|Sign-Up for Free Exclusive Services:||Portals|||||eNewsletters|||||Web Seminars|||||dataWarehouse.com|||||DM Review Magazine|
|Covering Business Intelligence, Integration & Analytics||Advanced Search|
The Intranet Data Warehouse:
Since the first decision support systems came into existence in the early seventies, software vendors have focused most of their attention on meeting the needs of a relatively small number of technically competent "power" users. These users spend hours, days or weeks analyzing important strategic business issues. They require access to huge amounts of raw data, often culled from multiple sources, and robust analytic tools to transform that data into informational content--primarily reports and graphs.
Although early mainframe-based decision support tools provided powerful analytic capabilities, they were complex and difficult to use, seldom reaching beyond the core group of power users within an enterprise. In the mid eighties, vendors began to introduce client/server technologies that extended decision support capabilities to users who were somewhat less technically knowledgeable than the traditional power users.
Second-generation decision support tools replaced the often complex menus of terminal/mainframe computing with GUIs and gradually resolved many of the ease-of-use issues associated with mainframe-based systems. Two of the major features of second-generation decision support tools--on-line analytical processing (OLAP) and data warehousing compatibility--significantly improved users' ability to acquire and analyze huge stores of information. Despite these important milestones in the evolution to client/server architecture, cost and user support issues have continued to restrict enterprise-wide adoption of decision support tools, preventing all but a relatively small group of users from taking full advantage of the technology.
Today, companies are beginning to recognize that by adapting Internet technologies to build intranets and extranets, they are creating a "universal client interface" that can provide access to the broad spectrum of the enterprise's information resources. The universal client interface, which combines browser, search engine and push technologies, offers very real advantages in terms of ease of use and cost-effective deployment. Companies are now exploring ways of using that same interface to expand user access to data warehouses and OLAP capabilities, thereby realizing the full potential of decision support technology.
After spending more than two decades focusing on the needs of the so-called power users, many decision support software vendors are beginning to address the information needs of a far larger user community--the front-line managers who are responsible for making the day-to-day tactical decisions that maintain the strategic course laid out by senior management. Addressing these users' information needs, however, requires combining the vast stores of information residing in data warehouses with OLAP capabilities and then deploying the information and analytical capabilities to the relevant decision-makers--wherever they are within the enterprise. Here again, companies are looking to extend lessons learned from the Internet to cost-effectively deploy the information and analytical capabilities in a timely manner and in a format that is useful to decision-makers.
The computing environment has changed dramatically in the past decade to meet the changing needs of its user base. The two-tier client/server systems of the past are no longer adequate to provide the flexibility and scalability required by today's complex business environments. Three-tier client/server architectures, Web-based computing and distributed-object technology are all good examples of technology advancements geared toward delivering processing power and information to the user. Data warehousing and OLAP are part of this change, but unfortunately in many cases they have failed to live up to their full potential.
The unfulfilled promise of data warehousing and OLAP technology is that the technology, once ease-of-use issues were resolved, would be broadly accessible and useful to decision-makers throughout an enterprise. This is hindered because organizations often overlook the cost of deployment and the support burden. The Internet appears to be a logical solution to these problems of deployment; but, all too often, software vendors have merely modified the analytical tools embraced by the power users, refitting them with graphical interfaces and declaring them suitable for all types and levels of users. In reality, most decision-makers within an enterprise have very different information and analysis requirements than the traditional power users. Power users typically demand application features for in-depth strategic decision support while the majority of users want information content with which to make tactical decisions.
Front-line managers need tactical decision support to wage an effective battle in today's competitive business environment. In any large corporate environment, hundreds of managers make thousands of tactical decisions every day--both within the enterprise walls and beyond. These decision-makers are not generally interested in spending time in unstructured searches for answers to business issues. They are concerned with answers to specific queries, such as initiating special pricing or rebate programs in response to a decline in product sales. In short, front-line managers need immediate access to actionable information--content, rather than features.
Feature wars have driven the software industry for the past several decades. In response to competitive pressures in the PC software arena, software vendors have become accustomed to packing all possible features into their application products, with remarkably little concern for the needs and wants of the majority of users. For too long, many software vendors have emphasized features over usefulness, developing and distributing products that are not well targeted for user needs.
