Systems; Infotech : My very own universal Access code

GENERATION 21X : INFOTECH
My very own Universal Access code

By Mahadevan Venkateshwaran

This is the age of the Business Information Revolution. Today, the defining concern is technology, especially infotech. How a firms gathers, manages, and uses information will determine whether it wins or loses tomorrow. Firms are increasingly realising that infotech can be used as a weapon to gain competitive advantage.

Enterprise-wide applications like Enterprise Resource Planning (ERP), which help a CEO obtain a consolidated view of enterprise-level data, speed up processes, slash inventories, and raise production-flexibility will become more popular this millennium. But although the sales of software-packages are rising dramatically most of these projects suffer from cost and time-overruns. This does not point to any obvious shortcomings of ERP itself; in fact, such packages can transform the way a company conducts its business. In a majority of cases, infotech is not the problem, but its management is. The role of infotech as a business enabler is known. It can help catalyse business transformations for companies that know how to use it effectively.

THE ROLE OF THE TOP MANAGEMENT. CEOs should become as engaged in infotech as in any other business function. The CEO must regard infotech as a strategic resource which can help the business get more out of its people; it should not be regarded as just a cost-centre. In line with these trends, the millennial organisation will deploy business intelligence software to deliver high-quality information for decision-making. Multi-dimensional data analysis to predict growth trends, benchmarking, and customer retention will be employed.

FROM DATA PROCESSING TO DATA REPRESENTATION. The emphasis of infotech-research will be representation. Companies have huge volumes of data, but often lack the expertise to process it. Multi-Dimensional Modeling (MDM) is one approach to making sense of this data. It helps managers to look at business models as a set of measures built around business functions. In contrast to On-Line Transaction Processing (OLTP) systems, which are designed around entities, relationships, functional decomposition, and transition analysis, MDM for data-warehousing is based on facts, dimensions, hierarchies, and sparsity. The models are not only understandable by businesspeople, but are also expressed in a way that is natural to the user.

A well-conceived MDM can be implemented in a relational database, a multidimensional database, or even an object-oriented database. For instance, in the analysis of customer buying patterns, the company would like to determine: what is the number of times a particular product was bought in a month by a customer? For what products is the average age of buyers over 30? And the number of customers, as well as their ages and gender by region, who bought a product.

To generate these details, the database of the company is structured as a cube, more formally known as a Multi-Dimensional Database (MDD). The cube would, thus, have 3 dimensions. One corresponds to the gender of the customer, with the axis split into two: half for males and the other for females. Similarly, the region axis is split into 4 and the product axis is split into, say, 50 pieces, one for each product. The larger cube, thus, contains 400 (2X4X50) sub-cubes. These are the building-blocks of all reports. From the perspective of On-Line Analytical Processing (OLAP), there are 2 types of columns. Dimensional columns contain data for summarisations. These take fixed numbers of values. Aggregate columns are calculated amounts. These are numeric values, such as dollar amounts or cash.

BUILDING AN MDD MODEL. The first step in building a MDD is to pick a business subject area-weekly sales reports, monthly financial statements, or insurance claims costs-and model it by asking these fundamental questions: what business process is being modeled? What are the measures ? At what level of detail is active analysis conducted? What do the measures have in common?

Active analysis is the ability to manipulate data. This can refer to mechanical manipulation, such as pivoting, or graphing; agent-based manipulation, like alert reporting, or exception reporting; and workflow manipulation, such as publishing or distributing documents. As companies build these models, they will avoid including everything. They will concentrate on a specific area such as customer profitability. This can be achieved by formulating a simple statement describing the intent of the model.

In many companies, huge investments are being made in infotech without any visible benefit in operational performance, let alone higher profitability. How will the millennial organisation leverage the power of infotech to gain competitive advantage? The first step is to integrate infotech into the strategy process. Infotech leaders should be drawn into the company's decision-making circle. Infotech must be presented as a corporate priority. Indian firms will, increasingly, take recourse to the MDM approach to slice and dice data to find the information they need. Retail chains, bookmarts, and fast food chains, will increasingly use OLAP techniques to drill down into databases. This will help them visualise data in a way that is readily understandable to business users. These will endow companies with the ability to identify clusters in the data by being able to look at data along several dimensions at the same time. Finally, success in infotech must be measured by profits, not cost. This will require companies to view technology as a revenue-driver rather than a cost-centre. When companies plan infotech investments, many miss the simple truth that even the most sophisticated of systems are nothing more than boxes and wires unless they are used effectively. Only the appropriate infotech infrastructure, coupled with the appropriate infotech usage, can lead to the realisation of benefits from infotech.

M.Venkateshwaran is a second-year MBA student at the
Indian Institute of Management, Lucknow

Issue Contents Write to us Subscriptions Syndication