Gestión de implementación de sistemas de inteligencia de negocios
| Autor | Klaus Solberg Søilen |
| Cargo | Blekinge Institute of Technology in Sweden |
| Páginas | 47-65 |
Page 47
There has been much focus on why companies should invest in Business Intelligence (BI) and there has been many contributions on the BI process, but there has been and continues to be less written on how to implement BI
Page 48
in the organization (see e.g. Jaworski, et al., 2002). The findings presented in this article build on a decade of experience and research in the implementation of Business Intelligence Systems.
BI is a relatively new organizational function, and most companies will have no or little experience in its implementation. Also, there is no one correct organizational solution for all companies. Instead intelligence needs and ways of monitoring and handling intelligence are different at different levels in the organization, as suggested by the figure below.
[VER PDF ADJUNTO]
Fig. 1: Different BI needs in the hierarchy (Van de Wiel, 2005, p. 4).
Some of the terms used in the figure need an explanation. Executives tend to have less time to read long reports. Instead they often look for the essentials, the «executive summaries». For one thing they enjoy following KPIs. Key Performance Indicators (KPI) consists of financial and non-financial variables used to evaluate the performance of a company or a department. This may be seen on a so called Dashboard. A dashboard in software is a business management tool used to get an overview of the organization’s performance. Most dashboards use simple graphical presentations, histograms, pies, maps, all in different colors to indicate the score of different KPIs.
Page 49
The example below is from Kramerica Industries, a web design company for nonprofit organizations and small businesses, operating out of Flushing, NY.
[VER PDF ADJUNTO]
Fig. 2: Kramerica Industries Dashboard (2009-04-18, http://enterprise-dashboard.com/)
This dashboard is based on SQL Server relational, Analysis services and Cognos data sources and lets end users build their own views of OLAP data. On Line Analytical Processing (OLAP) is like a fast calculating, on line excel spread sheet which can be shared by many users at the same time. We speak here of databases configured for OLAP. Like in excel we have different dimensions in the rows and columns, and different values at their intersections. The most common use of OLAP in organizations is to present sales figures, budgeting, forecasting and financial reports. The further we go down in the hierarchy in the figure above, the broader and more extensive the information will become. Thus down on the specific departments we
Page 50
find the classical management reports and at the lower levels, production reports. Each form of information is suited to the performance at each level in the organization.
Business Intelligence has now become everyday life in many organizations. A decade ago, when the research started in this field, business intelligence and competitive intelligence were met with mostly skepticism in top management. Today it has been recognized that to stay in touch with the market it is necessary to have timely information about financial operations, customers, in fact about everything that influences operations. As our competitors are likely to have this ability and are making better decisions so we too need to have it, simply to stay in business. This trend or tendency is not only clear today, but increasing. The battle to become market leader started a few years back. In November 2007 IBM purchased Cognos for five billion dollars.
Gartner Inc., the very company which is accredited for having introduced the term «business intelligence», reported in 2005 that from 1.300 CIO’s (Chief Information Officers) representing more than 57 billion USD in IT spending, business intelligence is the number two technology priority in their companies. Executives are now pushing BI initiatives as never before. Further indications can be found in the InfoWorld Business Intelligence Report 2005 and Gartner’s report Business Intelligence Predictions for 2009 and Beyond. Among its predictions are:
· Through 2012, more than 35 per cent of the top 5,000 global companies will regularly fail to make insightful decisions about significant changes in their business and markets.
· By 2012, business units will control at least 40 per cent of the total budget for BI.
· By 2012, one-third of analytic applications applied to business processes will be delivered through coarse-grained application mashups.
The awareness of the value of useful and timely information has never been greater. We could speak of an Information Turn, a paradigm shift in the
Page 51
way we run our organizations. The focus is no longer on hardware, but software. It is no longer machinery, but on the grey cells. We are more than ever aware of the fact that information or intelligence is the basis for all good decisions, and that decisions making is what business is really all about. Consequently, organizations are going out of their way to build electronic systems to help them in this process (see e.g. Gangadharan, Swami, 2004).
When an employee at the Gartner Group started to use the term «business intelligence» in the late 80s he was mostly thinking about what we today call «competitive intelligence» and «market intelligence» (Jenster, Solberg Søilen, 2009), at least in the Anglo Saxon world. However, when we say «business intelligence» today most people think about the software applied, about computer systems and the technology used for private intelligence tasks. As always it is the use of the term which at the end decides its meaning. There is nothing new or shocking about this. The meaning of many words change over time, some quicker than others. Note also that «business intelligence» mostly applies to the technical solutions applied in the private sector, not in the public. For the public sector we often speak of «public intelligence». We also separate between «state intelligence» and «military intelligence», even though most of the methodologies are the same. What separate these fields of studies is first of all the means and the applications (Solberg Søilen, 2005).
Most professionals in the software business who produce intelligence systems talk about business intelligence systems. The reason why we are hearing so much about «business intelligence» today is because the technical parts of the intelligence function has become so more important for the way organizations operate. The reason for this has to do with the introduction of new technologies, at ever greater speed and at ever lower prices, first of all software solutions. In other words, there is a lot more technology in the intelligence profession today than what was the case only a decade ago when this research was starting to become popular.
The Business Intelligence System consists of an Organizational structure and a Infrastructure, predominantly IT related, with a combination of se-
Page 52
lected hardware and software. Decision Support Systems (DSS) is another broader term used. DSS makes no discrimination as to the nature of the data entered and is helpful sorting and organizing large amounts of quantitative data. Consequently DSS technologies are used in all sorts of companies at all levels, from sales in IT companies to the status of patients in hospitals (see e.g. Sauter, Free, 2005). We also observe that the number of standardized systems is increasing. We see fewer BI systems which are being made in-house. Instead operational BI is increasingly bought by the big vendors, from companies such as the SAS Institute Inc., Information Builders Inc. and Cognos Inc. (Havenstein, Heather, 2005).
The real big dream for these companies and their customers is to develop a real time business intelligence system (Watson, et al., 2006). To take an example; we want to know what the sales in our store is right now, how many customers have entered, how much bread (if a bakery) has been bought and therefore how many needs to be produced. In many cases this kind of system it is already a reality, at least where processes and analyses can be standardized. The more difficult the analysis is, the more valuable the information will be and the longer it will take to come up with an answer. Good analysis demands time for reflection. Computer ability will help in this process, but is not yet ready to substitute the work of the intelligence analyst.
Managers want to know where the Intelligence function should be placed in the organization. There are many opinions about this, and companies have been trying out different versions and collecting experience for a number of years already. The list presented below is the joint summary of some of these experiences building on the theory available in the field.
As the intelligence function sprang out of the marketing function, and more particularly from the marketing research...
Para continuar leyendo
Solicita tu prueba