Feature-laden products designed to "do all things for all users" can cause problems in a number of areas. Such packages are, for example, generally more difficult to distribute because they typically require multiple disks or CD-ROMS; and many non-technical users find it difficult to install such products. These applications also need an increasing amount of space to support their functionality, thereby requiring more powerful, robust, costly PCs. Support costs are generally high for these products since users seldom use all of the available features. In fact, studies indicate that most users take advantage of only about 20 percent of a product's features. Clearly, the major problem with feature-laden software products is that users have difficulty learning to use these features effectively and can only create useful content after they have learned to select and manipulate the features that are most meaningful to them.
In many respects, Internet technology is affecting the way users view their computers, resulting in a change in how software vendors develop and distribute application software. The Internet's "universal client interface" emphasizes content over features, hiding the logic components of application software behind the content. In effect, the content itself becomes the interface, leaving the user blissfully unaware of the complexity of the application operating "behind the scenes." While this fundamental shift in the user interface software development model may seem obvious, it has some far reaching consequences for deploying application software throughout an enterprise.
Savvy IT organizations are beginning to use Internet technologies to deploy their business applications, including OLAP, throughout the enterprise and beyond. Such deployment, requires a thorough understanding of the target users and their specific decision support requirements. Merely replicating the look and feel of an existing application in a browser interface and deploying it through the Web is not the means to achieving broad-spectrum decision support. Instead, IT managers and software developers must focus on providing content-centric applications for effective tactical decision support, recognizing the information and analysis requirements of the new community of non-technical (casual) users that the Web allows them to reach.
The Internet is teaching many valuable lessons to software developers and users alike. Perhaps the most dramatic is the shift in importance from an application's interface (the "look and feel") to its functionality (the ability to deliver content). A data warehouse that is accessible via the Internet or a corporate intranet can expand decision support capabilities to the front-line managers--that large community of casual users that need ready access to information resources and analytical tools for day-to-day tactical decisions. However, in deploying data warehouse and OLAP capabilities over an intranet, it is essential to mask the underlying complexities of the applications from the users to ensure the communication of meaningful information is helped, not hindered, by technology. The information that a front-line manager retrieves from the data warehouse should look like any other resource accessible via the Web browser. In essence, the user interface that has been the key in determining an application's ease of use becomes largely irrelevant in the content-driven world of Internet/intranet technology.
Many companies are now using intranet technologies to deploy data warehouse and OLAP capabilities throughout the enterprise. As a next step, many--recognizing the value of business partnerships--are establishing extranets to provide external business partners (suppliers, wholesalers, retailers and affiliates) with secure access to their data warehouses and OLAP tools. This type of extended decision support will enable mutually dependent organizations to share information resources and identify new business opportunities. Such access clearly raises questions of privacy and security. Once again, organizations need to carefully consider the information and analysis needs of the target users before beginning to deploy decision support capabilities over an extranet. The information and analytical tools useful for strategic decision making are not suitable for sharing with external business partners that need support for tactical decision making. Suppliers, for example, may need to access last month's sales figures to determine appropriate shipping quantities for the next month but should not have access to long-term sales or revenue projections. External business partners, such as front-line managers, have little need for powerful, ad hoc analytical tools that are the mainstay of strategic decision support.
Worldwide organizations like Gap Inc. conduct business with a variety of partners. Each planning season retailers need to determine shipping quantities, assortments, item replenishments and more. They need to respond to the following types of questions:
The company is being positioned to provide these external business partners with the critical information necessary to make timely, effective decisions. The Web is a viable way to deliver information for everyone involved in the decision-making pipeline. Using OLAP technology over the Web and its push technology through agents, managers at the front line tasked with decision making are directed to top priority items that are tied to specific business rules for their immediate attention.
The Web assists with addressing the challenges that are inherent to software distribution in a non-centralized type of environment prevalent in many organizations today. The Web browser supports interactive OLAP reporting for a worldwide organization that has numerous suppliers around the world, regardless of the platform they are running. OLAP analysis training is minimized by the use of familiar Internet functions such as hyperlinks for drilling and Internet native navigation.
