The art of extracting strategic information
TO employ a highly hackneyed expression, information is the key to success. Today, as information technology is making an incursion in our lives, Corporations world-wide are accumulating mountains of data. Now, all that remains is to extract useful information from the accumulated data. This task is anything but trivial!
Just for the sake of not leaving behind the who use the terms data and information synonymously, it is imperative that before I delve deeply into the above mentioned topic, the difference between the two words is grasped well by the readers. Data consists of facts, figures and plain numbers where as information on the other hand refers to the knowledge that can be learned by analyzing data; data is collected whereas information is extracted; data even if it is authentic may be misleading since the information extorted from it may be erroneous; data collection is an iterative process, however information retrieval is a highly specialized art, there are many online MCSE, CCNA, CompTIA, IBM and more exams solution provider on the internet but for if you are willing to pass the exam and become certified you can get info from I think they are the best in solution provider and study material on the net.
It is a widely believed that the more data you have, the more information you will get. This is a fallacy. It’s true that more data means more information, but that information is hidden and to recover it is anything but trivial. This was first realized in the 1990’s when all the large corporations of the world had mostly become IT savvy and were using high-scale efficient information management systems to record the operations of their businesses. All of their business units were recording sales receipts, purchase invoices, order items, financial figures, units sold etc. electronically and the data collected by these corporations was increasing exponentially. Indeed, the amount of raw data was colossal, but the nature of the data generated was operational and could not be used to make strategic decisions. This may sound absurd to those individuals who have just come to know the abilities of relational database management systems (RDBMS), but it’s a fact that in the early nineties, managers of large Corporations faced an unusual predicament, they had tons of data, yet they could not extract useful information from it that could assist them in making strategic decisions. This inability to retrieve useful information was termed as the information crisis.
Conventional database applications and strategic information: Now, why was there an information crisis and what is this strategic information? These issues definitely need further elaboration. Strategic information is the knowledge needed by manager and senior executives to formulate and implement strategies that will provide a competitively superior fit between the organization and its environment so as to achieve organizational goals; in short it is used to take strategic decisions, those decisions that will affect the long-term policies of the organization. It is used to develop business and operational strategies, set objectives, develop goals and scrutinize results. Strategic information is therefore, said to be the lifeline of any Corporation in today’s highly competitive ambiance. Consider the following questions usually asked by decision makers:
What if discontinue product A and start re-distributing product B in the Southern division, with the new tax laws implemented? How will this affect our total revenues? How much should the sale price be reduced to increase the sales percentage by 10 per cent in the North Eastern region? How can we retain the present customer base after the advent of two new rivals? What will be the impact of the sale price of our Products A, B and C, what if we change our approach from profit maximization to sales maximization? Tell me the names of problem regions (with respect to all possible parameters). Show the highest margins in comparison to our competitors.
Any experienced information system’s manager would have guessed it right away that such questions cannot possibly be answered by applications which were designed to store operational data such as recording an invoice, drawing an amount from the bank, settling an order, managing inventory items, etc.
The basic issue is that these systems were meant to record the daily routines of the businesses, not to provide such high level of information which itself is dependent on certain external factors as well. However, there are many specific shortcomings in the conventional information management applications which thwart any attempt to retrieve long-term strategic information. Before we find out what cons are associated with conventional system, let’s define what conventional system are.
A conventional system is nothing more than a slave and answers only those questions which you ask. It is not smart enough to point out the various anomalies or regularities reflected in the given data. Executives need patterns; they are looking for trends in the figures as these patterns suggest a regularity or irregularity that is crucial in healthy decision making. Here are some cons of conventional systems.
Disparate storage forms: The domain of information technology is undoubtedly the fastest evolving science mankind has witnessed. An application developed 2-3 years ago will now definitely be totally obsolete according to the standards of the applications developed today. The back-end database would have changed; the front-end programming language would have undergone fundamental changes. The data that has been collected during several years, although it is present is scattered across incongruent storage mechanisms, making it virtually impossible for anyone to analyze it as one unit. Thus, the applications developed by companies to record day to day information have highly diversified data structures which cannot interact with one and other to provide the much needed high level information
Event-driven nature: These conventional database applications are essentially event-driven. Information is available with respect to events such as sale, order, invoices, receipts, etc., not with respect to subjects such as product, region or revenue. A system that answers questions which begin with What ifâ€¦ need to be subject oriented rather than event driven. This fundamental characteristic of operational systems makes the task to extracting useful information challenging, not to mention impractical.
