We're all familiar with the growth rate of the public Web. Regardless of the metric used to measure its growth ­ attached networks, servers, users or pages ­ the growth rate continues to exhibit an exponential pattern. In the same vein, the adoption rate of intranet and extranet data warehouses (i.e., Web warehouses) has exhibited a similar pattern, although the pattern has lagged public adoption. While data warehouse and business intelligence vendors have offered Web-enabled versions of their tools for the past three to four years, it was not until the last year that the sales of their Web offerings began to substantially exceed the sales of their traditional client/server versions. However, the patterns of adoption differ by more than a simple lag.

To date, the usage of Web warehouses has been business as usual with increased efficiency as the primary focus. The public Web has been anything but business as usual. Not only has the public Web proven to be an efficient medium for business communications, but ­ more importantly ­ it has also engendered an amazing array of new business forms that are changing the overall structure of the global economy. Is the same sort of evolutionary pattern in store for Web warehouses? Will the adoption of these technologies lead to substantial changes in the way business is conducted within and between enterprises? Not likely, unless we understand the differences in the two patterns and consider the range of possibilities for Web warehouses.

Nine years ago Thomas W. Malone and John F. Rockart, best known for development of the critical success factors method for information and information systems planning and for initiating executive support systems, documented a three-stage pattern in the adoption of number of new technologies.1 The stages reflected "three orders of effects of decreasing costs" and included:

First Order Effect: Simple substitution of the new technology for the old.

Second Order Effect: Increased demand for the function(s) provided by the technologies.

Third Order Effect: The rise of new technology-intensive structures.

As one example of this pattern, Malone and Rockart pointed to the adoption of transportation technologies in the older industrial economy. The first order effect was simply substituting the new for the old ­ trains and cars for horses and buggies. The overall effect was a reduction in transportation costs. With improvements in these technologies came the second order effect ­ a substantial increase in the amount of travel and the distances traveled. People began to commute to work, to travel longer distances to business meetings and to vacation at far away places. Then came the third order effect. These new transportation technologies resulted in new transportation-intensive social structures and organizations such as suburbs and malls that were impossible with the older technologies that they transplanted.

The adoption of Web technologies for e-commerce and e- business is undergoing similar stages. Consider, for instance, online stock trading from the consumer's standpoint. The first order effect has been a simple substitution of online trading for broker trading with the attendant reduction in transaction costs. An average online trade is between $10 and $20, while trading through brokers is at least five times as much. The second order effect has been an increase in the volume and number of individuals involved in trading ­ day trading and after-hours trading (when the major markets are closed). Finally, online trading along with encouragement from the SEC is hastening changes in the overall structure of the exchanges as we know them today. Electronic communication networks (ECNs) ­ that match buyers with sellers of stock finding the best possible price for limited orders ­ already handle about 25 percent of NASDAQ's trading volume and are gaining more ground with the push for after-hours trading. Both the NYSE and NASDAQ are in talks with the ECNs. An alliance between these entities could result in NASDAQ stocks being traded on the NYSE and vice versa. The end game is likely to be a global electronic exchange, accessible over the Internet with trading on a 24x7 basis.

Can we expect similar effects on the structure and organization of enterprises from Web warehousing and other intranet and extranet applications? To date, most enterprises have experienced the first and second order effects. Virtually none have encountered third order effects.

Initially, the driving force behind Web warehousing was simple substitution with the expressed aim of reducing costs. The key phrases used to promote the replacement of more traditional client/server applications with Web-based applications were (and still are) lower total cost of ownership (TCO) and zero administration. Relative to client/server applications, the per seat costs of Web-based applications are less; although in the case of Web warehouses, the savings have been less than expected because the costs of concurrent database "port" licenses have remained high. Similarly, training costs have been reduced somewhat because Web applications are easier to learn for the average user than client/server applications. However, the real savings have come from the reduction in administrative costs. Web warehouses no longer saddle administrators with the task of distributing upgrades, patches and applications to individual desktops. Instead, the changes take place on the server and are automatically available to application users, thus reducing software distribution and data synchronization costs.