There is little argument that intranets can serve as incredibly valuable sources of information for corporate decision-makers. Without navigation tools or direction, however, intranets can be as daunting as the Internet itself, presenting unwary users with thousands of documents pertaining to virtually any aspect of the organization. Decision-makers, faced with such a glut of information, must be able to quickly find the specific resource they need and determine if it is current and if it correctly reflects the rules or goals of the business.
Many organizations, including Gap Inc., are turning to "push" technologies to resolve these important questions. Push technologies enable businesses to selectively disseminate critical information to decision-makers via the corporate intranet or Internet. They can, for example, disseminate information pertaining to inventory counts, sales history or other important metrics based on schedules or specific events (referred to as triggers) recorded on the intranet or Internet. The goal of push technology is to present end users with information that is directly relevant to their tactical decision making, emphasizing high-priority items and eliminating the need for users to "surf" through all available information in hopes of discovering some relevant fact.
Push technology enables an organization to seamlessly expand its decision support capabilities beyond the enterprise strategic decision-makers and tactical managers to a range of suppliers. Because suppliers use a variety of computer platforms, an intranet is critical to deliver the information necessary to make sound, tactical decisions regarding production regardless of their desktop.
The concept of subscription is critical to organizations like Gap Inc. Intranet users can subscribe to a collection of content information, much like subscribing to periodicals that are delivered via mail. Defined by the individual, workgroup or location, a user can manually or automatically subscribe to receive relevant content/information found within the enormous repository of available information sources on the intranet, bringing critical issues to the forefront for immediate action. For example, suppliers could automatically receive items pertaining only to the specific items they manufacture. Critical orders and product design changes are immediately brought to their attention by pushing this information automatically to their browser. However, suppliers would not have appropriate security clearance levels in the intranet to surf the data warehouse. With push technology they are automatically alerted to any problems that need their attention.
Push strategies are well suited to Gap Inc.'s implementation of data warehousing. The company uses push technology to disseminate trend reports and sales history provided by their OLAP application to various subscribers within the enterprise; these users understand the information being presented and can suggest possible solutions to address a business rule that indicates a problem or missed service level. With push technology, an organization's knowledge workers do not have to sift through e-mail, hard copy reports or other sources to locate information that is of interest to them. Having immediate access to relevant information helps them make timely, effective tactical decisions about their production lines.
In reality, the so-called push technologies actually include offerings that encompass both push and pull mechanisms. Software agents can be pre-programmed to track and send information from the data warehouse to desktop and/or server environments. Organizations can specify business events as triggers to activate the delivery of information (in the form of charts or reports) to users to, for example, notify relevant business managers of meaningful trends.
Push technology is a natural extension of knowledge dissemination for data warehousing and should be considered an extension of its comprehensive architecture. When Web-based solutions, event notification and intelligent agents are combined with data warehousing and OLAP capabilities, they can significantly expand the scope of an organization's decision support system, providing the front-line "knowledge workers" with vital information for sound tactical decision making.
As organizations move toward the intelligent enterprise approach of information sharing and delivery, they are indeed fulfilling the promise originally made by the data warehousing and OLAP communities to deliver broadly accessible and useful information to decision-makers throughout the enterprise. With the evolution of the Web, followed by subscription and push technology, organizations are beginning to realize that business intelligence solutions can no longer be selected based exclusively on features. Rather, the selection should focus on the ability to deploy the knowledge-filled data warehouse to the masses.
Phil Wilkerson has over 18 years of software engineering experience. The director of technical architecture at Gap Inc., Wilkerson is an advocate of de-facto and open standards and speaks at industry trade shows such as RISCON, Comdex and Retail Systems Alert.
Richard Tanler is the founder of eiVia Inc., the leader in providing technology that the company refers to as an ON-Demand Spreadsheet. Tanler was the founder of Information Advantage, Inc. a leader in the business intelligence market (acquired by Computer Associates) and is the author of the Intranet Data Warehouse (John Wiley & Sons) and numerous articles on business intelligence and data warehousing. You can contact him at firstname.lastname@example.org.
|View Full Magazine Issue|
|E-Mail This Article|