Complexity: The transaction processing systems usually face some restrictions vis-Ã -vis the time needed to complete a transaction. These systems have been optimized for transaction processing, not for answering complex queries. Hence, when such systems were grilled with composite queries, they were not able to cope up with the transaction time restrictions.
OLTP and DSS: It was clear that in order to provide information needed for in depth analysis, something else was needed. That something should be separate from what the straight forward transaction processing systems does. This need for thorough analytical information characterized the information management systems into two parts:
Online transaction processing systems: According to textbooks, OLTP makes the wheels of the business turn. They are needed to record all the operational information. They deal with events and usually revolve around a single entity such as sale receipts, cash transactions, invoices, etc.
Decision support systems: DSS watches the wheels of the business turn. They are designed to watch how the various operations of the business are running and are meant to answer questions related to strategic decision making. DSS extracts data from OLTPS and various external sources and keeps in a format suitable for responding to high level queries.
How DSS evolved
In the past the IT departments of many companies tried to fulfil the needs of their managers regarding more information in more suitable formats, but all their efforts have been more or less bit of a damp squib, yet they were significant in the sense that they actually led to the development of full-scale decision support systems.
Initially, for all the unconventional detailed reports the IT departments used to write programs that generated ad hoc reports for each request.
Globalization ushered in a new era in which each and every department was looking for reports generated in new formats, reports that were relative to some external benchmark. Since, it was not feasible to write programs for every such request the IT experts came up with special extract programs. These introduction of these programs was made as an attempt to generate reports that were anticipated. This was an ignorant attempt as mostly the needed information was unexpected, but data specialists were learning to store data in such a format in which it could be easily synthesized, that is all kinds of information could be easily extracted. Then in the 70s came the primitive forms of DSS, which at that time were called information centresâ€. They were the complex and highly generic applications which could instantaneously generate requested reports and special analytical information. After the era of information centres, we see the current DSSs which are under a constant stage of evolution.
Today’s DSS can be operated by a manager who knows nothing about databases; they are being designed as the essential tool required in fulfilling the compelling need for strategic information in the present competitive landscape.
Components of DSS: According to a study conducted by the University of California, 1-2 exabytes of unique information is produced every year worldwide and just in case you did not know one exabyte equals 1 billion gigabytes or 1018 bytes, hence making it virtually impossible to derive useful knowledge.
To tackle an issue such as this effectively, the various functionalities of the DSS are handled by different components and to even name all of them is perhaps easily beyond the scope of this article. However, if we broadly analyze the DSSs, they are all associated with two fundamental concepts: (1) data warehousing; and (2) data mining.
The implementation of these concepts has taken decades to mature but, as we shall se ahead their rudiments can still be easily grasped.
Data warehousing: It is the name given to the process of transformation and storage of data in a suitable format that is best for complex analytical purposes. Data warehouse is not an application, it is an environment, it refers to the warehouse in which all the operational data collected by OLTPS is residing, but prior to the storage of operational data into a data warehouse, the raw data goes through radical renovation.
Basically, a data warehouse helps in integrating, categorizing, codifying and arranging the data from various parts of the enterprise so that it can project a unified view of the organization and help in strategic decision making. It can be easily seen that this data is bears no resemblance to the nature of data we see in traditional OLTPS, yet the data is extracted from traditional database applications and fed to the data warehouses. This is what makes the implementation of data warehousing such a specialized task. Data warehousing provides the environment, in which a DSS can help make a strategic decision, but how well that decision is taken depends upon how intelligently the data in a data warehouse is exploited and this is precisely where the implementation of data mining comes in.
Data mining: Nowadays, most large organizations generates more information in a week than most people can read in a lifetime. We humans have several limitations, we can only know that which we want to know. So, what about the information we don’t know about, the unique knowledge of our own business which is hidden from us? What if it points to a unique buying pattern of the customers in the South Asian region? What if that hidden information uncovers an unusual link between your company’s line of products which can be exploited to achieve effective marketing. What if a concealed pattern is present in your corporation’s data warehouse, which can predict the number of production units needed in the next year with respect to the future population, employment and income statistics of the country?
Surely, no one will ever want to miss such valuable information, but it is humanly impossible to try to discover such patterns that are buried under mountains of data. This is where the concept of data mining steps in. Data mining just like mineral exploration, tries to discover hidden valuables. In both the processes, we are looking for something precious and we don’t exactly know for sure where we are going to find it, what we know that there is something precious out there that needs to be found!
