analytics start from the inside out: from the perspective of data and information sources themselves. We suggest starting from the outside in — from the actual users' needs and their typical decision-making behavior. In this article, we'll outline three broad categories of users and discuss how an enterprise analytic architecture might serve them. We'll argue that understanding how they make decisions — and in this way, clarifying their data and information needs — will help articulate a better conceptual enterprise analytic architecture. Modeling Complexity: ApproachesBusiness enterprises are extremely complex systems, difficult to manage and hard to model. Nonetheless, we model business enterprises to better understand how to manage them, where and how to evolve them, and how to improve business and financial performance. Our experience tells us that at least two-thirds of enterprise complexity is attributable to organizational complexity, while one-third is due to IT complexity. Taking an abstract, holistic view, we can stratify enterprises into three layers:
In this simplified view, IT systems in layers 2 and 3 function as mediators and serve as the business conduit ("fabrics") for monitoring business operations, grasping the holistic enterprise situation, and executing the necessary actions that optimize a limited set of top business parameters, such as revenue, profit, growth, and market position. Corporate managers and senior executives are typically in the business cockpit and therefore take these actions. Another useful way to consider an enterprise is to split it into two distinct parts: static or slowly changing (such as the IT infrastructure); and dynamic (such as the typically enormous number of operational parameters), which includes data and information flows within and outside the enterprise. Both parts are critical to supporting corporate decision-making. The static part acts as an overall context, based on a current set of operational parameters. However, decision-making is dynamic by nature: It has urgency and involves different types of people and styles.
|
 |
Most Popular This Week
IE Weekly Newsletter
Subscribe to the newsletter
| |
| |||||||||||||||||||||||||||||||
