Business Information Goes Self-Service
The advent of the World Wide Web has transformed information from a scarce commodity, available only from "experts" or through painstaking research, into an easily accessible resource, available to anyone who knows how to use intuitive, graphical browser technology. Not only is the information on the Web generally free; it's also indexed, searchable and organized by subject matter, enabling users to quickly "surf" through related information stored on the network via hypertext links.
However, when the same Net-savvy user goes to work, he or she finds a very different situation. Critical business information needed to make decisions is frequently not available in a timely and efficient manner. Users may not even be aware of what information exists or what is available to them: The typical large corporation maintains hundreds of separate data stores (such as databases, ERP systems and data warehouses), and runs applications that generate tens of thousands of different reports in a wide range of formats. And that's not to mention the information being generated every day by end users in the form of word processing documents, spreadsheets, groupware, e-mail or output from analytical tools (such as OLAP cubes).
As organizations increasingly rely on consolidated business information for strategic -- rather than merely tactical -- decision making, they face the critical task of getting that information to end users. As a result of downsizing and process re-engineering, managers at the middle and lower tiers of management have gained increased responsibility for making important business decisions. Complicating the situation further, businesses are finding that, as they optimize the supply chain and streamline operations, they must also share information in a timely manner with users outside the firewall: subsidiaries, distributors, suppliers and even customers.
Unfortunately, despite heroic and often costly efforts over the past several years to consolidate and leverage corporate data in data warehouses and datamarts (and to provide analytical tools) most end users still have limited access to the information that could help them make decisions more effectively.
Users require self-service, personalized access to enterprise-wide business information, just as they've come to expect on the Web. The technology that will meet these needs, analysts agree, is the enterprise portal, a customized, browser-based single point of access to the entire corporate business information infrastructure. Through enterprise portals, users can "help themselves" to myriad reports from data marts, real-time feeds, Web content, text documents and more, quickly navigating to the specific information they need and making informed decisions on the spot.
From Data to "Decisional" Information
Organizations have always faced the challenge of transforming raw data into information useful to business decision-makers. This process occurs in two distinct steps: 1) the transformation of raw data into business information (such as reports, spreadsheets, documents, etc.) through the application of business logic, and 2) the consolidation of information into "decisional" information -- the specific figures, text, charts, etc., culled from all the various business information sources that bear on a specific decision by a specific user.
In the mainframe days, the process of transforming data into decisional information was laborious to the extreme. After the raw data had been entered into the system, business logic was run against it to produce the infamous green-bar reports, which were then shipped (occasionally by the multiple box-load) to the lucky recipients. In order to transform the green-bar information into decisional information, the user would then sift through the hundreds or thousands of pages in search of the specific numbers he or she required, and compare them to whatever other information he or she had managed to collect from other sources.
In the distributed environment, the process has grown vastly more complex. Data collected by different applications exists in databases throughout the enterprise, in a wide range of formats. In particular, the widespread adoption of ERP applications has resulted in "information silos" that represent discrete datastores, each with its own embedded business logic. While reports can be generated from these applications, transforming the results into "decisional" information is problematic: Individual reports may offer incomplete information, and similar figures may not be comparable across applications, depending on the standards used to input the data and the business logic used by the application to generate the figures from the data.
In recent years, IT organizations have made efforts to consolidate and unify this disparate corporate information in data warehouses and, more recently, data marts, which are miniature data warehouses focusing on a subset of corporate data. Unfortunately, extracting data collected by the various applications from the underlying business logic has proven extremely difficult in practice: "Scrubbing" and re-organizing all the data is complex, time consuming and extremely costly. Even then, the data in the warehouse or mart must be transformed into information using analytical tools, such as OLAP, which require extensive training to master. As a result, only a minority of users is able to generate information directly from the data warehouse or mart; the others must rely on IT to do the work for them -- creating yet another IT burden or backlog.
Empowering End Users
Currently, only a very small subset of users -- "power users" and IT administrators -- are able to extract decisional information from the corporate datastore (databases, ERP systems, data warehouses and data marts) by running queries to generate custom reports, or by using analytical tools. For the others -- 95 percent of all users at a typical corporation, the information is not conveniently available.
Even the fortunate five percent, the power users who can generate their own business information, still lack full "decisional information," because they have no access to "unstructured" data -- the data that lurks outside of "structured" datastores in end user documents, spreadsheets and e-mail, as well as in live feeds, video and audio format. Nor do they have a mechanism to correlate reports across applications, or to determine what other structured data might be relevant to the particular "decision" they are researching.
Corporate intranets may make it somewhat easier to find relevant information. However, the typical corporate intranet maintains only a small subset of information; users must navigate to multiple internal Web sites (each with its own security) to find information; and the information tends to be poorly indexed, unsearchable and not standardized among the various departments and business units. Even when a user has found the information, he or she has no ability to drill down to detailed information, update the data, create new reports synthesizing the information or publish results in a format accessible to other users.
The problem becomes even more acute when the information recipient is a third party. Many companies must regularly collect, consolidate and distribute business information to distributors, partners, subsidiaries and customers, resulting in a significant burden to IT.
The Enterprise Portal Model
In order to offer end users true decisional information, organizations are moving toward a new technology layer that unites all business information technologies under an intuitive GUI: This is the enterprise portal.