There are many algorithms used in data mining tools such as decision trees, memory-based reasoning, neural networks, genetic algorithms, etc. All are highly advanced applications of computer science, so I shall make no endeavours to explain them! However, what must be known is that the serve a two-fold purpose.
1. Knowledge discovery, this is achieved by uncovering the relationships that exists in the pre-processed data of the data warehouse.
2. Future prediction by means of pattern identification.
Statistics and data mining pursue the same aim, which is to build compact and understandable models incorporating the relationships (dependencies) between the description of a situation and a result (judgment) concerning this description. The only variation being that a sophisticated data mining tool builds the relationship model much quickly which is generally more efficient as it is developed using algorithms which have been proven over and over again.
There are countless domains in which data warehousing solutions can be implemented and its full potential is still not known, but the IT specialists have learned that wherever it has been implemented properly it has produced significant benefits.
Data warehousing solutions has helped organizations keeping a very close and personal relationship with its customers. In DSS we refer to such applications as Customer Relationship Management (CRM) solutions, which specifically focus on customers needs by identifying relationships between products, customers and their spending habits.
DSS have successfully predicted which fiscal policies will help the corporation achieve its objectives in different external environments by analyzing past financial data. Also, DSS profitability analysis satisfies the need for detailed enterprise profitability reporting and analysis leading to better informed decisions which in-turn maximize profitability.
Through DSS, organizations have attained highly professional supply/demand chain intelligence which has significantly reduced costs at various levels in the industrial/management operations. Similarly, many other industries such as E-business, internet banking, targeting advertising etc. have generated more profits after adopting DSS.
DSS can help any corporation perform to the best of its abilities. It’s a proven fact that through professional data warehousing solutions corporations have reduced costs, achieved higher productivity, attained customer centricity and improved the overall organization of their internal operations. Still, there is a dark side of the picture too. Data warehousing is not something which any person/organization can provide as it is a highly specialized field and because there is no absolute solution to any decision support requirement; it varies from organization to organization. Furthermore, the users are usually unable to elucidate their requirements when it comes to specifying the information they wish to seek. That’s the reason why a popular DSS provider markets its products by asserting Data warehousing solutions that workâ€; indicating that if not executed correctly, DSS may also fail to achieve what was expected from it.
A true data warehousing solution is not just about computer science, it’s also about understanding the customer and foreseeing its potential requirements and companies which have realized this, like Teradata, Oracle and IBM have successfully provided DSS to numerous organizations. Corporations all over the world pay such companies millions of dollars for their data orgnizational needs. In the current knowledge economy, it is now an indisputable fact that information is the key to organizations for gaining competitive advantage and as more and more companies realize this, the better DSSs we will see in the future.
For those of you, who not convinced whether such theoretical concepts actually works, refer to the study conducted by International Data Corporation (IDC) which showed that data warehousing could provide significant and impressive Return on Investment (ROI) numbers. The study, which included 62 participants, demonstrated that the overall ROI on warehouse projects was 401 per cent with payback periods of two to three years. ROI, being a standard by which any investment is measured, suggests that data warehousing solutions is an extremely profitable endeavor. Established benchmarks have proved that the decision support systems are not just topics of post graduate research; they are highly practical and as more research and development takes place in this branch of computer science, experts predict that in the future all of us will rely heavily on DSS’s predictions!
If your preparing for career change and looking for MCTS Training or MCITP Certification the best online training provider that provide the all the and complete MCTS certification exams training in just one package, certkingdom self study training kits, save your money on bootcamps, training institutes, It’s also save your traveling and time. All training materials are “Guaranteed” to pass your exams and get you certified on the fist attempt, due to best training they become no1 site 2009 & 2010.
In addition I recommend Certkindom.com is best and No1 site of 2008 which provide the complete Windows Server 2003 certified professionals training, Microsoft MCITP, Microsoft MCTS, Cisco CCNA, Cisco CCIE, CompTIA A+, IBM, Citrix, PMP, ISC, and lots more online training self study kits, saving your time and money on all those expensive bootcamps, conventional training institutes where you have take admission pay fees first and if you don’t want to continue no refunds no transfer to any other training course, If you planed to take CCNA or specialization in MCSE 2003 all the process starts again; as for getting online training can be much beneficial and you don’t need to take for fill any from to switch your training on any desire certification
Comments are closed.