Second order effects of increasing demand for the function(s) provided by technologies are also being experienced by a number of enterprises. For most enterprises, the audience for client/server data warehousing has been restricted to, at most, a few hundred users often located at a central headquarters facility. Wide-spread usage of client/server applications has not only been hindered by application costs but also by network costs. By relying on the public Internet as if it were a private network and focusing on HTML production reports (rather than interactive analysis), enterprises have been able to substantially broaden the recipients of warehouse analysis from hundreds to thousands and to deliver this analysis across the globe to users both inside and outside the enterprise.

Increased demand is also being created as Web warehouse tools and applications are morphing into enterprise portals and analytical applications are being extended to transactional systems. Portals combine structured and unstructured data from a variety of sources including data warehouses and are designed to bring this information to all the desktops throughout an enterprise. In the same vein, the audience for Web warehousing has expanded as analytical applications have rapidly moved (indirectly) into the transactional world of enterprise resource planning (ERP), supply chain management (SCM) and customer relationship management (CRM).

Although the first and second order effects of Web warehousing are substantial, it is still business as usual ­ just more efficient. The fact that a data warehouse is Web-based hasn't changed its primary function. Web warehouses are still used primarily for reporting and analysis in support of decision making. There has been little impact on the way in which enterprises use the data, and there has not been a concomitant change in the structure or organization of most enterprises.

A number of benchmarking studies, such as those from the Hackett Group, have highlighted some of the inefficiencies surrounding management reporting and the use of management reports. As these studies indicate, most of our efforts are expended in gathering the data on which the reports are based, not in actually analyzing the results. In many cases, the analysis that is performed still requires eyeballing and manually navigating through an excruciating amount of detail in search of patterns of interest ­ even when those reports are color coded to highlight significant variances and exceptions. Web warehouses can inadvertently result in much wider participation in "value-diminishing," rather than value-added, activities.

What are some of the third order effects ­ the new technology-intensive structures ­ of Web warehousing that are being realized by a handful of enterprises? Three somewhat interrelated activities immediately come to mind:

Extranet and Consumer Ware-houses and Data Marts: Just as the airline industries recognized a number of years ago, some companies have recognized that there is market value in their warehouse data. For instance, a supplier of industry-wide insurance claim data provides extranet access to auto insurance companies so they can benchmark their claims- resolution performance. While this is a reporting and analysis application, applications of this sort can eliminate a number of steps in the data supply chain and bring customers and consumers closer to the enterprise.

Management by Exception: Manage-ment by exception relies heavily on computer systems to sift through masses of routine data and events in search of exceptions. Only when exceptions occur are organizational participants notified, thus freeing them from tasks that are value diminishing. Even the client/server versions of most data warehouse and business intelligence tools have alerting capabilities. Very few enterprises have taken advantage of these capabilities. The one exception has been the financial services industry, where alerting has been used to eliminate superfluous processes and organizational layers.

Automated Supply and Value Chains: Most of the reports garnered from Web warehouses end up in a browser where an end user is required to review and analyze the data. Once analyzed, recommendations are made and actions performed. In the world of e-commerce, some companies have begun to short-circuit the process, eliminating the need for manual review. For instance, some e-commerce operations are feeding clickstream data to their data warehouses where it is automatically analyzed. The results are fed to downstream processes that utilize the results to create dynamic, one-to-one marketing programs for Web site visitors. In the future, the adoption of XML as an interchange standard will simplify the process and encourage the elimination of manual reporting and analysis.

Such third order effects signal profound changes in the ways we will use and think about the Web for business. Ann Winblad, of the high-tech investment firm of Hummer Winblad Venture Partners, in an interview conducted at the beginning of last year observed: "As we enter the next millennium, rather than automating processes, companies are looking at who even belongs in the process. We'll see a change from automating processes to changing the process themselves."2

The same thing can be said of Web warehouses. Instead of substituting one form of reporting and analysis for another, enterprises should consider the possible third order effects that can be realized by eliminating those processes through a combination of data warehouses, analytical applications and Internet technologies. Only then will the power of the Web be truly exposed.


  1. "Computers, Networks and the Corporation." Scientific American. 1991. P 92-99.
  2. King, Julia. "Into Thin Air." ComputerWorld. January 4, 1999.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access