The enterprise portal offers a Web-like solution to the problem of distributing business information and consolidating business intelligence objects (reports, documents, spreadsheets, data cubes, etc.) generated anywhere in the enterprise by making them easily accessible to non-technical users via standard browser technology. The portal serves as a window, providing transparent access to "information objects" generated by various applications and stored in a central repository. The portal also provides access to analytical tools, and to the applications themselves for "dynamic execution" of pre-defined reports or generating new reports.
The key features of enterprise portal technology include:
Scalability. An effective enterprise portal solution must be supported by a multi-tier, distributed architecture in order to scale effectively. The portal must scale to support a huge number of discrete information objects without affecting availability or response time -- a significant requirement, when companies may generate a hundred thousand reports in a year, and when individual reports can consume hundreds of megabytes. In addition, the portal architecture must be capable of scaling to support a large number of users (up to hundreds of thousands) and applications on a wide range of platforms, from mainframe hosts and UNIX and NT servers to desktops, mobile PCs and hand-held devices.
Search/Navigability. Like popular Web portals, such as Yahoo, an enterprise portal should offer multiple ways to identify potentially valuable information, including a search engine for text-based searches, an indexing system that is standard across all information objects and hypertext linking within documents to enable "jumps" to associated information objects. For example, a report showing sales for a particular product might contain links to a Word document detailing sales strategy for the year, and to an Excel spreadsheet with a hard-dollar forecast for the quarter.
Security. One of the most crucial features of the enterprise portal is enforcing security for the hundreds of thousands of information objects that can be accessed. Most business intelligence objects are assigned security levels when they are generated by the business application, so to avoid undue administrative overhead, the enterprise portal must be capable of plug-and-play integration with existing authentication schemes. Security enforcement at the portal enables users to log in once to access all business intelligence objects for which they are authorized. It has the additional effect of increasing customization and ease-of-use.
Dynamic Execution. The portal must provide the ability to execute reports off of production systems or databases in real time, giving users access to the most up-to-date information. Moreover, this must occur transparently to the user. For example, when a user clicks on a report object, that report might update automatically from the underlying application and display the most current data. Enterprise portals can even serve as a user-friendly front-end to business applications, enabling the user to generate a report from, for example, Oracle Financials based on queries formulated in the browser window, and delivering that report back to the user via the browser.
Ease of Use. Perhaps the greatest benefit of the enterprise portal is its ability to make decisional information available to untrained users. Enterprise portals employ standard browser technology as the interface, eliminating the learning curve for anyone who has ever surfed the Web. Effective portal technology meets users at their skill level, providing a basic interface with simple choices for the novice, and a range of interactive business information analysis and reporting tools for power users. The portal should enable intelligent report viewing, formatted in a static report, but offering interactive features, such as drill down and hypertext linking.
Ease of Administration. Ease of use must also extend to the tools used by administrators and "power users" to create reports, to file and index business intelligence objects in the repository and to administer the repository. To be effective, the enterprise portal must also leverage existing data and existing business logic, rather than require the company to rebuild their datastore, as with data warehousing technology.
Extranet Support. A crucial feature of the enterprise portal is its ability to function on either side of the corporate fire-wall. This enables a corporation to effectively open its business intelligence object repository to mobile users, customers, distributors, subsidiaries, partners and any other parties who need access to company information (subject to security clearance, of course). In addition to reducing administrator costs (since the outside companies are retrieving their own information), this free exchange of information is a key enabler of corporate electronic commerce. Third-party users must also be able to administer their own users -- for example, assigning and deleting passwords.
Personalization/Customization. In order to offer true self-service access to business information, the portal should permit customization by the end user, including the arrangement of the browser (incorporating real-time Web feeds, headlines, notification of report availability, etc.). The portal must permit both "push" and "pull" report distribution, so that users can "subscribe" to information based on interest or by "exception" (for example, receive an inventory report when levels of item X fall below a user-defined level), or call up information from the repository upon request.
The Benefits of Self-Service BI
The explosive growth of the World Wide Web demonstrates the immense value of information on demand, even in such everyday decision-making tasks as shopping for a new stereo. In the corporate environment, the advantages of giving users fast access to decisional information are even greater:
Increased Business Efficiency. The less time users spend looking for information, the more time they're spending on income-generating activities, and the less time administrators must waste assisting them.
Increased Productivity. Better information leads directly to improved decision-making, which has a clear effect on the bottom line. In addition, the ability to make quick decisions can often mean the difference between making a sale or losing it. Some organizations are even selling business information back to their customers, making the enterprise portal a profit center in its own right.
Reduced Costs. The self-service portal model reduces costs by offloading administrators from spending hours generating reports and documents for end users.
End User Empowerment. While empowering end users has ancillary benefits, such as reduced burden on IT and better decision-making, it also has a positive effect on the users themselves, increasing their confidence and independence while reducing frustration.
New Perspectives on Information. When users have easy access to information, they frequently hit upon solutions or insights that would otherwise have eluded them -- again, with a clear impact on the bottom line.
E-commerce. The ability to efficiently and securely share data with third parties enables increased efficiency in the supply chain, as well as improved relationships (and easier, faster negotiations) with suppliers, distributors and customers.
These advantages, combined with the proven effectiveness of portals in the Web environment, demonstrate why enterprise portals will soon be considered a necessity by corporations seeking a competitive edge and a knowledge-empowered employee community.
John Schroeder is Executive Vice President of Products and Services for Brio Technology, where he is responsible for overseeing the company's technology strategy, development, and services and support. He can be reached via e-mail at jschroeder@brio.com.
